Just because your overall cost per sale is £6, it doesn’t necessarily follow that every sale you make is profitable. Consider for example, your brand name. Chances are, the cost per click on your name is much lower than for the rest of your account, and there’s a good chance that the conversion rate is higher as well. So what? Consider the impact of splitting this data out:

Your performance looks good – you’ve made £5000 (from 500 sales making £10 each) and only paid £3000, so you’re £2000 in profit. But what happens if you split it into brand and non-brand?

Interesting… The sales via your brand name have cost £2 each, but your non-brand sales have been costing £10 each. It appears that non-brand conversions aren’t profitable at all!

However, it gets worse. Overall, your non-brand sales may be breaking even, but if you’ve ever tried to generate more sales by increasing your budget, you’ll know that this is an over-simplification.

Any time that you increase your bids to generate additional sales, your cost per conversion will increase sharply. This is because you will not just be paying more for the incremental clicks that you pick up, you’ll also be paying more for the clicks that you were already getting. Conversely, reducing your bids will reduce your CPA disproportionately. Consider the ‘Other’ campaign above. What would happen if you were to reduce the cost per click to £0.20 (from £0.25)? It’s impossible to tell for certain (since you don’t know whether you’ll drop one place or three, or what impact this will have on your click through rate), but for the sake of argument, we’ll say that clicks also fall by 20%, and that conversion rate doesn’t change.

Your CPA has also dropped by 20% (obviously, since your conversion rate hasn’t changed, and your CPC has dropped by 20%), but if you look at the figures more closely, there’s a bit more to it than that.

By paying £0.25 per click instead of £0.20, you’ve increased your click spend by £900. But if you look at the sales volumes, you’ve only got 50 more. The additional sales have cost you £18 each – and they are only worth £10!

This is the basis of sweet-spot analysis – one way to identify the most profitable position is to estimate the cost per click at which the incremental cost per sale becomes unprofitable.

In fact, extending this analysis further, you’d probably want to reduce your bids to just 12.5p – halving them:

By reducing your bids by half, you’re now making £625 in profit on this campaign (125 x £5), but even then, you’d be missing a trick.

The performance of the different Ad Groups within the campaign may be completely different, so in the same way that your overall figures masked the fact that brand was propping up non-brand, some Ad Groups may be masking the underperformance of other Ad Groups. Continuing the example from above, suppose that having reduced the cost per click to 12.5p as above, you see that your two Ad Groups are performing as follows:

Neither of these bids is optimal – increasing bids on Ad Group 1, and reducing them on Ad Group 2 will result in higher profits on both:

Theoretically, you could extend this analysis even further, down to keywords, though I would argue that if the performance of keywords in the same Ad Group varies, why are they in the same Ad Group (surely the fact that their conversion rates are different should indicate that the people searching are different, and may respond to different adverts).

There are a number of flaws with this type of analysis. Firstly, we haven’t taken into account the average order value. If Ad Group 2 had a much higher AOV than Ad Group 1, then you’d want to pay more per conversion. If this data is available, then you should be using it to calculate the profit for different levels of bids, rather than the overall figure.

Secondly, there’s no way to predict accurately the impact on clicks of increasing or reducing your bids. Since this is the case, I would advise against making dramatic changes to your bids – changing them slowly over a period of time will allow you to assess the impact on clicks with more confidence, and reduce the risks inherent in basing your bids on estimates.

Another concern is the conversion rate. Most of the time, the conversion rate isn’t affected that much by the position in the search results (something Hal Varian at Google acknowledged in 2009), but there are cases when the conversion rate can be affected significantly. In these cases, the effect of bid adjustments cannot be estimated with the same degree of confidence, and you should rely on empirical data far more.

However, perhaps the biggest assumption in this analysis was that your brand conversions weren’t caused by non-brand searches initially. People searching for you by name must have been led to your site by something, and whilst some may be via word-of-mouth or other advertising, much of it may be people that have visited your site previously via a non-brand search, and are returning to make a purchase. Clearly, these sales are at least partially (almost entirely) due to the non-brand clicks, and excluding their value from non-brand understates the true value of these clicks.

On this last point, the solution is multi-touchpoint conversion attribution. By assigning part of the value of the sale to the first visit, and part to each subsequent visit, you’ll get a much truer idea of how your Adwords campaigns are performing.

The two points to take away from this blog are these:

• Don’t rely on the overall performance of your account. Look at the performance of the individual Ad Groups and campaigns (particularly non-brand campaigns or Ad Groups). If you aren’t being given this information by your marketing team or agency, request it.

• Don’t be distracted by the total cost per sale – consider how much the incremental sales have been costing. Could you make more money with a higher or lower bid?

• Use conversion attribution to understand the true value of your PPC campaigns (and the rest of your marketing), and attribute your sales appropriately. This data is now available in Google Analytics, so your marketing team or agency should be able to analyse it for you easily.

Please share your thoughts below – @CustardMite

But for many retailers, it’s also a time to make lots of money – hopefully enough to get you through the rest of the year with a bit to spare.

For one of our clients in particular, Christmas is particularly important – they sell jewellery. It always interests me that when we manage accounts for jewellers, the peak days are much lower than for other clients. For this client, the peak period this year was the 11 to 15 December – is it possible that ladies jewellery is a favourite amongst ‘last minute’ Christmas shoppers?

There’s always a problem with buying Christmas presents online in December, though – will they arrive in time? The later you get in the month, the more of a worry it becomes, and as a result, about a week before Christmas conversion rates drop off a cliff.

However, our client used a delivery service that offered delivery in time for Christmas right up until 21 December – but in the past, they hadn’t recognised that this was a key selling point.

We took over the management of the account in late November, and recommended that they push this message far more aggressively on their website, and in the last few weeks before Christmas, we included it in the adverts as well.

The result? Their click through rate increased in the weeks before Christmas, and the conversion rate actually increased slightly in the two weeks before their last date for delivery. Immediately after the cut-off for delivery was reached, the conversion rate dropped by more than a half – had they not highlighted the final date, the conversion rate would have dropped to this level much earlier.

The upshot of all of this was that simply by recognising that fast delivery is a huge selling point in the run-up to Christmas, the uplift in their revenue was equivalent to their total revenue for a month any other time of year. So for retailers, perhaps the true Christmas Message should be ‘Buy before x and get it in time for Christmas’.

What are your thoughts? Please leave your comments below.

1)    Going into battle without the knowledge and tools of the trade at your disposal isn’t a good idea. You may be able to hack and slash your way through a few bandits, but you’re going to get your posterior handed to you a number of times as you learn what you’re doing. There are plenty of guides to help you to avoid the common mistakes in AdWords (and a book for Skyrim), and it’s a much less painful way to learn the ropes.

There are obvious similarities between a good shield or defensive spell, and understanding how to adjust your bids to reflect the value of your clicks. If things aren’t going well, cutting your bids intelligently can be as effective as a top quality shield to protect you from losing money if your conversion rate dips (or somebody tries to cut your head off).

2)    Dragon slaying isn’t just a child’s dream anymore – you really can do it (well – not really, but you know what I mean). On Skyrim, the dragons are standing between you and progress through the game. In AdWords, the dragons are more metaphorical – big advertisers with long teeth (financially speaking).

AdWords gives you the tools you need to bring them down – if your clicks generate more revenue than the dragons, you can outbid them, and if you can write more compelling adverts, you can level up your Quality Score.

3)    It has to be said, it’s very easy to lose sight of your goal in Skyrim. Whether you’re chasing butterflies, beating up bandits or simply accosting innocent travellers (come on – we’ve all done it), you can spend days heading off in completely the wrong direction. If you’re going to kill the big scary dragon that’s causing all the trouble, you’re going to have to keep your eyes on the prize.

AdWords is very similar (sort of). You are advertising in order to make money. There are lots of distractions, like Quality Score, maximising your click through rate, the various advertising options, search query reports, bidding options and others. But whilst you should be taking advantage of these things if they push you towards your objective, they should never be mistaken for the objective – to make money.

4)    I loved Oblivion (the previous Elder Scrolls game). I put many, many hours into it, and I wasn’t totally convinced that improved graphics on Skyrim would add much to the experience. It’s a bit like watching a football match in high definition – you love the picture quality for the first couple of minutes, but after that, you completely lose yourself in the game, and don’t notice it any more.

However, so far, I haven’t become used to the improved detail in Skyrim – it’s still impressing me, and I think it’ll continue to do so for a long time to come. And again, this is comparable to AdWords.

The level of detail that you can get on the performance of your AdWords account is improving all the time, and it’s really improving the user experience. It wasn’t that long ago that you couldn’t split out the performance of Mobile Phones, see how your adverts performed in the top positions compared to the side positions, or see an hourly breakdown of conversions. To be fair, on Skyrim the differences don’t actually make it any easier to achieve your objective as far as I can see, whereas on AdWords it can potentially help you to improve your performance.

5)    The scope of each successive Elder Scrolls game is a step beyond what I’ve seen previously. I had thought that the size and scale of Oblivion was staggering, but already it’s looking limited compared to what I’m seeing on Skyrim.

AdWords is a bit like that – over time, it’s becoming bigger and bigger. At first, you simply advertised on Google Search, but over time, they added the Display Network, Search Partners, Remarketing, YouTube, Interest Targeting, Topic Targeting and many others to expend the reach and targeting of your advertising.

I recently blogged (on another site) about how AdWords could threaten the big banner advertising networks, if they can persuade advertisers to view the Display Network the way that they do the rest of their advertising spend. Who else can put your adverts on Google Mail and You Tube?

6)    There’s nothing more enjoyable than a jaunt into a mine (on Skyrim, not AdWords). There are treasures to be uncovered, and who knows what manner of beasties and baddies you may come across.

Obviously, this is just like looking at a list of your search queries. Who knows what little gem of a profitable search query you may find, and add to your keyword list? And those nasty little goblin-esque irrelevant search queries that waste your money and attack you (metaphorically speaking).

Mining your searches is clearly similar to searching mines – whilst you can achieve your objective without doing it, it can be a profitable and enjoyable way to spend some time.

Clearly, there are many more ways in which optimising AdWords accounts is just like playing Skyrim – these are just a few examples. Hopefully these comparisons will be helpful to you whether you are playing Skyrim, optimising an AdWords account, or both.

What are your thoughts? Please leave them below.

Their announcement in full reads as follows:

When searching on Google, users appreciate results that are relevant and deliver a great experience after they click. In August, we announced trials in Brazil, Spanish-speaking Latin America, Spain, and Portugal¹ that increased the weight given to relevance and landing page quality in determining Quality Score and how ads are ranked on Google. The goal was to improve the user experience with search ads. Based on the results we’ve been seeing, we’re now rolling these changes out globally over the coming weeks.

As the changes roll out, some campaigns will see variation in keyword Quality Scores and typical ad position. Within a couple weeks, things should stabilize and we expect most campaigns will not see a significant change in overall performance.

Just like today, campaigns with better-performing ads for user queries will continue to see higher Quality Scores, lower average cost per click and higher position on results pages. The 1-10 Quality Score in the AdWords interface will continue to represent the general quality of your keyword when exactly matching a search on Google. Increases in bid and Quality Score will continue to increase Ad Rank, and the same basic approach to improving your results with AdWords applies.

This is interesting, as this is effectively the same as saying that they are reducing the role of the click through rate in the ranking algorithm. But the use of click through rate is critical to Google – this is what they use to maximise their revenue from Adwords.

By promoting adverts that are more likely to get clicks and demoting those less likely to get clicks, Google ensures that the adverts that appear in the top positions (with the highest cost per click) are the ones most likely to generate clicks, and that an advert with a lower click through rate has to effectively make up the loss of revenue for Google by bidding higher, in order to get the position.

This isn’t a criticism of Google at all – I believe that it’s generally in the user’s interest to see adverts that they are likely to click on – this is a strong indicator that the advert is relevant to the user. So by moving the focus of the algorithm from pre-click to post-click, Google are likely to cost themselves money. But does this really improve the user experience?

Certainly, if the keyword matches the search query, it’s an indication of a more diligently managed account – a campaign that utilises Exact Match extensively rather than relying on Broad Match requires more work. And more targeted landing pages can be achieved by effective keyword grouping – but does this mean that user gets a better experience?

To be honest, I’m not convinced. I doubt that it will make any discernible difference to users, but anything that encourages PPC account managers to manage their keyword lists effectively (or rewards those of us that do) can’t be a bad thing.

-CustardMite

Interflora is delighted by the judgment of the Court of Justice of the European Union which today ruled in their favour. This ruling will enable brand holders across Europe to deliver quality service and ensure that trade marks guarantee the origin of the goods bought by consumers online.

I’ve always been of the opinion that as long as you are making it clear in your advert who you are, and aren’t trying to pass yourself off as the brand being searched for, advertising on a competitor’s brand name could only improve the experience for the searcher.

There was a time when Interflora was the only company that most people could name if they wanted to send flowers through the post. As such, when they search for Interflora in order to send flowers, being presented with a number of other companies offering the same service only improves the choice of services available to them.

I was a little surprised that the court hadn’t seen it that way, so I wandered over to their website to find out what their arguments were. I was far more surprised with what I read. The final paragraph of the ruling is perhaps the most interesting:

Article 5(1)(a) … to approximate the laws of the Member States relating to trade marks and … must be interpreted as meaning that the proprietor of a trade mark is entitled to prevent a competitor from advertising – on the basis of a keyword which is identical with the trade mark and which has been selected in an internet referencing service by the competitor without the proprietor’s consent – goods or services identical with those for which that mark is registered, where that use is liable to have an adverse effect on one of the functions of the trade mark. Such use:

–      adversely affects the trade mark’s function of indicating origin where the advertising displayed on the basis of that keyword does not enable reasonably well-informed and reasonably observant internet users, or enables them only with difficulty, to ascertain whether the goods or services concerned by the advertisement originate from the proprietor of the trade mark or an undertaking economically linked to that proprietor or, on the contrary, originate from a third party;

–      does not adversely affect, in the context of an internet referencing service having the characteristics of the service at issue in the main proceedings, the trade mark’s advertising function; and

–      adversely affects the trade mark’s investment function if it substantially interferes with the proprietor’s use of its trade mark to acquire or preserve a reputation capable of attracting consumers and retaining their loyalty.

There’s clearly a bit of wriggle room here. What they appear to be saying is that you can’t bid on somebody else’s brand name, if doing so makes it unclear which is their website, or reduces the value of their trademark.

The second part reads:

Article 5(2)… must be interpreted as meaning that the proprietor of a trade mark with a reputation is entitled to prevent a competitor from advertising on the basis of a keyword corresponding to that trade mark, which the competitor has, without the proprietor’s consent, selected in an internet referencing service, where the competitor thereby takes unfair advantage of the distinctive character or repute of the trade mark (free-riding) or where the advertising is detrimental to that distinctive character (dilution) or to that repute (tarnishment).

Advertising on the basis of such a keyword is detrimental to the distinctive character of a trade mark with a reputation (dilution) if, for example, it contributes to turning that trade mark into a generic term.

This is reasonable. You can’t bid on their name in order to pass yourself off as them or do anything to tarnish their name. The third and final paragraph is the really interesting one…

By contrast, the proprietor of a trade mark with a reputation is not entitled to prevent, inter alia, advertisements displayed by competitors on the basis of keywords corresponding to that trade mark, which put forward – without offering a mere imitation of the goods or services of the proprietor of that trade mark, without causing dilution or tarnishment and without, moreover, adversely affecting the functions of the trade mark with a reputation – an alternative to the goods or services of the proprietor of that mark.

If I understand this correctly, if you are not imitating them, diluting or tarnishing their brand name, or affect the trademark’s effectiveness, then you can bid on their brand name.

Clearly, you could argue that as long as you clearly present yourself as an alternative purveyor of that product or service, you can still bid on their trademark. From reading this, I would suggest that Marks and Spencer are in the clear – they never claimed to be Interflora, and they didn’t insult them in any way – they simply put forward “an alternative to the goods or services of the proprietor of that mark”.

Ironically, I would argue that Interflora are currently the ones more likely to fall foul of this ruling. Here’s what their advert says when you search for “Marks and Spencer Flowers”:

Unlike John Lewis, Asda and Teleflorist, they have decided not to include their brand name in the title, instead advertising that they sell flowers in Spencer. If you search for “John Lewis Flowers”, you also discover that they sell flowers in Lewis:

Maybe they are just bidding on every town and city in the country, but I know that when I search for “Castleford Flowers”, they are nowhere in sight. I leave it up to you to decide whether the pot’s calling the kettle black, or it’s just a big coincidence.

But if I were Interflora, I wouldn’t be shouting my victory from the rooftops, I’d be keeping very, very quiet…

What are your thoughts? - @CustardMite

6. Be Patient

Everybody is guilty of impatience with their accounts sometimes. If you had a bad day yesterday, the immediate impulse is to hit the big red panic button, and start making changes. Obviously, in the cold light of day, this is not wise. In fact, unless the performance is truly dreadful (or you’re really bored), spending a couple of hours trying to identify what went wrong is a waste of time. You’ll probably find nothing, and the performance will usually have recovered by tomorrow.

It’s the same with bid adjustments. If you changed the bids on Monday, then you can only look at the performance since then in order to assess what the impact of those bid adjustments was. If your overall performance has been poor for a few days, it doesn’t mean that you should change the bids back – assuming that you changed them based on sufficient data in the first place, they are probably not the cause of the poor performance.

The only exception to the ‘be patient’ rule is with advert testing. If you’ve been waiting a while, and neither advert is far ahead, then end the test, keep the original advert (probably) and try something else. The longer the test takes to get a result, the less impact choosing the better advert will have. Run another advert, and try to improve things a lot – big improvements are better than small ones, and they take less time to spot.

7. Get Their Attention

If nobody sees your advert, nobody is going to click on it. What looks like a good advert on an Excel spreadsheet, or in the Adwords interface, may disappear completely when you put it in the middle of ten others…

A search for Bosch Drills shows a lot of very similar adverts. How many ways are there to say that you have good prices, a good range and quick delivery? If you’re at the top, this isn’t a problem, but what about the lower adverts? A few advertisers have used product feeds to make their advert more visible, which is definitely a good move. Screwfix have a rating, which also makes the advert stand out more. A few advertisers mention prices, which can help – particularly if you’re cheap. Eyes are drawn to numbers more than they are to text; just don’t do it if your prices are high!

The point is this – if nobody even sees your advert, they won’t click on it. There are a number of ways to make your advert stand out – parroting the same ‘unique’ selling points as everyone else isn’t one of them.

8. Split Your Performance Data

So your account’s performing well, and you’re content. That’s nice, but unless you’ve sliced and diced your performance every way that it can be sliced and diced, you may be missing a trick. Google allows you to see your performance in Adwords split in many different ways. Your overall performance may be masking some big opportunities.

For example, what happens when you split your campaign performance into Google Search vs Search Partners? The click through rates are likely to be very different, but that’s not really the point. We’ve found that for some campaigns, the performance of Search Partners is simply too poor to justify it, and that by turning it off and pushing the spend into Google Search, we can improve performance significantly.

Unfortunately, whilst you can bid on Google without Search Partners, you can’t bid on Search Partners without Google. As a result, if both networks work to a greater or lesser extent, you can’t split your campaign into two, and bid on them separately…

Another split worth investigating is Computers vs Mobile Devices. Frequently, the conversion rates can be very different, but this is only part of the story. As there are less paid adverts on Google on mobiles, click through rates can vary, and different adverts may be effective.

It’s probably not worth splitting Mobiles out unless you’re getting enough clicks to make it worthwhile, but if you are spending enough on Mobiles, and the performance is significantly different, then you may be able to improve the overall performance just by splitting them out.

By far the most important split is Brand vs Non-Brand. If you can afford to pay £20 per conversion, and you are paying £20 per conversion, you may be happy. But if Brand conversions are costing £1, and others are costing £50, things may not be so rosy.

Since your brand performance is almost always better than the rest of the account, you should always split this out before assessing how you are doing. Otherwise, you could be bidding too much on everything else in your account, wasting big piles of money.

You may find that the performance of your account varies by region (particularly if you are more well-known in some regions, or you have stores in some cities) – again, this may be worth checking.

Finally, check whether your performance varies on certain days of the week, or times of day. This is a little bit iffy, as people may research in work hours, then convert in the evening (particularly for expensive commitments, where spousal sign-off is required!). But it’s quite possible that your bids should be different on different days, or at different times – it’s certainly worth investigating.

9. Use Advert Extensions

If your advert appears in the top few positions, and you don’t have advert extensions of some kind, you’re missing a big opportunity.

Searches with high local intent will throw up a map – and you can put your marker on it at no additional cost:

Alternatively, you can put a nice map to your location:

Or add site-links:

Or even phone numbers that you can click-to-call on mobile phones.

All of these things have one thing in common – they make your advert, bigger, brighter and more noticeable.

10. Be Brave

If you try something new, and it doesn’t work, it’ll cost you a few pounds (or dollars, or yen, or zloty or whatever your local currency is). If you try something new, and it does work, it’ll make you money for years to come.

Google are constantly rolling out new toys to play with – whether it’s Remarketing, Interest Targeting, Topic Targeting, In-Video YouTube adverts, or something completely different, it doesn’t hurt to try it out.

Even if they don’t work, you can tell your boss (or client) what you’ve been trying, and they’ll be impressed with how proactive you are. And when they do work, you’ll get a warm and fuzzy feeling inside.

The only exception to this is when Google offers you something to ‘reduce your workload.’ Don’t try automated bidding (unless your account is absolutely massive, and you really haven’t got time to adjust the bids yourself), ‘optimising’ your advert delivery, or anything else that means that Google will do things that you should be doing yourself.

You decide when the bids need to be adjusted, and you decide when one advert is better than the other. That way, you know that the account is being managed to optimise performance, not to maximise spend.

Last Bit…

There are many other pieces of advice that I could give. I could go into detail, talking about writing effective adverts, the relationship between advert position, advert text and click through rate, or the importance (or unimportance) of the Quality Score. But this is only a blog, not a book, and I didn’t really mean to go as long as I have. That said, it’s possible that I’ve missed a critical tip – feel free to point any out that I’ve overlooked…

-@CustardMite

Over the years, I’ve seen a lot of PPC accounts, and what never ceases to amaze me is how the same mistakes are being made time and time again, even by some of the biggest agencies in the country.

I’ve always maintained that PPC is easy, and that if you do it correctly, it doesn’t really take all that much time to get great results.

So what I’ve decided to do is to put together ten pieces of advice that I would offer to anybody managing an Adwords account. This isn’t a ‘how to do PPC’ manual, rather it’s a set of guidelines for somebody that’s managing an account, and isn’t sure how to get its performance moving in the right direction. This blog post will feature the first five pieces of advice, so watch out for my next blog post with the final five recommendations!

1. Start Simple

If you are building an account from scratch, there’s a real temptation to produce a keyword list with thousands of keywords split across hundreds of Ad Groups, and dozens of campaigns. You may want to try out the Display Network, some Interest Targeting, advertising on You Tube, possibly a bit of Remarketing, and so on…

Slow down there! The more complex an account is, the more time and effort it will take to optimise its performance. Build a relatively simply account structure with an easily manageable number of keywords and Ad Groups, and get that working properly first. Once your main campaign is performing well, you can introduce additional levels of complexity one at a time.

Just bid on the keywords that you are confident will deliver good results, don’t worry too much about the long tail, and use Phrase Matching or Modified Broad Matching to pick up other relevant searches.

Run search query reports to pick up new keywords with enough traffic to justify their inclusion (if it’s unlikely to get another click this month, it’s not worth worrying about), or exclusion (again, only worthwhile if people are likely to search for things containing that word regularly).

Regarding negative keywords, don’t include search queries as negatives if you can avoid it – use single words where possible. If you don’t sell red widgets, don’t include [red widgets] as a negative – you’ll still appear for cheap red widgets, red widget and red widgets uk. Instead, include ‘red’ as a negative keyword.

Also, decide whether negative keywords are appropriate just for the Ad Group, or for the whole campaign. If you only sell widgets, ‘red’ should be added at campaign level (that way, you don’t need to add it to your ‘cheap widgets’ Ad Group), but if you also sell thingies, including red ones. It needs to be applied at keyword level. Believe me, it’s a lot easier to manage your keyword list if you split them like this…

2. Be Organised

One of the biggest frustrations that I have when I review a potential client’s account is when I’m confronted with chaos. Whilst I have nothing against chaos per se, it has no place in an Adwords account. An organised structure isn’t just important for optimising your account, it’s absolutely essential.

What additional value do you get from having 100 campaigns in your account? It’s not essential to do this in order to report on performance – if you consider your Ad Group names carefully, you can still reflect multiple levels of structure easily enough (with a bit of Excel jiggery-pokery).

For example, if you want to report on different types of clothing, but also on brands, you may have a campaign for each brand, with an Ad Group for each type of clothing, or vice-versa. However, simply labelling your Ad Groups like this:

Nike – Shoes

Adidas – Shoes

Nike – Tops

Reebok – Tops

Etc.

…will allow you to report on brands or clothing types easily.

There’s no disadvantage to doing this, and there are a number of advantages. Firstly, you have fewer campaigns, so you don’t have to split your budget into as many pieces. I reviewed a campaign recently that had a daily budget of more than £10,000, but there were so many campaigns that some had a daily budget of just £2. Clearly, the risk of running out of budget in this case was high, despite the overall budget being more than adequate. The more campaigns you have, the more likely you are to run out of money on one campaign, whilst others carry a surplus.

Keywords are also easier to manage. If you find that people are searching for ‘nike training shoes,’ you can easily add ‘reebok training shoes,’ ‘adidas training shoes’ etc. to the appropriate Ad Groups as well in seconds. Similarly, adding negative keywords is simple.

More than anything else, it’s important that your Ad Groups and Campaigns don’t overlap. The same search query shouldn’t be able to trigger adverts from two different Ad Groups or campaigns. If it can, then it becomes difficult to tailor your advert message to the search, and it’s impossible to adjust bids effectively – all that’s liable to happen if you reduce a bid is that you’ll push clicks into a different Ad Group.

This is a common problem, particularly where people use generic Ad Groups. There’s nothing wrong in the above structure in having Ad Groups for generic (non-branded) clothing and footwear searches – but assuming that you are using Phrase or Broad Match, you need to include the various brands (Nike, Adidas etc) as negative keywords. Similarly, if you have a generic Nike Ad Group, you will need to include the product types as negative keywords. Otherwise, if somebody searches for Nike Trainers, they are in danger of seeing a generic advert for Nike, or one for trainers, rather than a specific Nike Trainers Ad Group.

3. Be Relevant

In the last point, I highlighted that if somebody searches for Nike Trainers, you want to ensure that you show an advert for Nike Trainers, not a generic Nike or Trainers advert. There are a number of reasons for this.

Firstly, obviously people are more likely to believe that you sell what they are looking for if your advert says that you do. Secondly, you can choose a more appropriate landing page. As a result, you can potentially increase click through rate and conversion rate.

Consider the simple account structure that you started out with from my first point. By gradually expanding out your keyword list over time, and segmenting Ad Groups when you find groups of keywords that indicate a different type of searcher (maybe from adjectives that they use, like cheap or luxury, or from conversion rate), perhaps some keywords clearly perform better than others. Even from click through rate, maybe your advert is speaking more clearly to some people than others; you can then improve the relevance of your adverts, as well as tailoring your bids.

4. Bid To Your Objective

I really hope that you understand exactly what you are trying to achieve with your account. It needs to be very clearly defined and measurable not just overall, but at the level at which you manage your bids (Ad Group level, in the main part, if you’ve got any sense!).

Why do you need to be able to see the performance at Ad Group level if you are adjusting your bids at Ad Group level? Well,  how are you going to know which bids to increase or decrease, if you have no idea which Ad Groups are delivering sales or leads or whatever you want?

Ideally, your objective should be as close to the business’s objective as possible. If they are trying to maximise profit, and you know revenue and their margin on that revenue, at Ad Group level, then you can optimise on contribution to profit. You can work out how much they make from a click via an Ad Group, and adjust the bids based on this to maximise their profitability. If you don’t have their margin data, then you should be looking to maximise revenue. If you don’t know the order value, then maximise sales volumes. If you can’t track actual sales (possible if payments are processed on a third-party website), then maximise shopping baskets created.

Whatever your objective is, you should set your bids solely with this in mind. It doesn’t matter if that means that some searches show your adverts, whilst others show you in tenth. You should try to get the same return on investment on every Ad Group (or keyword, if you insist on trying to optimise at that level).

5. Don’t Run Out Of Money

Often the biggest mistake that advertisers make is to run out of budget before the end of the day. If you run out of budget by lunchtime, you’re sabotaging your account. Think about it. What happens if you halve your bids? Your adverts appear lower, you get a much lower click through rate, and your budget lasts until the end of the day. But if you’re still spending the same amount, and you’ve halved your cost per click, you’ve doubled your clicks, and conversions, and revenue. Your profit goes through the ceiling, and you’ve not had to do any work to achieve it.

In order to maximise the number of conversions that you get for your budget, it should just run out as the day ends. In practice, this is impossible to achieve due to random variation, but since Google allows you to overspend some days to compensate for days when you underspend, you should be able to get close…

Check your lost impression share (budget) at campaign level. If any campaign is running out of budget every day, either move some from a campaign with a surplus, or reduce the bids.

I hope you’ve learned something and stay tuned for recommendations six to ten in my next blog post! In the meantime, please feel free to leave comments or questions below or on Twitter - @CustardMite.

Looking at it, the distinction between the paid search results and the organic results is clear. The paid search results are in a coloured box saying ‘sponsored links,’ and they have a totally different format to the organic search results – short titles, and only two lines with the URL at the start of the second line.

I ran the same search last week, and the difference was striking:

The shaded box is much lighter, ‘Sponsored Links’ has been replaced with ‘Ads’ and the PPC adverts now have longer titles and a distinct line for the URL. Basically, it’s very easy looking at the screenshot above to be confused as to which results were organic and which were paid.

These are just a few ways in which the PPC search results are being made to look like organic search results – you can now have your location on a map – and it’s interesting that the PPC locations (in blue) appear on top:

In these examples, the address and phone number are also been displayed on additional lines of the advert, and Google even gives you a link to directions.

Site Links are also looking more and more similar. Sometimes, the organic site links appear on a single line, and sometimes the paid search site links appear in columns:

Paid:

Organic:

Paid:

Organic:

Google may argue that they are working to improve both sets of search results to make them as useful as possible, and to an extent, that does make sense. People searching for a cosmetic surgeon may want to know where the different advertisers are actually located, and the site links are definitely useful sometimes.

It could actually be argued that a pay-per-click model should deliver more relevant search results than an organic model under certain circumstances. After all, the amount that companies can afford to pay per click should be related to the amount that they receive from a click – which is clearly related to the conversion rate, which must clearly be a strong indication of relevance. Of course, for informational searches, the paid search results are likely to be weak – there’s not much money to be made from providing people with information, in general.

But the fact remains that a substantial number of users simply don’t trust paid search results, and always click on the organic results. Whether or not this is rational, it is their right, and Google have always stated that they would keep the two sets of results separate.

So are Google doing anything wrong here? Personally, I believe that they are improving the user experience by delivering clearer, more easy-to-read adverts on both paid and organic search, but there’s no doubt in my mind that they are deliberately obfuscating the division between the paid and organic adverts together – which is clearly a change of position for them.

Right now, the only real difference that remains between the two sets of adverts is the limitation of 35 characters on the third line of the advert text. Whilst this makes perfect sense for adverts on the right hand side, for paid adverts above the organic search results, it really doesn’t, and the advert looks a bit odd as a result. It wouldn’t surprise me at all if Google released a new trial soon, allowing you to write alternative advert text dependent on whether your adverts appear at the top or on the right. If this happens, then the only demarcation left is the (increasingly faint) shaded box. I doubt Google will get rid of that any time soon, as it would effectively be an admission that they want to combine the search results, but I suspect it’s just a matter of time…

I won’t go into the details of everything that he said that was patently ridiculous, though if you’re interested, Melissa Mackey has covered most of the salient points on Search Engine Watch here: http://bit.ly/mCBAvw

Instead, I’ll offer a counter-argument as to why PPC is a really great way to advertise for many (most?) businesses.

Firstly, I should make it clear that the title is intentionally ridiculous – there is no ‘best’ or ‘worst’ form of advertising for everyone (whatever Mr Shatkin-Margolis believes), as different businesses may find a variety of advertising media to be the most effective.  However, there are many advantages to Paid Search, that may explain why it’s become a multi-billion dollar industry in just a handful of years.

1)     Scalability

Whether you are a plumber in Milton Keynes, or a multi-national retailer with a million dollar per month budget, advertising via Paid Search is a viable way to contact potential customers.

2)     It’s A Level Playing Field

It really is! If I start up a little online computer shop, then if I generate more money from a visit than PC World, I can afford to outbid them on my target keywords. Further, if I can convince people to click on my advert instead of theirs (OK – that may be a bit tricky), then my advert should appear above them, at least until my budget runs out.

3)     Transparency

Imagine that instead of running a PPC campaign, I ran a televsion advert and some adverts in The Times. Leaving aside all of the practical issues, how would I know whether my sales were caused by one or the other (or just word-of-mouth). But Lord Leverhulme’s complaint (not knowing which half of his marketing budget he’s wasting) doesn’t really apply to PPC – you know exactly how many clicks and sales PPC has generated. OK, it gets a little complicated sometimes, but if you’ve got no organic rankings, and somebody searches for ‘computers,’ how are they going to buy from you if you aren’t doing PPC?

4)     Targeting

With PPC, you can currently target individual cities, regions or even an area that you’ve doodled on a map. It goes without saying that you can also target your keywords so that the only people that see your advert are the ones that are interested in your product or service. OK – maybe it didn’t go without saying…

5)     Speed

PPC is quick – really quick. Think about it – how long does it take to get rankings on your new product range via SEO? How long does it take to make a television advert and broadcast it? Even newspaper adverts can take days or even weeks. With PPC, you can be up and running in a few minutes. Not happy with your message? It only takes a couple of minutes to fix it.

6)     Affordability

OK – so the clicks may not be particularly cheap – the prices are largely determined by market forces. But our plumber friend from earlier can set up his own PPC campaign for just a few pounds. No need for a production company to create his television adverts, and he can do it himself – he doesn’t even need an agency. Don’t get me wrong – if he’s spending a lot of money, he should probably hire us to make the campaign as profitable as possible, but for small advertisers, there really are no overheads.

7)     Simplicity

There are plenty of ways for you to make a mess of your PPC campaign, but as long as you keep it simple, it’s really not that hard to do a half-decent job of running a campaign. Just don’t start believing that you’re an expert, and try all the ‘bells and whistles’ that Google give you to play with.

8)     Flexibility

PPC can be used to support your other marketing activity, whatever it is. Whether you simply bid on your brand name to ensure that people find you, you bid on keywords related to your advert (Day V. Lately anyone?) or you run a major PPC campaign to raise your business’s profile, PPC can be a useful part of the marketing mix.

9)     Profitable

I’ve said it before (repeatedly), and I’ll say it again. It should be impossible to lose money on Adwords in the long run. As long as you know how much a click is worth to you (multiply your conversion rate by the value of a conversion to you), and bid less than that, you’ll make money. Maybe not much (if your conversion rate is rubbish), but how many other media can make the same claim?

10)  Erm… OK, so I only came up with 9. Any suggestions for number ten?

Now I’m not a doctor, but Scientific American declared last year that Google Flu Trends is “nearly on par with CDC surveillance data.” And if it’s good enough for them, clearly there must be something of value there.

Of course, now that they’ve done the analysis for the flu, everyone has been beating a path to Google’s door trying to find a way to predict their data. So Google decided to release a tool so that anyone that wants to can try to find searches that correlate to their data. Cunningly, they’ve called it Google Correlate.

The idea is that you can feed in the trend data for whatever it is you’re looking for, whether it’s car sales or swine flu, and see which keywords follow similar trends.

To show you what it does, I’ve drawn an imaginary trend on their system:

Hitting Correlate! tells Google to go and find the keyword whose search volumes most closely match this shape. In this case, the term that comes up is ‘6710b,’ a laptop made by Compaq, released in August 2007:

This is a fairly straightforward example, and given the line that I drew, it’s unremarkable – I was matching to something that spiked sharply in August 2007, and has steadily declined since – a new piece of technology was always likely to fit the profile. Had I drawn something seasonal, I’d get a totally different result:

It appears that people search for wedding toasts in the summer, but that fewer and fewer people search each year – are less people getting married, or are best men just writing their own toasts now?

Levity aside, this tool could be very useful to build a model to predict the near future, particularly in areas where collecting accurate data is either difficult, or the data simply takes a long time to collect.

On top of this, Google has added another interesting feature, which allows you to compare keywords to each other. Is it possible that this could be used to generate new keywords that you may have missed?

Apparently not.

I should probably finish by adding the caveat that Google give – that correlation and causality are not the same thing. A strong correlation may simply be the result of a third factor that influences both – such as time.

For example, according to the Pastafarian religious beliefs, pirates are good for the environment – as the number of pirates has fallen, the average global temperature has gradually increased – therefore pirates combat global warming. I wouldn’t rely too much on this – they also believe that a flying spaghetti monster is the supreme being.

The religion and this article of their faith are obviously spoofs, but it highlights the fact that assuming a direct relationship between two variables that show a similar trend is unwise. Should I interpret from my data that when people buy levis jeans, they all stock up on black powder firearms as well? Or that pregnant women in India look to enhance their natural figure?

From a PPC perspective, I suspect this tool is going to be of no use whatsoever, but then, I suspect that Google didn’t build it for me, but to stop the CDC from phoning them every week asking about the latest disease…

So what is the best position to put your advert in, and how do you calculate it (this is effectively what you are choosing with your bid)?

Ask a lot of PPC analysts about this in the past, and they’d have reeled off an answer like “fourth” or “second,” but the industry is slowly catching on to the fact that the question is much more complex than that. After all, fourth place is going to be quite expensive on competitive terms, and if your site has a low conversion rate, you could easily lose money. On the other hand, if the term has few people bidding on it, then a higher position will only cost a few pence more and generate loads more conversions.

Clearly, a simple one word answer isn’t quite going to cut it…

Before answering the question, there are a number of general rules of thumb that can be applied to PPC adverts:

1. The higher your advert appears, the more clicks you will get.
2. The higher your advert appears, the more you pay per click.
3. Moving your advert to a higher position won’t improve the conversion rate.

So what do you need? Well, we’ve said that the conversion rate increases as you move down the results, and that the click through rate and cost per click fall. You need to put numbers to this. Something along the lines of:

Don’t use the numbers above, I’ve just made them up, and they will be different for each campaign, and probably for each Ad Group or even each keyword. The only way to get these is to move your advert around for a little while, and get them. Once you’ve done this a few times, you can get some feel for them and make a few guesstimates. This is really the only time that the Quality Score influences your bidding – in the amount that you have to pay to appear in each position.

Once you’ve got this table, calculate the cost per conversion (by dividing the cost per click by the conversion rate). Subtract this from your approximate income per conversion (take into account the cost of the goods/services!), and you get the ‘profit’ per conversion. So for the above figures, with product that retails for £150, with a cost to you of £75, the results would be:

As mentioned earlier, this will always increase, as you move further down the rankings, especially if the conversion rate is higher further down the results. This makes sense, since the cost per conversion will be lower in lower positions. It should also level out quite quickly – the cost per click for 11^th and 20^th are usually reasonably close together …

Next, multiply this by the conversion rate to get the profit per click. This is, in effect, the maximum that you can bid – bid less and you’ll make money, bid more and you’ll lose money. Multiply this by the click through rate, and you’ll have the profit per impression. Multiply this in turn by the impressions per day (or month or year) to get the profit over that period. Here, we’ll assume 1000 impressions per day, but remember that the lower your adverts are, the fewer impressions you’ll get.

So, in the above example, the sweet spot is fourth, though any position below fourth will also show a small profit (as the income for each unit sold will always be more than the cost below this point).

Is all this work worth it? Perhaps not, if you are in fourth position, and haven’t got the budget to move into first or second – but the general rule is very useful. If you lower your bid, you’ll get fewer conversions, but each one will cost less, and if you raise it, you’ll get more conversions, but each one will cost a bit more.

What you will find is that there are few situations where the top position is the most profitable. Even if your site converts better than your competitors, your advert has a better click through rate, and your margin on sales and average order value are better, you are still relying on them to do the same calculations that you have.

Otherwise they could be bidding over the odds, and competing with them just means that you lose money, they lose money and Google gets richer!

Note that the cost per click here is the actual cost per click, not your bid (refer to the last section on Quality Score if you don’t understand why your cost per click is lower than your bid).

At the start of the last blog, I identified the three things that you need to do in order to optimise your account (assuming that it’s correctly set up) – Maximise the Conversion Rate, Maximise the Quality Score, and optimise your bids. I’ve looked at the conversion rate, and now it’s time to look at the Quality Score.

How Adwords Uses The Quality Score To Decide Your Advert’s Position

Most people understand that Google Adwords assigns a Quality Score to your keyword and advert, and that this is combined with the cost per click to somehow decide how much you have to pay per click, and what position your advert appears in.

But few actually understand exactly how it happens, and just how important your Quality Score really is, or how it works.

The first thing that Google decides is which position your advert appears in. That’s fairly straightforward – the cost per click is multiplied by the Quality Score, and the adverts are ranked according to this. So, if you’ve got the following seven adverts, with the cost per clicks and Quality Scores given, the rankings would be as follows:

Now comes the hard part – how much do you actually bid to get that position?

Google starts with the bottom position – in this case G which is seventh – and charges a minimum cost per click. For the sake of argument, we’ll call this £0.10

To get sixth, advert E needs an adjusted bid of £0.69 (to beat the £0.68 for G). Given that it has a Quality Score of 1.2, the actual cost per click would need to be £0.57 (as £0.57 * 1.2 = £0.69).

To get fifth, advert F needs to beat the adjusted bid of £0.72 from G, which means that the actual cost per click for F would be (0.72/1.5) = £0.48

And so on…

Filling in the remaining numbers, you get the following:

It is useful to know, but how is this important?

Consider advert A, which is pretty hopeless with a Quality Score of 1.0. If A can get onto the same level as G, and improve the Quality Score to 1.7, it jumps to the top of the search ranking, and even ends up paying less! Instead of paying £0.88 to appear third, they would pay £0.85 to be top! Cheaper clicks, and twice as many of them! That’s a big return for just changing the advert text a bit, and making sure that the advert is more closely matched to the search term (you can do this by trying to turn one phrase match term into many using the search query report).

But what if you’ve already got a good Quality Score? Is it worth trying to squeeze a little bit more out of your advert, or is your time better spent doing something else?

Consider advert F. What happens in the above example, if they increase their Quality Score from 1.5 up to 1.55? They don’t move up the rankings, so it’s probably not worthwhile? Wrong. Two things happen here. First, their advert costs £0.46 instead of £0.48, so they get about 5% more clicks for their money (if they are limited by their budget), or pay 5% less for their traffic (if they aren’t). That’s like getting 2.5 weeks of traffic free every year.

But as always, Google doesn’t lose out (they’re very clever here!) as advert C pays £0.70 instead of £0.68 for their advert. So your competitors are effectively giving you their money! Depending on how you feel about your main rivals, this may give you a very warm and fuzzy feeling inside…

One point worth remembering here is that the Quality Score isn’t static – one of the attributes that impacts it is the click through rate, which is constantly changing.

So, whilst B may be paying £0.40 today, it may be £0.38 or £0.42 tomorrow, due to variations in both B’s and A’s click through rates.

The Quality Score uses the CTR, Keyword Relevance, Advert Text and Landing Page. Let’s look at these factors in turn.

Click Through Rate (CTR)

Not much to say, really. The higher your click through rate, the more relevant your advert must be (since so many people are clicking on it).

This highlights the problem that if your product is of interest to only some of the people searching for a term, then Google will penalise you for only targeting them with your advert text. This is another reason why bidding on generic terms is dangerous if you don’t have the whole area covered by your product range…

You can improve your click through rate by making your advert text more eye-catching and appealing, but just remember that you don’t want to sacrifice your conversion rate to do so. Be aware that Google does take your advert position into account when assessing your click through rate, so increasing your bids won’t automatically result in a higher Quality Score.

Keyword Relevance

There are a few parts here – Google compares your keyword, the search query and the advert, and looks at how similar they are.

If somebody is searching for “blue widgets,” then your keyword matches the search more closely if you are bidding on “blue widgets” than if you were just bidding on “widgets,” though your advert would appear in either case. And if your advert says “blue widgets,” then your keyword and the search query will both match the advert closely as well. This is one reason that an extensive keyword list, appropriately grouped, is important (though there are also benefits in terms of bid adjustments and click through rates from more targeted adverts).

This approach by Google is reasonable – if somebody searches for “Leather Reclining Chairs,” then you’re more likely to have what they want if you tell Google that you have “leather reclining chairs” than if you tell it you have “reclining chairs,” and you’re much less likely if you tell Google that you just have leather chairs (broad match).

Advert Text

Advert text seems to come up in just about every section, doesn’t it!

In this instance, I’m looking at advert text from the perspective of maximising the Quality Scores.

It’s actually quite simple, in principle. Google wants to see the search term, or at the very least, part of the search term, in the advert - ideally, in the title.

The catch?

The adverts are written at Adgroup level, remember? Therefore, you have to be very careful how you group your keywords - it’s not enough just to have keywords that mean the same grouped together, they need common words in them, if possible.

At one extreme, you could put every single keyword in its own Adgroup. This would certainly give you the opportunity to target your adverts very well from a Quality Score perspective. However, when the time comes to analyse how your campaign is performing, it makes things quite difficult. And it also makes optimising the long-tail more difficult, since the keywords haven’t been grouped together, and as a result can’t be analysed together, making bid adjustments difficult.

So your keywords should be grouped not just on the sort of keyword they are, but also the actual words within them.

Now it’s not always possible to do this – if your keyword is more than 25 characters long, then you can’t put it in the title, and if it’s more than 35 characters long, it won’t fit anywhere. Just do the best you can.

It’s worth noting here that you may run into another Catch-22 situation here. You want to put your keywords in the title of your advert, but you also want your advert to stand out. What if everyone else already has that term in their title?

If your competitors are rank amateurs, who don’t really know how Adwords works, it’s not likely to be a major problem, but try running a Google search for “PPC Management!” You can bet that most of these guys know the score! At the moment, I’m seeing the following titles:

Which one stands out? The one with the lowest Quality Score! But they might well be making a lot of ground up with a good click through rate, and they have PPC in the second line, so perhaps they are ‘winning’ at the moment…

In this instance, it may not matter too much, as most people looking for an advertising agency would click on every link on the page (I think. I would, anyway).

There’s no ‘right’ answer to this one – but personally, I’d recommend improving the clickthrough rate. Test it yourself, and see what happens…

Landing Page

I’ve already covered this in a fair amount of detail in previous posts – there’s not really much more to add.

Basically, Google go and look at your website, and decide whether it’s the sort of place that they think people want to shop or visit. If it is, they reward you, if not, they slap you (not literally, obviously).

And that’s about it – understand these principles, and you should always be able to get a good Quality Score.

Comments are welcome and if you missed the last five parts, you can find them here: Part One, Part Two, Part Three, Part Four, and Part Five.

So, you’ve set up a campaign, and now you want to start fine-tuning it? Understandable, though depending on your business, optimisation may mean slightly different things…

For the majority of businesses, their ultimate objective is profit - making as much money as possible. And for most of these, the purpose of their PPC campaign is the same.

There are a few cases where this isn’t true. If you have a bricks-and-mortar store somewhere, you may run a PPC campaign to get people to visit you. This would make a lot of sense if you’re selling something that people aren’t likely to buy online, such as a second-hand car.

Alternatively, if your website is designed to give information, such as how to stop smoking, why you should convert to a new religion or who you should vote for in the next election, then you may have no objective, other than to be heard.

In all of these cases, you can use Adwords to track some form of conversion – placing the tracking code on the appropriate page will give you an indication of which keywords are the most effective. Online Stores should (if possible) place the code on their ‘Thank You For Ordering’ page, tradesmen may want to tag the ‘Thank You For Contacting Us’ page or the ‘Contact Details’ page, bricks-and-mortar stores may want to tag the ‘Find Your Nearest Store’ page. Even informational websites could tag pages containing information, in order to identify users that interact with the site.

Aside from identifying the keywords that you should be bidding on (which has already been covered), there are three aspects of optimising - one is true for all of these, and two are only true when you are trying to make as much money as possible from the site.

• Maximise your conversion rate
• Improve Google’s opinion of your campaign (in terms of its relevancy, known as Quality Score) as far as you can.
• Adjust your bids until you find the position in the search results that generates the most profit (higher = more sales but a higher bid, lower = less sales, but each one costs less).

This blog covers the first of these – maximising the conversion rate.

Optimising Your Conversion Rate

The conversion rate is probably the most critical number there is when trying to make money online, and with PPC, it’s even more critical.

Think about it. If your campaign costs £0.50 per click, and you make £30 from each conversion, with a 2% conversion rate, how much money do you make from 100 clicks on your site?

What happens if you double the conversion rate? It doesn’t double your profits!

Your profits just went up by 600%. That’s a lot!

In this example, reducing your cost per click couldn’t generate this level of profit, even if the clicks were free!

Even a small increase can have a disproportionate impact on your profit - in the above example, even an increase in conversion rate from 2% to 2.2% would increase profit from £10 to £16, a 60% increase.

So, how do you improve the conversion rate of your campaign?

Website Design

The first place to look has to be the website itself. This isn’t really related to your campaign, it’s a more general thing. It’s just more important with your PPC campaign, because you are paying for each and every visitor to your website – if your cost per visitor was £0, then a 10% increase in your conversion rate would yield a 10% increase in your profits.

There are plenty of books and websites about how to maximise the conversion rate of your website, so I’ll make just a few points on this.

• If people are giving you money, they need to trust you. If your website looks like it was put together by a spotty teenager in his bedroom, they may be reluctant to give you their credit card details.
• If people can’t find what they are looking for, they aren’t likely to buy it. Don’t tell people how to buy things on your website, let them choose their own way. Include a simple, straightforward navigation on the left-hand side of the page, and a clear, easily visible search box at the top (that actually returns what they search for!).
• Don’t distract them. You want a big, clear button telling them what to do next. “Buy Now” on the product page. “Proceed To Checkout” on the basket page.
• Don’t annoy them. Everyone hates pop-ups. Adverts for other products, such as cross-sells and up-sells, are fine as long as they are relevant and not too obtrusive.
• Don’t bore them. Only ask for the details you need when they are paying. You don’t need their full life history, their mother’s maiden name or their favourite type of cheese, so don’t ask. Ideally, put all of your questions on one or two pages.
• Sell your product. Tell people why they should buy this particular product, and why they should buy from you. You put shop assistants in a shop, and they do more than just recite product specifications at you.

Landing Pages

I’ve already told you (in a previous instalment) what a good landing page needs, but sometimes, it’s not quite that simple. Suppose that you are selling digital cameras, as you were in the example earlier.

Somebody who searches for digital cameras may have a range of technical knowledge – some will be familiar with all of the technical terms, whereas others will just know that they want something that takes pictures.

Each of these visitors would probably look for different things on your website. A less technical person may want a guide, to tell them what to look for in a digital camera, and whether they really need the different features, whereas a more technical person may prefer to see the products with simply stated features, without being lectured on things they already know.

But both are searching for the same term, so both will land on the same page. Should you land them on a guide to buying digital cameras, or a thumbnail page?

The correct solution here would probably be to test different landing pages, and see which one works better. Adwords has the facility to let you run different adverts at the same time – in this case, the adverts would have the same text, but with different target URL’s.

For more advanced tests, Adwords has a tool called Website Optimiser. To directly quote Google:

Website Optimizer works by testing different versions of the content on your live site to see what will lead to the highest conversion rates. You provide us with the original content of your site and the different variations of that you’d like to try out. We run an experiment on a portion of your site traffic using those variations, to see the effect of each variation on your conversion rate. When we’ve collected sufficient data, we’ll provide you with a full report of our findings that can help you determine what changes to make to your site.

It works quite well most of the time, and gives you a free and easy way to serve up alternative landing pages and track their performance.

Advert Text

Many experts never look beyond the website when talking about conversion rates, but writing the correct advert text has a role as well.

Remember my example from an earlier post?

If you sell luxury leather furniture, and you bid on “leather sofas”, there will be some searchers who are looking for cheap, low quality leather furniture. Writing an advert to maximise the click through rate will attract many of the wrong people, and end up costing you money. On the other hand, you don’t want to put people off by implying that your sofas are expensive. So you need to write an advert that gets as many potential buyers through, but as few inappropriate visitors as possible.

In this case, you might use a title like “Top Quality Leather Sofas.”

The objective here isn’t to get people to click on your advert - it’s to stop them from clicking on your advert.

Suppose that you sell armchairs, specifically aimed at the male market (maybe they’ve got a built-in beer fridge?). Your advert reads:

Luxury Leather Armchairs

Five-Star Leather Quality

Free Delivery, Money Back Guarantee

Overall, your advert has had the following results:

So your campaign made a total profit of £500. Nice.

But this may be split as follows:

Writing your advert more like this:

Leather Armchairs For Men

Five-Star Leather Quality

Free Delivery, Money Back Guarantee

That may put off most of the unprofitable lady visitors, and make you more money.

In this case, you were trying to get rid of unprofitable clicks, in order to make more profit overall. Note that in the above example, your click through rate would drop from 5% to 4%, but the 60% increase in profit should outweigh any additional costs from increased bids (since click through rate is a factor in your Quality Score).

Not Keyword Selection

Pardon?

What I’m saying is that you should not select keywords solely on their ability to convert. You should get rid of keywords that have a 0% conversion rate after a reasonable number of clicks, but if your campaign has an overall conversion rate of 5%, then getting rid of a keyword with a 3% conversion rate may not be in your best interests.

Remember, the objective is to maximise profits, and improving your conversion rate is merely a way to do that.

There are two situations to consider here – either you are spending your full daily budget or you aren’t.

If you aren’t, then clearly, any keyword that is making you money should be kept, and any that isn’t should be optimised (and removed if it can’t be made profitable).

If you are, then you should always look at the keywords that generate the lowest ROI, and either remove them, or make them more profitable (see later). These are not necessarily the ones with the lowest conversion rate.

Suppose that you’ve got two keywords, [leather chair] and [reclining chair].

Their performance figures are as follows:

The first keyword has a conversion rate of 10%, the second 5%, and yet both make the same profit, from the same level of spend. So, where possible, you’ve got to do the calculation, and base your decisions on which keywords make the most money per pound of spend, if your budget is restrictive.

Of course, if a keyword is getting no conversions at all, then it can’t make money. Just make sure that you give it time. If a keyword has a profit per sale of £500, and a cost per click of £0.50, you only need a 0.1% conversion rate to be profitable, so don’t panic if you’ve got no conversions after 100 clicks…

Not The Amount That You Bid

In most cases, the conversion rate is not affected very much by your position in the search results. I can only think of a few cases when this is not true.

One may be if you are bidding on a generic keyword like “digital camera,” where a lot of people are doing research on what they are looking to buy, they are more likely to click on a few of the sites at the top of the page while they collect information. So the top positions may see a lower conversion rate than the ones further down the page. This is less likely to be true with more specific keywords like “Sony Cyber-Shot DSC-T100S.”

Then again, the way that the majority of people advertise on Adwords, the bids for the top positions on generic terms are frequently way too high, so a well-run campaign probably won’t appear in these positions anyway – it’s just not profitable.

Advert Text

Advert text is arguably the most critical aspect to get right in your campaign. You have three key objectives, all linked to the advert text, and often directly conflicting with each other.

• Quality Score
• Conversion Rate
• Clicks (or Click through Rate)

The advert is critical to the Quality Score, as it is checked for ‘relevancy.’ In addition, click through rate affects the Quality Score, so you want to maximise this as well. Of course, if your campaign is profitable, you want as many clicks as possible anyway.

The problems start when not everyone that sees your advert is particularly interested in what you’re selling. Maximising the click through rate could just lead to you paying for a lot of clicks that are worthless.

For example, if you sell luxury leather furniture, and you bid on “leather sofas,” there will be some searchers who are looking for cheap, low quality leather sofas. Writing an advert to maximise the click through rate will attract many of the wrong people, and end up costing you money. On the other hand, you don’t want to put people off by implying that your sofas are expensive. So you need to write an advert that gets as many potential buyers through, but as few inappropriate visitors as possible.

In this case, you might use a title like “Top Quality Leather Sofas.”

So, what does Google look for in your advert, when it’s deciding how relevant it is to the searcher? Simply put, the search phrase should appear prominently in the advert, ideally in the title. It doesn’t hurt to repeat them in the body of the advert, but you probably don’t want to sacrifice the appeal of your advert to achieve this.

In order to maximise the click through rate, you need to do a few things, but the most critical is to make your advert stand out. If your advert looks just like everyone else’s, why are people going to click on your advert, or even see it? Searchers do not read every advert in full, as a rule, and they’ll click on the one or two that catch their eye.

So you want to write an advert using different words to your competitors. Numbers, particularly with a £ at the start, will always draw the eye. If your prices are competitive, advertising them could certainly be worth a try.

You want a strong call to action - Google won’t let you say “click here” or “visit us,” but “Buy xxx Now” is allowed. There are plenty of other good words, such as “Free,” “New,”  “Save,” “Fast,” “Great,” and “Amazing.”

You should certainly be looking to extol the virtues of your product. Why should people buy your widget, rather than everyone else’s? Does it have free delivery or an extended warranty? A price promise? Is it the best quality on the market? Is your range extensive? If so, TELL PEOPLE.

But this raises a question – what happens if all of your competitors are also using the search term in the title of the advert? What if they all have the same product with the same selling points?

If you search for injury compensation, everyone’s offering the same deal. You get 100% of the compensation, and it’s no win, no fee. And they all say “Injury Compensation” in the title.

So what do you do? Sacrifice your Quality Score a bit by putting something eye-catching in the title? Fail to mention these selling points?

I’d say that the answer to the first question would be that you should do anything to maximise your click through rate. If your advert doesn’t generate any clicks because it looks just like everyone else’s advert, then your Quality Score doesn’t matter – and a good click through rate can also improve your Quality Score.

I’ve found in the past that challenging people with a question often works – you’ll see that television adverts do this a lot.

“Are you looking for cheaper car insurance?”

“Been involved in an accident that wasn’t your fault?”

“Overworked And Underpaid?”

Try a similar approach, it may work…

Another thought is to make your advert look different to the others – if it appears on the right of the page, and all of the adverts surrounding it are using the full 35 characters on each line, then yours could stand out if it’s very short, with a lot of space around it.

You can’t capitalise whole words in your advert unless they’re acronyms, but one thing that you should certainly be doing is to capitalise the first letter of each word. It Makes The Advert Stand Out Much More Clearly.

The display URL does not need to match the actual URL that the advert lands on. One of the things that Google does when somebody searches for a term is to embolden the keyword in any adverts. So there’s no good reason not to put the keyword at the end of the display URL, separated by a slash. For example, if you sell blue widgets, and your landing page is

www.mywidgets.com/widgets/blue.html

you could try

www.mywidgets.com/blue-widgets

Another option is to use Dynamic Keyword Insertion (DKI). This is a little tool that Adwords gives you that allows you to place the search term into your adverts. You have to specify a default word to put in place if the search term is too long for the line, and this is what Adwords uses to calculate your Quality Score (so this isn’t an easy way to improve your Quality Score), but it can give the impression that you have exactly what the searcher is looking for.

Note that this won’t put the search term that the searcher typed in, rather it uses the search term that it matched to from your list.

For example, if you are bidding on “leather sofas” on phrase match, and somebody searched for “quality leather sofas,” then the words “leather sofas” will be implanted into your adverts.

The syntax for this quite straightforward:

{KeyWord:Leather Chairs}

In this case, if the advert doesn’t fit, Leather Chairs would appear by default. Also, note that if you wanted the words to appear in lower case, use keyword (all lower case) rather than KeyWord. Keyword (first letter capitalised) would capitalise the first word only.

So this could be implanted into an advert as follows:

Buy {KeyWord: Leather Chairs} Here

Finest Leather Sofas And Chairs

Free Next Day Delivery

If somebody searched for “Leather Sofas,” the advert would say:

Buy Leather Sofas Here

Finest Leather Sofas And Chairs

Free Next Day Delivery

If they were to search for “Leather Furniture” (which would be too long to fit), then the advert would say:

Buy Leather Chairs Here

Finest Leather Sofas And Chairs

Free Next Day Delivery

If there’s one absolute fact with advert text, it’s that you should keep testing constantly. Adwords gives you the opportunity to run multiple adverts at the same time, and see which ones are more effective. Use it. Remember that every time you increase your click through rate, you improve your Quality Score, and hence your advert moves up the search results (or you can reduce your bids and stay put). So you win twice!

Landing Page

Google are a little tight-lipped about exactly what makes a good landing page, and indeed website, in terms of their relevancy factor, the Quality Score.

They do, however, offer the following guidelines:

• Use a page that contains relevant and accurate information about your product or service.
• Distinguish advertising from the rest of the content on your site.
• Don’t ask users to register without telling them what they’ll get.
• Have unique content - don’t just copy content from other websites.
• Users should be able to find what your advert promises quickly and easily.
• Be open and honest about what your business is/does (about us page).
• If you are collecting personal information, explain what you intend to do with it (privacy policy page).
• Keep the navigation on your website simple.
• Provide a simple process for purchasing from your site.
• Avoid pop-ups, fiddling around with the back-button and other ‘obtrusive elements.’

To be honest, most of these are good practice anyway, and will build up trust with your visitors, as well as with Google.

One of the points there was a simple purchasing process. This is also critical from a conversion rate perspective – confuse visitors, or make them work hard to buy things, and they’ll go next door.

You should tailor your landing page to what people are searching for. In the digital cameras example, if somebody was searching for Sony Cyber-Shot DSC-T100S, then a good landing page would have

• A good quality picture of this product
• Bullet points with the major selling points of the product
• Technical information available, but not too obtrusive
• Cross-sells like a case, and batteries
• Product reviews (again, not too obtrusive)
• A clear price and call to action (Buy) button visible without scrolling down the page.

On the other hand, if somebody was just searching for a Sony Cyber-Shot, then your landing page would ideally be a bit different. They don’t necessarily know which product they want, so a comparison page, or failing that a thumbnail page would be appropriate. It should contain

• A list of all of the appropriate products
• Prices and product attributes for comparison
• A small editorial, explaining why the user should consider each.
• Product reviews

Given all of this information, visitors will be more likely to make a decision whilst on your site, and buy from you, rather than looking at other sites.

Bidding

So, how much should you bid when you first create a campaign?

Obviously, it’ll depend on the keywords that you are bidding on. It’ll also depend on your daily budget. You need to get a sensible number of clicks in order to make it possible to begin optimising the campaign - I’d suggest an absolute minimum of 30 clicks per day, so 1/30 of your daily budget should be an absolute upper limit.

At this point, you can turn to the traffic forecast tool built into Adwords. After you input the keywords when setting up an Adgroup, you are taken to a page asking for your bid. Enter a bid, and hit the button to forecast the amount of traffic and position that your advert may generate. Try a few different bids, and pick one that you’re comfortable with. I would suggest that for less competitive terms, you may want to appear in about fourth, and for more competitive terms, about seventh.

Once you start optimising your campaign, these will change anyway – at the moment you’re just getting things going.

One of these traditions is wild variations in your search traffic, but is this really true? And if so, should you be expecting your traffic to skyrocket, or plummet?

It’s certainly the case that Bank Holidays are big business for the high street. According to the British Retail Consortium, the ‘eight days of Easter’ will be worth £8Bn to UK retailers, but should retailers expect this spike to be reflected on their websites, or will online sales actually fall as a result?

One fundamental truth in retail is that people like a bargain. The high street has developed a culture of ‘Bank Holiday Sales’, to cash in on this, and many people flock to the shops to cash in on this. Whether the retailers are responding to the customers or the customers to the retailers is a chicken-and-egg question – an Ouroboros of discounts and shopping.

But the same culture doesn’t really exist online – short term sales are far less common, for a number of reasons. Perhaps the biggest reason for this is the difficulty in advertising such a sale – whilst a high street shop can simply stick a big pink banner in their window, e-commerce sites have no such facility.

Of course, established retailers can raise awareness of their sales by e-mailing their customer base, but this medium has been so heavily utilised that customers just don’t respond in the way that they would in a shop. If you’ve ever purchased anything from Play.com, you’ll know that there’s almost always a sale on!

Maybe it’s not just about shoppers responding to a bargain – there is a distinctive feeling in searching out a great deal in a store, trawling the high street for something that until now you’ve never known you need – whether it’s yet another little black dress or a six-speed bread-maker.

Online shopping isn’t the same experience; no matter how well-designed a website is, window-shopping is just not the same somehow. Maybe it’s the way that you have to navigate a website – basically, you have to know what you’re looking for before you go onto the site, or maybe it’s just the experience of being physically surrounded by ‘objects of desire.’ For many people however, online shopping just isn’t as much fun as ‘the real thing.’

But what about other advertisers? For many business-to-business advertisers, bank holidays are almost a thing of dread – in the words of Bob Cratchit, “there’s nobody to do business with.”

Whilst it’s true that there is sometimes a small dip in traffic on a Bank Holiday Monday, our experience has been that this is surprisingly limited (with the exception of Christmas and New Year), and any loss of business is caught up the following day.

So perhaps the answer is that whilst Bank Holidays are never going to make a dramatic difference to your search marketing campaigns, for savvy retailers perhaps there is a way to leverage them to your advantage.

(Photo: Dorchester Dorset)

These are questions that everyone that manages an Adwords account has asked at some point. So I decided to find out. As a large agency, we are in the position of having far more data to analyse than most people, and by manipulating this data, we’ve found some very clear and consistent trends, that have given us unique insights into what click through rate Google are looking for in order to consider you ‘relevant’, and how they take your position into account.

So here’s what I did. I pulled out the data for the last 30 days for every keyword, in every Ad Group, in every campaign, in every account that we manage. I stripped out any Phrase and Broad Match keywords (since QS is calculated on Exact Match), Google’s Search Partners data (since this is ignored in the calculation) and brand terms (which would skew the click through rate analysis significantly – more about this later).

I then looked at the average click through rate in each position from 1 – 8 (in increments of 0.1), for each keyword Quality Score. By graphing the results, it wasn’t hard to detect a pattern. Here’s the CTR’s by position for keywords with a QS of 10:

Note that I took the total impressions and clicks for each position, effectively weighting the CTR for each keyword by the number of impressions, so that low volume keywords didn’t have an exaggerated impact on the performance. You can also see why I removed brand names from the above analysis – the position 1 CTR would have been influenced excessively…

From the above graph, it’s possible to get some sense of what click through rate Google expects from a keyword in order to give it a QS of 10. And by fitting a curve through the data, it’s possible to get a reasonable indication of where you should be in order to pick up a Quality Score of 10:

Based on this line, it appears that Google expect the click through rate in any position to be about 65% of the next position up. So where position 1.0 has an average click through rate of 34%, position 2 has an average click through rate of 22.1%.

Whilst this line fits nicely through this data, what happens if you look at other Quality Scores? Will Google expect a similar drop-off in click through rate? If not, then this number is just a convenient fit for this data. Here’s the graph for Quality Scores of 9:

And 7:

And 6:

The same decay of 0.65 has been drawn on each, and in all cases, it appears to fit well. So this appears to be Google’s estimate of what ‘should’ happen to your click through rate every time you drop a position – you lose just over 1/3 of your clicks. Of course, this doesn’t really work once you drop below about 8^th, but it’s certainly a useful thing to know. Indeed, the line fits just as well with a Quality Score of 5, 4 or 3 as well.

Based on these fitted lines, you can estimate the following average click through rates by position, and how they are converted into Quality Scores:

Using this, you can potentially ‘health check’ your account. If you have a click through rate of 4.5% in position 4, you should have a Quality Score of around 7 or so. If you are getting less than the predicted Quality Score across the bulk of your keywords (excluding brand, on Google only, on Exact Match), then it’s a sign that your account has other issues, possibly with the landing page, keyword relevance or the overall account quality.

I tried to take this analysis to the next level, by calculating the predicted Quality Score for each keyword, based on its click through rate and average position. Unfortunately, whilst the fits are very good looking at large amounts of data, on a keyword-by-keyword basis, the results can be quite inaccurate, underlining the fact that advert relevance, landing page relevance and account quality also have a significant bearing.

One final point worth noting is that the overall figures for the Quality Score of 7 are a little misleading. Most keywords seem to be given a Quality Score of 7 initially, until Google calculates a value based on their performance. Because far more of our keywords get Quality Scores higher than 7 than get less than 7, the effect of new keywords on the ‘7’ performance is pushing up the click through rate significantly.

Realistically, none of this has any bearing on the way that you manage your account. Regardless of your click through rate, you should always be looking to improve it (though obviously not to the detriment of your conversion rate), and the Quality Scores that Google quotes in the data are only indicative – the true Quality Score figure is probably much more precise, and keywords that appear to have the same Quality Score could actually have quite different values (6.6 and 7.4 could both round to 7).

But in terms of offering insight into what may happen to your traffic volumes if your position were to change, it’s very interesting. If one of the advertisers above you were to drop out, you could potentially be looking at a 50% uplift in clicks. If you’ve got a restrictive budget, this could cause problems. Similarly, a new competitor appearing above you could cost you 1/3 of your clicks…

Generating Lists of Keywords

There’s a big contradiction in PPC at the moment. In general, the more specific a keyword is, the more likely the visitor is to convert (as the searcher has a better idea of what they want, and is likely to be further along the decision-making and buying process), and as a result, the more you should be able to pay per click.

This is sensible, because the cost per conversion can be expressed as:

Cost per Conversion = Cost Per Click / Conversion Rate.

So, the higher the conversion rate is, the more you can pay per click, and still get an acceptable cost per conversion.

As a result, you’d expect the bids on specific terms that convert well to be higher than for more generic terms. But in fact, the opposite is frequently true.

If, for example, you were to bid £0.25 per click for “Sony Cyber-Shot DSC-T100S”, your advert would almost certainly appear much higher than if you were to bid the same amount for “Digital Camera”.

There are a few possible reasons for this. The more generic the term, the more traffic it will generate. It’s possible that companies feel that it’s worth bidding more for these terms, hoping to make the extra spend back on volume of sales. I can’t advise you strongly enough not to go down this road…

Alternatively, it could simply be that fewer companies are bidding on the more detailed terms, as many companies do not feel inclined to put in the work involved for low volumes of traffic. But as these are the most profitable terms (high conversion rate, low cost), this is short-sighted on their part.

Whatever the reasons, this is the current situation, and you need to know how to make the most of it.

Clearly, attacking this ‘long-tail’ of keywords should be a priority, whatever your situation. And it applies to almost all areas of online retail – a search for “estate agents In Manchester” is more likely to convert than a simple search for “estate agents”, and a search for “14 inch blue widgets” will convert better than “portable widgets”.

So the first step should be to list your products or services in their entirety. Then consider words that should or could mean the same.

In the example above, your product was:

Sony Cyber-Shot DSC-T100S

But people may also search for

DSC-T100S

Cyber-Shot DSC-T100S

Sony Cyber-Shot DSC-T100S Digital Camera

Sony Cyber-Shot DSC-T100S Camera

DSC-T100S Digital Camera

Cyber-Shot DSC-T100S Camera

DSC-T100S Digital Camera

Cyber-Shot DSC-T100S Camera

Sony Cyber-Shot DSC-T100S Digital Cameras

Sony Cyber-Shot DSC-T100S Cameras

DSC-T100S Digital Cameras

Cyber-Shot DSC-T100S Cameras

DSC-T100S Digital Cameras

Cyber-Shot DSC-T100S Cameras

Once you’ve done this for all of your products, you need to consider the next level up, if there is one.

So for the above example, you might want

Cyber-Shot

Cyber-Shot Camera

Cyber-Shot Digital Camera

Cyber-Shot Cameras

Cyber-Shot Digital Cameras

Sony Cyber-Shot

Sony Cyber-Shot Digital Camera

Sony Cyber-Shot Camera

Sony Cyber-Shot Digital Cameras

Sony Cyber-Shot Cameras

Next, you repeat it for the next level up again (if there is one).

Sony Digital Camera

Sony Camera

Sony Digital Cameras

Sony Cameras

And again…

Digital Camera

Digital Cameras

Camera

Cameras

Next, look for possible misspellings. For example, Cyber-Shot could be spelled Cybershot or Cyber Shot quite easily. And DSC-T100S could be written DSC T100S or DSCT100S, or even just DSC-T100.

You may not want to bid on the last few terms, unless you’ve got a very large budget to spend. The conversion rate is likely to be poor, since people searching for Digital Cameras don’t know which camera they want, how much it’ll cost or even what features they are looking for. These people are looking for information first and foremost, and will probably bounce around a few different websites deciding what they are looking for. If you’re going to make a success of these terms, you probably want to have a lot of comparison information and product reviews and the suchlike. If you can persuade people to make their decisions based on the information on your website, then you can still get a respectable conversion rate. Personally, I’d steer clear of them, at least until you’ve got a stable, profitable campaign…

On top of these keywords, you may want to add keywords with certain features, e.g.

8 megapixel digital cameras

7 megapixel digital cameras

Bear in mind that if you’re planning on bidding on these terms, you need a landing page that’s appropriate.

If you start out with all of these keywords, you can always trim them back once you can see which keywords are working and which ones aren’t.

Match Types

One of the features of Adwords is the ability to bid on the same keywords using different types of matching – specifically, Exact Match, Phrase Match and Broad Match.

If you were to bid on “Sony Cyber-Shot” on Exact Match only, your advert would only appear if the searcher was to search for the exact term “Sony Cyber-Shot”. If they searched for “Sony Cyber-Shot DSC-T100S Camera’,’ your advert would not appear.

If you were to bid on “Sony Cyber-Shot” on Phrase Match, your advert would appear if the search contained those words, even if there were other words before or afterwards. So, your advert would appear for “Sony Cyber-Shot DSC-T100S Camera”.

Broad Matching has two forms. Firstly, it’ll return your advert if people search for terms containing your words, even if there are other words in the middle, or the words were in a different order. So, if the search term was “Cyber-man Episode Of Doctor Who Shot On Sony Film”, your advert would appear (maybe). Secondly, Adwords will think of similar words (in its opinion), and bid on those on your behalf. You can limit this by using Modified Broad Matching – by putting a + sign in front of certain critical words in the keyword, you can ensure that that exact word, or a close variant such as a plural, must appear in the search query.

One of the components of the Quality Score (Google’s measure of the relevancy of your advert - this will be discussed in detail later) is the similarity between the search term and the keyword that you’re bidding on.

Suppose that you are bidding on [blue widgets] (square brackets mean ‘Exact Match’ and somebody else is bidding on “widgets” (double quotes means ‘Phrase Match’). If somebody searches for blue widgets, the keyword relevancy aspect of the Quality Score would be higher for you than for the other advertiser.

However, if you were bidding on [blue widgets] and somebody else was bidding on “blue widgets,” then you wouldn’t get any advantage, as Google considers that the two are equally relevant.

So, how should you use the different match types? I’d suggest that you should start with Phrase Match on everything, and use the Search Query report to look for keywords that perform particularly well or badly. Also put all of your keywords on Exact Match – it may be that the exact term performs very differently from variants, and you need to be able to see this variation.

If some searches perform badly, use negative match to stop your adverts from appearing for them. If they perform very well, add them on exact match and phrase match, and crank up the bid on exact match (since you can afford to pay more per click on a better-converting keyword).

Broad match is a bit of a gamble, as it can throw up some very different keywords. Once you’ve got your keyword list in place, run broad match for a few days, and see if you can get some new ideas for keywords.

Here’s an example: You start out with a keyword list as follows:

“widgets”

“widget”

“blue widget”

“blue widgets”

“red widget”

“red widgets”

After a few days, you look at the search query report, and find that a lot of your searches are for cheap widgets. Your widgets are very reasonably priced, and so these convert quite well (about the same as the rest of the campaign). So you should add these in on phrase match. On the other hand, you’re also getting searches for rechargeable widgets and waterproof widgets, which you don’t sell. So you change your keyword list to:

“widget”

“widgets”

“blue widget”

“blue widgets”

“red widget”

“red widgets”

“cheap widget”

“cheap widgets”

“cheap blue widget”

“cheap blue widgets”

“cheap red widget”

“cheap red widgets”

-rechargeable

-waterproof

-”water proof”

Your campaign runs for a few more days, and you realise that “designer widgets” has had significant traffic, and it’s converting much better than everything else. So you decide to bid more on this keyword. However, you don’t want to bid more on “cheap designer widgets” as your designer widgets are very exclusive, and it doesn’t convert as well. So you add in [designer widgets] as an exact match, and pump up your bid on that keyword alone. You also add in “designer widgets” on phrase match, but with the same bid.

And so on. Eventually, your keyword list will stop expanding, and you’ll have a lot of keywords on phrase match, and a handful on exact match with higher bids.

At this point, try broad match. You may find some interesting alternative keywords that mean the same thing. It’s unlikely, and you should monitor it closely, as Google includes something called expanded matching. The idea is that if you bid on a term, Google comes up with other terms that it thinks are similar, and that you’ve probably just forgotten to include.

The results of this can be variable – sometimes broad matching returns keywords that convert at a better price than your existing keywords, but other times it can be completely pants. If you sell wooden chairs, and use broad match, it’s more than possible that your advert will appear with a search for metal chairs. This can drag down your clickthrough rate, and hence your Quality Score.

Grouping Keywords Into Adgroups

As I’ve already said, since keywords in the same Adgroup will all get the same advert, it’s important to group them carefully. If you are a retailer, you should probably put each individual product within its own Adgroup, and then write an advert that specifically mentions this product. This is also important as Adwords considers an advert more relevant if the keyword appears in the advert (particularly in the title) and so will position your advert more highly for your bid.

There’s a useful tool called Dynamic Keyword Insertion which is worth a mention here – you can insert what was searched for into your advert (sort of, more about that in a minute), and so make an advert that relates to a number of products. For example, you could write an advert that says “Buy a 7 Megapixel Camera Here” if they search for a 7 Megapixel camera, and says “Buy an 8 Megapixel Camera Here” if they search for an 8 Megapixel camera. Generally speaking, this is more trouble than it’s worth, though, so I’d avoid using it, at least at first.

Here is Part One and Two if you missed them. Comments are always welcome!

An Adwords account is like a set of Russian dolls; open each element, and you find a smaller one.

An account comprises one of more campaigns, each made up of a number of Ad Groups, containing a set of keywords that in turn can have different match types.

One of the most common mistakes made in Adwords is to fail to utilise the functionality that this structure provides. An account comprising a single campaign, containing a single Ad group, with every keyword appearing with just one match type, is almost certainly not optimal. This is a structure that I’ve seen far too often when assessing accounts that advertisers have created themselves – they don’t understand the elements, so they tend to ignore them.

So how should it be done?

Accounts

Unless you have multiple websites, each with a significantly different offering (in terms of products, services, or general purpose), then you should only have a single account. If you run multiple accounts bidding on the same keywords, you will display multiple times for that search result. Unless your websites fit into the exceptions above, this is called double-serving, and will get you told off by Google.

Campaigns

Your account comprises a number of campaigns, each of which has its own budget.

There is a clear downside to using different campaigns if you have a tight budget. If one campaign runs out of money, and another has a surplus, then some of your adverts won’t appear even though you have money left over. Campaigns cannot take budget from other campaigns, so you should never split your account without a good reason.

There are times when it is absolutely essential to run separate campaigns, however. Suppose you are selling toasters, and advertise on both PC’s and mobile phones. It’s very likely that you’ll want to bid on ‘toasters’ on both, but the performance is likely to be different (users are more or less likely to buy), so you’ll want to make a different bids. In order to put different bids on the same keyword for different media, you’ll need separate campaigns.

Another example would be if you know that your campaign performs completely differently at different times of the day, or on different days of the week. Google provides you with basic tools to manipulate your budgets and bids at different times, but you are generally far better off running different campaigns.

Finally, most of your targeting settings are set at campaign level. As we’ll see in a later blog, the Display Network works completely differently to the Search Network, and so you need to manage it completely separately. The same is true if you are running more advanced options like Remarketing.

If budgets are unlikely to be an issue, you may want to use different campaigns to target different subsets of your products, or different types of searcher.

For example, one of our advertisers has a campaign for each product type (dresses, trousers etc). Another has separate campaigns to try to target males and females, based on the type of searches that they enter.

The first example allows them to have an additional level of grouping, so that they can have adverts targeting different clothes styles and sizes – but more about that later.

The second example allows the advertiser to try to manipulate the gender split visiting their site. As they are a dating agency, this is a very powerful tool to them.

Ad Groups

Every Ad Group has its own set of adverts. You can also set bids at Ad Group level (this is recommended – again, this is covered in detail in a later blog).

As I said in the first part of this series, Adwords has become seriously competitive. If your advert is competing with ten others for the users’ attention, it is absolutely critical to make it as relevant as possible.

I ran a search for Panasonic DVD Players, and these are the nine results that appear above the fold:

Of these nine adverts, eight mention Panasonic, four mention DVD players, and only three mention Panasonic DVD Players.

Two of the adverts aren’t even for DVD Players!

So which would you, as a user, be likely to click on? I’d suggest either M&S, Pixmania (though this is B2B only, so perhaps not) or Shopcompare.

It’s slightly ironic, then, that none of these sites actually sell Panasonic DVD players!

So what’s happened here? We’ll look at keyword selection and matching types in a later section, but it appears that companies like Superfi and Play have put all of their Panasonic keywords into one Ad Group, and Argos have done the same with their DVD Players keywords.

The lesson here is to make sure that your adverts relate to your keywords. The more similar your keywords are within an Ad Group, the more targeted your advert will be; the more disparate they are, the less targeted your advert will be.

So group your keywords carefully. You’ll get better click through rates, and you’ll also be able to manage your bids more effectively.

Keywords

Finally, select your keywords based on what you actually sell! I’ll look at matching types, which have probably caused the downfall of Panasonic and EBay above, in a later episode.

Whenever somebody searches for something, clicks on your advert, but doesn’t buy anything, you’ve wasted your money. If you bid on products or services that you don’t sell, you’re in danger of wasting your money. If your advert is accurate, you’ve got a fighting chance, but it’s still not a good idea in the majority of situations

Setting Up Your Campaign

Before you switch your computer on, you need a credit or debit card. Google will accept bank transfers as well, but if you do this, you’ll have to pre-pay, and every time there’s a problem, your adverts (and customers) will disappear.

You’ll also need to dig out your VAT details – the rules are probably different from country to country, but in the UK, you’ll need to declare your own VAT, as Google Europe is based in Ireland.

You also need to pick your daily budget - this is entirely up to you, but if this is your first campaign, it’ll probably take a little while to make it profitable, so don’t spend more than you can afford. Once your campaign’s making money, of course, things will change and you can spend as much as you like (subject to cashflow limitations).

Once you’ve got these things, you need to create your list of keywords, and group them into Adgroups. Then write some adverts for each Adgroup, pick appropriate landing pages and set your daily bids. I’ll go through each of these stages in the next part.

In case you missed my last post, on Google AdWords, here is the link: http://www.epiphanysolutions.co.uk/blog/google-adwords-expert-agency-secrets-revealed-part-one/. Look out for next week’s post on Thursday!

Even experienced users of Adwords, such as agencies, make mistakes, and can cost their clients money and ultimately cost themselves accounts. It’s very rare that I’ve accessed an account, whether managed by an agency or by an individual, where mistakes aren’t being made.

But fear not! Help is at hand. Perhaps the most easy-to-understand, but powerful guide ever written on how to advertise on Adwords is now being serialised on the Epiphany Solutions Blog for the first time.

This is the first part of that serialisation. It gives an overview of what Adwords is, where it is and isn’t useful, and the advantages and disadvantages of PPC against SEO. Future parts will look at each aspect of an Adwords account, and show what you should do in order to deliver amazing results, and perhaps more importantly, why the account should be set up and managed in this way.

Why Advertise Using Paid Search?

It’s a fair question. The days of PPC being a goldmine, with undiscovered pots of gold (also known as searches with few or no adverts) have gone. It’s one of the competitive – if not the most competitive – forms of advertising that there are, and you are going head to head with your rivals.

It’s definitely not for the faint-hearted. Your advert is likely to be in the middle of a bunch of competing adverts, all vying for the searcher’s attention. Even if you get the searcher to your website, they probably won’t do what you want – many sites see only one in every hundred actually making a purchase. Is it worth paying for the 99 that don’t?

It’s interesting, then, that a properly managed account is virtually guaranteed to make money.

No seriously, I mean it. Do the sums. Assuming that the visitors to your site have a value (i.e. you make money from at least some of them), then as long as you bid less than the visitor is worth, you can’t lose money.

For example, suppose 2.5% of your visitors buy something, the average order value is £100, and 25% of the order value contributes to profit. Each sale is worth £25 to you, and so each click is worth £0.625 to you (since every 40 clicks generate one sale, which is worth £25 to you).

So if you only pay £0.50, you will make money. Of course, if 30 other advertisers are all making £5 from each click, they are all going to outbid you, pushing your advert down into the nether regions, and you’ll get very few clicks – so you won’t make much money.

Optimising Your PPC Account

This simple example highlights two of the themes that will keep repeating throughout this whole guide.

Firstly, you should be basing the amount that you bid for a keyword on the amount that the click is worth to you. This is pretty obvious if you look at the example above. But there are still many advertisers out there that view success as achieving the highest positions that they can for their budget!

Secondly, since the amount that you bid is limited by the value of the click, you should be doing everything you can to maximise this. The simplest way to do this is to maximise your conversion rate. Every click that doesn’t convert is wasted money, every click that does convert is very profitable.

There is one more critical aspect to successful account management. Adwords is an auction – if you bid more, you appear higher. However, it is not a fair auction. Google multiply your bid by your Quality Score in order to determine your advert’s position, and the amount that you actually pay per click for your keyword.

Quality Score is Google’s measure of how relevant your advert is. Google want to make sure that their adverts are as relevant to users as possible for two reasons. Firstly, if their search results aren’t high quality, users will use another search engine. Secondly, if their search results aren’t high quality, nobody clicks on them, and Google don’t make any money. And Google likes making money.

Quality Score is covered in a lot of detail later on, as it is really quite important. But for now, suffice to say that every change that you make to your account should serve one or more of the following purposes:

• Increase the volume of clicks that you are generating
• Increase the profitability of a click, either by
  • Increasing the value of the click, or
  • Reducing the cost of the click

Often these two objectives directly conflict. For example, bidding more increases your clicks, but results in each click being less profitable. A new advert may improve your click through rate, but by driving less qualified traffic. New keywords may cause a similar trade-off.

Each individual section will look at this trade-off, and how to make the right decisions, and every recommendation made will be based on improving one or both of these objectives.

Isn’t It Cheaper And Easier To Stick To SEO?

Well – it can be, and you should probably do this regardless of whether or not you do PPC, but optimising your website to appear high in the natural, ‘free’ results takes a lot of time, work and money. In the long run, SEO generally pays off, but you’ll be spending a lot of money for a long time before you get results on popular search terms. And if your competitors are also doing SEO, then you may never catch up.

Pay Per Click gives you immediate results, and unlike other forms of online (and offline) advertising, you know for certain that you’ll get results. Instead of paying for an advert that people see, who then may not even visit your site/store, with PPC you pay only when somebody walks through the door. As long as there’s a good chance of that person buying something, then there’s a clear link between what you’re paying and how much you sell.

Also, you have total control over how much you pay for your visitors, and who visits your site. If something isn’t working, you can stop it immediately, and if something’s very successful, you can do more of it.

PPC can also support your SEO strategy, showing you which keywords convert, which landing pages you should be trying to drive visitors to, and which messages appeal to the searchers. And many studies have shown that websites with both natural and paid search listings get more traffic than they would expect to achieve from the two individually – the whole can be greater than the sum of its parts.

Hope you enjoyed Part One of Google AdWords Expert — keep your eyes peeled each Friday when I’ll post the next part.

In the meantime, please feel free to comment or leave questions.

It appears that an innovative entrepreneur in Twickenham has spotted that all you need to do in order to achieve a high ranking in Google is to buy a domain with the target keyword in it.

OK, it’s a little more complicated than that. For those not in the know, here’s how it works…

Google wants to display relevant websites, and two of the key things that they look at are the number of links to a website (which shows how ‘important’ the site is), and what the anchor text says (the text on the link to the website, which tells Google what the page is talking about).

But clearly this system can be gamed – if you want to rank for a keyword, all you have to do is to buy huge numbers of links, and use the same anchor text. Google has all kinds of checks in place to make sure that this doesn’t work – if all of your links have the same anchor text, Google will ignore them and possibly penalise you.

So how does this relate to World Stores? Well, there’s an exception that bypasses all of Google’s monitoring – they can hardly penalise a website for having the name of the website in the anchor text, can they? If a website were to link to Amazon, and the text said ‘Amazon,’ it would hardly be a sign of dodgy activity.

So if you have a website called ‘Wooden Beds World,’ you can put ‘wooden beds world’ in your anchor text as often as you like, and Google won’t penalise you. And as a result, you get great rankings for wooden beds, and similar keywords.

This probably isn’t news to anybody that uses Google, let alone anyone in the industry – search for almost anything nowadays, and it seems that half of the results are obscure websites that appear to rank for no other reason than they have the keyword in their domain name.

Clearly, this doesn’t offer a great user experience – search for Wooden Beds, and two of the top three websites belong to World Stores. You’d hardly call the user experience different:

The two websites have very similar ranges, the same product images, prices, and the same delivery options. They are even using the same fonts and buttons!

The same is true on Bing, where they rank first and second, and to a lesser extent on Yahoo, where they rank fifth and sixth.

“So why don’t Google do something about it?” Well to be honest, it’s not all that simple. After all, it’s commonplace to put the name of a website in the anchor text – as I said earlier, it’s hardly an indication that the website is up to something naughty. However, as this seems to be becoming more and more of an issue, I think it’s just a matter of time before Google clamp down on this. When Matt Cutts covered this topic recently (http://www.youtube.com/watch?v=rAWFv43qubI), he indicated that Google were going to take action about the problem soon.

If I were a company like World Stores, I’d be making a point of keeping as quiet as possible, rather than doing interviews with the BBC, and making statements about how they “reverse-engineered the retail process.” They are virtually spitting in Google’s all-seeing eye, and sticking their tongue out!

It’s interesting – there’s not a single ‘traditional’ natural search in sight. I can see eight paid search adverts, three news results, three real-time results, and not a single ‘natural’ search result. OK – so you could argue that the news results are natural, but the point is that no matter how well-designed your website is or how much time and effort you’ve spent getting other websites to link to you, if you haven’t put any new content on your website in the last couple of days, you’re not getting above the fold.

If the paid search results contain product listings, things can be even worse:

Google now allow paid search adverts to have up to 60 characters on the first line of the advert, making them appear as long as the natural search results. The image adverts and website reviews also draw your attention away from the organic listings.

One more – here’s a search for ‘cosmetic surgery london’:

I’m seeing a big map and four paid search results. No sign at all of any organic listings…

Cynics would say that Google are just trying to cash in – they don’t make any money from organic listings, so they don’t care about them. However, examples like the last one reduce the number of paid search listings as well, so there must be more motivating these changes than just money.

Google would tell you that they’re trying to produce more relevant search results. If you search for something that’s been in the news recently, then a page of search results from last year may not be relevant. If you are shopping for a specific product, then shopping results and images of the products and prices that an advertiser actually sells may well be useful. Also, if you’re looking for a chemist in London, a map showing you the closest ones would certainly be useful.

How much value are they really adding to the search results though? I searched for a Samsung Galaxy S, because I was recently in the market for a new phone, and I wanted to see phone retailers, not news articles about phones. I know what an Ottoman bed looks like, and (going out on a limb here), I’m guessing that they are going to look quite similar to each other. If I was looking for some cosmetic surgery, I’d probably be willing to travel across town to get to it, rather than just going to the nearest one – or maybe that’s just me…

I suspect that the real reason that Google are constantly fiddling with their results pages is nothing more complex than a desire to keep ahead of the competition. If Yahoo or Bing try something different, Google need to add something new, to show that they’re ‘still the daddy.’

If this is the case, it’s a little ironic. After all, one of the reasons that users flocked to Google in the first place was the simple, uncluttered layout of their homepage and search results.

Sergey Brin and Larry Page wrote a paper when they were at Stanford, stating that “we expect that advertising funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers” (http://infolab.stanford.edu/~backrub/google.html). As I said earlier, some people would argue that they are proving themselves right. Personally, I think that the truth is slightly different – that the needs of the consumers have been sacrificed (at least in part) by Google’s determination to maintain their dominant market position.

I don’t believe that this is an intentional shift in focus by Google – every one of the additions they’ve made to their search results could be useful to users in the right context. However, it appears to me that by trying to be too innovative, too quickly, they are just cluttering the search results with irrelevance. If Google want to improve their search results, then for me, the key isn’t developing as many augmentations as possible for their SERPs, but learning how to target them better.

If you assume that the quality of traffic driven by the two different is the same, then you can simply look at the click through rate – the conversion rate won’t vary. This is a reasonable assumption if the key messages and tone are similar, and if the website doesn’t receive many conversions (or can’t track them), it’s necessary.

However, if there is a chance that the different adverts are likely to drive a different type of visitor to the website, and sufficient conversions pass through the adverts, then it’s important to look at conversion rates.

But does this really go far enough?

Consider a casino website. They can promote their site using a small, easy-to-claim bonus, which small-time players can probably claim, or they can promote a much larger bonus, that only high-stakes players are likely to qualify for. Which will work better?

It’s likely that promoting a higher bonus is likely to deliver a higher click through rate. Or is it? Are searchers savvy enough to realise that they can’t claim this bonus? If so, they may be put off by this, and would respond better to a lower, more realistic bonus.

What about the conversion rate? If users weren’t put off by the higher bonus, will they be put off once they read the terms and conditions? If the landing page also has more realistic bonuses as well, this may not be a serious problem. On the other hand, has the bigger bonus attracted newbies – potential players that aren’t aware of how these bonuses work? If so, they may click on multiple websites before converting.

Is this sufficient? Can you tell from the click through rate and the conversion rate which is the more effective advert? Well – no.

Consider the likely behaviour of the visitors that do convert. If they’ve been attracted by different bonuses, their behaviour could vary significantly. If the bigger bonus appeals to the high stakes gamblers, then their lifetime value could be substantially more. On the other hand, if the people attracted by large bonuses are just bonus-hunters, playing just enough to claim their bonuses before moving on, then their lifetime values may be lower.

It may be that this can’t be answered – it depends on the back end systems that the advertiser has. But if you add a variable to the end of the destination url, and the back end can tag up new accounts with this variable, then after a few months, you’ll be able to ascertain which advert generates the more valuable players. Combining this with the conversion rate and the size of the bonus that was paid out, you can establish the value of a click from each advert. And from this, you can make an informed decision.

If this seems like an unusual situation, it really isn’t. Consider another example – a retailer selling discounted designer labels.

The advertiser decides to lead with the size of the discounts in one advert, and promotes the designer labels in another. Whilst the former may generate a higher click through rate and conversion rate, it’s likely that average order values (and possibly margins) will be lower for these customers.

Of course, it takes a lot of customers/players to generate reliable results. One big order or degenerate gambler shouldn’t be sufficient to make a major impact on the outcome. But for advertisers that can perform this kind of analysis, the benefits could be substantial. After all, there are probably thousands of advertisers who have squeezed their average order values, chasing improvements in the click through rate.

It’s called Call Metrics, and if you generate a lot of phone calls from your Adwords adverts, it may of use to you.

Google uses its Google Voice technology (also only available in the USA, so maybe it won’t be appearing quite so soon!) to assign a unique phone number to each campaign. With this number, you’ll be able to track which campaigns have been generating your phone calls, and adjust your bids accordingly.

It’ll be free for the moment, though Google are talking about billing for it in the future. Of course, this kind of thing is nothing new – companies like AdInsight have been offering unique phone numbers that can be assigned to different Adwords campaigns (as well as organic search, referrals or direct visits) that can be displayed on your website.

Still, this is an interesting development – allowing you to include phone numbers on your Adwords adverts without losing the transparency of using multiple numbers.

So who has the power, is it the paid search advertisers, who pay Google’s bills, is it the millions of people that use Google every day, or is it Google themselves?

Google Need PPC Adverts

In 2008, 97% of Google’s revenue came from paid search advertising. Without it, Sergey and Larry would have to move back out of the Googleplex, and back into their (admittedly, probably quite large) garage.

Paid Search is their only real source of income, and yet the advertisers don’t always feel as though Google loves them.

PPC Advertisers Need Google

Well – in most countries they do, anyway. Depending on where you live, Google can command up to 90% (or even more) of the search market.

If an advertiser wants to do paid search, they need Google. Do Google really take advantage of this? The cost per click on Google doesn’t seem to be unreasonable – in fact we’ve got a number of accounts where Google is cheaper than the other search engines.

Google also seem to be quite keen to support paid search advertisers (at least, they do if you spend a lot of money, or you’re an agency). I know that if I have a problem or a question, I can phone or e-mail one of our friends at Google, and they’ll be able to help, or pass me on to somebody else that can.

Google Need Users

But if PPC Advertisers need Google because it’s got most of the market, then clearly, Google have to place the needs of the users ahead of their advertisers. Pandering to the needs of the advertisers at the cost of market share would, in the long term, be very damaging to Google.

So it seems that Google are having to perform a balancing act – they want to make as much money as possible from their advertisers, but at the same time, they want to keep their market share in terms of users.

Google Wants More Money

This is an interesting one. I’ve said that Google wants more money, not that they need more money. The amount of money that Google gets is based on the number of people using it (more clicks = more money) and the amount that advertisers are willing to spend on it.

Perhaps then, this is why Google is so keen to improve the advertiser’s experience. Speaking as Chief Analyst at an agency, I’m going to be willing to pay more for a click if that click is worth more. The best way for me to increase the value of a click is by understanding better which clicks work and which don’t. Knowing that some search queries have a higher conversion rate than others, or that some areas of the country work better than others (and being able to target accordingly) means that I can bid more when and where I get the best results.

On an aside, why don’t Google allow you to see hourly conversions? This would allow advertisers to push their spend into the most profitable times of day, and I believe it would probably increase click spend. This seems like a short-sighted decision on Google’s part (unless it’s for technical reasons). Whilst it would reduce the average cpc when the clicks are worth less, it would also increase them when the traffic is of the highest quality.

If I have a criticism of Google’s approach, it seems that they do seem very keen (too keen) to get their advertisers to spend more. Recommended bid adjustments seem to be up more often than down, and many of their ‘advertiser tools’ seem to be designed to generate more money for Google, not the advertisers.

On the other hand, Google are doing some things for their paid search advertisers, that perhaps aren’t appreciated as much as they might be.

Consider Google’s instant results for example. Whilst some paid search advertisers think this is a bad thing, look at the search results with and without it. With the instant results, there are more paid search adverts above the fold, and fewer natural search results. Whilst my colleagues in the SEO team here at Epiphany may view this as a bad thing, from a PPC perspective, it’s certainly not hurting us!

After all, something that Google and their advertisers both want is more users clicking on paid search results. Whilst Google mitigate this by their need to retain and even improve the integrity of their search results, certainly Google and the advertisers should be on the same team…

Users Need Google. Um – No, Not Really

Why do people use Google? Why is their market share so large? Is it because their search results are much better than their competitors?

The truth is that they aren’t really. There are plenty of tools out there to compare the search results, and a number of studies have concluded that Google’s search results aren’t any more relevant or useful than those of Bing or Yahoo.

So why do people use Google? Perhaps it’s simply a force of habit. I always use Google because I’m used to it. I know what the search results are going to look like, and it works as well as anything else.

To ‘Google’ is synonymous with to search. It’s a part of everyday language.

But so is to ‘Hoover’ and I use a Dyson. It certainly helps to have a well-known name, but it’s not enough by itself.

The truth is that it probably wouldn’t take a very large push to make a lot of Google users switch to another search engine. They just need a good reason to. Look at what happened when MSN became Bing – everyone tried it out to see whether it was any better. As far as I could see, the results were about the same, so I went back to Google – along with most other people, I think.

So Google’s market position isn’t as solid as it seems – they only get advertisers because they have market share. They only get users because they haven’t got a good reason to switch search engine.

At the moment, it seems that the users have the power – Google and the advertisers need them far more than they need Google. The advertisers will always follow the users, so without them, Google has nothing.

I don’t think Google’s in any danger of losing its market share any time soon, but putting my soothsayers hat on for a moment, I think it’s just a matter of time before somebody invents a better search engine, whether it’s better results, a better layout, or it simply becomes more fashionable. Ten years from now, Google won’t be the force that it is now. Until then, perhaps the only way that they can keep everyone happy is to keep the users happy, and ignore the advertisers.

Where their standard bid optimization is completely automated, this option allows you to set your bid, and they’ll adjust it based on the conversion rate. The tool will increase your bids up to a maximum of 30% above your bids, though there does not appear to be a minimum.

The bids are adjusted in order to maximise the number of conversions that you achieve for your current CPA. If this is done correctly, then assuming that nothing changes in terms of the performance of your account due to external factors, then theoretically the number of conversions achieved at the same CPA should improve.

Which is all very nice, but I do have a few concerns…

1. If the tool works, you’ll spend a lot more money. Is this made obvious to the advertiser? It’s inevitable that if you get more conversions at the same CPA, then you’ll spend more.
2. Since this tool will spend more of your money, what happens if you’re already spending your full budget? Since your cost can’t increase, and your CPA won’t change, you should see no difference in your conversions.
3. What happens if your market changes? Seasonality can change your conversion rate significantly, as can a number of other market-specific factors, and circumstances within your individual business (such as stock levels). If your conversion rate drops, you’ll get fewer conversions (obviously), but Google will then kick you while you’re down, lowering your bids to hit your target CPA, and reducing your conversions further. Ouch!
4. Perhaps the most critical question of all, and the one that I ask whenever Google rolls out a new automated tool, ‘WHY?’ Why would you use this tool? Think about it. How long does it take to adjust your bids based on their performance, and how often does it need doing? If you spend a huge amount of cash, you may need to adjust your bids every day or two – but in that case, you should be either doing it manually, or using an expensive, cutting edge tool (depending on the complexity of the account). If you aren’t spending that much cash, you should only really be adjusting your bids occasionally – maybe every week or two. Otherwise, you’ll just be responding to random variation. How long does this really take? An hour? Two? Doesn’t the amount that you’re spending on Adwords justify that much of your time?

The truth is that assuming that you can crunch the numbers yourself, you will perform better than an Adwords automated tool. You can choose how far back to look when assessing the performance (balancing the volume of data with its recency), and if certain Ad Groups are performing better or worse, you can decide whether to increase the bids on the good groups, reduce the bids on the bad groups, or both. You can choose whether to spend more, or to use the improved performance to reduce your spend whilst retaining the same level of conversions, or to spend the same and pick up more conversions for your money.

Seriously, I don’t see who this tool is aimed at. Clearly, agencies shouldn’t touch it – after all, they are being paid to manage the campaign as well as possible, not with as little effort as possible. Perhaps it’s aimed at small companies who want to advertise on Adwords, but how are they even going to find out about it? It’s an option hidden away in the account settings, so unless they are subscribers to the Adwords blog, they will probably never know about it.

I don’t believe that this is just another ‘Google Money Grabber’. Assuming that it works in the way that it says that it does, it will certainly increase spend, but the user will be getting extra conversions for their money. I won’t be trying it on any of the campaigns that I manage – our clients pay for better service than that – but it’ll be interesting to see what sort of results the people that do try it get…

Generally, these periods are good for businesses, but it can be the case that the performance doesn’t quite match the increase in traffic. Too often, businesses simply label the visitors as lower-quality traffic, and adjust their bids accordingly, but this can be totally wrong.

Suppose that a business has a significant lead time between a visitor hitting their website for the first time, and the time of the sale – more than a week, perhaps. When traffic volumes are stable, so are the conversions, but what happens when you add seasonality to the mix?

Here’s some (obviously) fictional traffic data to a website that sees a spike in its clicks in the middle of summer:

Imagine first that when somebody makes a purchase, they always do so on the same day. For the sake of an easy-to-read graph, we’ll say they have an unlikely 20% conversion rate:

So far, no problems. The conversions increase at the same time as the clicks, and the conversion rate remains constant, at 20%.

Now consider a business where people take more time to consider the alternatives before converting.

What happens to the above example if you still have a 20% conversion rate, but 5% convert one week after the initial visit, 10% after two weeks, and 5% after three weeks?

Suddenly, the conversion rate is all over the place, dropping without any obvious reason in the early spring, then increasing through until late autumn. What’s caused this?

Quite simply, if you see an increase in your clicks, but the conversions don’t come until later, your conversion rate will fall. Put another way, your conversions are still being driven from past clicks (which were lower).

In the above case, I’ve used a fairly improbable sinusoidal curve for sales. What happens if sales are usually flat, but there’s a sudden, sharp increase?

I’ll use the conversion-delay distribution from before, with the following click distribution:

The variation in conversion rate caused by this seasonality is huge:

The conversion rate drops from 20% to 15% immediately, the moment the clicks start to increase. It starts to recover, but the moment the clicks start to fall, the conversion rate goes through the roof.

The Catch

Looking at the above graph, it’s fairly clear what’s happened. But understanding what’s happened in real time from looking at the data isn’t all that easy. Look at the performance data from the above graph up to week 21:

What are you going to conclude at this point? A lot of analysts, even quite experienced ones, would assume that the additional traffic to the site is of lower quality, and will cut the bids.

Of course, this is completely the wrong thing to do – the clicks aren’t worth any less.

Similarly, when the conversion rate increases, there’s a temptation to increase the bids:

Again, this is the wrong move – the clicks are worth the same in reality, so the bids shouldn’t change.

The Bottom Line

Individually, seasonality and delays between clicks and conversions don’t really do much harm. Combining them, you see some eye-opening variations to your conversion rate, which can lead you to make some very bad decisions.

This is also a problem if you decide to significantly change your budget and bids. Suddenly driving more or less traffic to your site can make a big change to your conversion rate. Have you ever seen that your conversion rate drops when you increase the bids, or climbs when you cut back? It may not be a reflection of the position of your adverts, but a natural result of the lead time on your conversions.

As long as you know that this is a possibility, you should be able to respond in a more measured way, but how many PPC Analysts look at the delay between the click and the conversion on a campaign? I would strongly urge anyone with seasonal demand or long delays between first click and conversion to check this out before their next peak or dip in business.

Once upon a Monday dreary, while I entered a search query
Into Google, haven of quaint volumes of forgotten lore
While I nodded, nearly napping, I considered words, which tapping
Brought forth ads attention trapping, trapping my attention’s core
‘Tis a trickster,’ this I muttered, ‘trapping my attention’s core
Only this, and nothing more.’

Ah, distinctly I remember, it was July or December,
When the advert’s dying ember printed on my memory store
Eagerly, I clicked to follow, seeking to receive tomorrow
Copies of ‘The Mask Of Zorro’, a movie that I adore,
For the rare and radiant maiden, Catherine – she whom I adore
Ever since she first I saw

As I passed onto the webpage, suddenly I felt a great rage,
Anger’s dagger thrust in me, through my heart it brutally tore
No delivery tomorrow, I discovered to my sorrow,
The parcel would later follow, follow to my home’s front door
‘What of claims you made,’ I cried out, ‘to deliver to my door
The next day, if not before?’

Presently, my soul grew stronger, turning lassitude to anger,
‘Sir,’ said I, ‘Or madam. Your advert gave me rancour.
I shall not this film be buying, your false promise left me sighing
Even a part of me was dying, all for Cath whom I adore!
She who starred in America’s Sweethearts, raven beauty I adore
Now I shall see her no more!’

Deep into my dark soul peering, I thought of adverts, that endearing
Led consumers gladly cheering, to that company’s distant shore
Beating Adwords could be easy, to write adverts that, though cheesy
Make promises to tease me, tease me to their website store?
This I decided to trial then, increase clicks to my own store
With untrue claims, base and poor.

Back to my computer turning, all my soul within me burning,
Again I started writing adverts, with more vigour than before
‘Surely,’ said I, ‘this will work well, though visitors think it hell,
Nonetheless with it I shall sell, this ad shall be my cat’s paw
Tricking customers with untruths, this ad shall be my cat’s paw
Empty wind, and nothing more’

As my adverts became alive, I doubted not that they would thrive,
My cost per sale would start to dive, dive below its previous floor,
Not the least doubt pierced my soul yet, my expectations would be met,
I detected not the least threat, my campaign would start to soar.
For my advert was compelling – hence my sales would no doubt soar
But my sales remained quite poor.

As I watched with mounting fear, my ad began to disappear
Falling down the SERPs descending, past competitors it tore.
Turning to my bust of Larry, on my knees I begged ‘don’t tarry –
Tell me why this pain I carry – tell me now, I do implore!’
Vainly waiting for an answer to the question I implore
Quoth Larry ‘Quality Score’.

Much I marvelled this ungainly statue to hear discourse plainly
Though its answer little meaning – little relevancy bore
Nothing further then he uttered, even though my mind was cluttered
Even yet I quietly muttered, ‘Larry, your thoughts I’d explore –
What is this thing that you speak of – something that I must explore.
What is this Quality Score?’

Larry spoke, unlike a raven – if one did enter my haven
He would not offer advice far beyond a raucous caw.
‘If you want your adverts high, there is one thing left to try –
Keep your relevancy high – lest your ads remain as poor’
You who sought to game the system, have been penalised as poor
Thus repair your Quality Score’

‘Wretch’, cried Larry, ‘Don’t be evil, lest you turn into a devil
Adwords must be used for good, so put the product at the core
Don’t let greed make you dishonest, deliver all that you have promised
Or your adverts will be dismissed – this has been your fatal flaw
Make good on all claims that you make, don’t let greed become your flaw’
This he said, and nothing more

‘Thanks Larry’ I said, then turning to my PC, once more burning
With the need to do no evil, wrote ads whose motive was pure
So it was I typed with relish, telling truths I’d not embellished,
Lest my ads fell to that hellish place they’d seen the day before
Clicks I gained thus were not worthless, worthless like they’d been before
Thus I saw conversions soar

The bust of Larry, never flitting, still is sitting, still is sitting
On my desk – it still reminds me of the lessons learned before
So it is I sit here writing, cautioning against inviting
new techniques that sound exciting but harm your Quality Score
Relevancy is the answer; don’t risk your Quality Score
I will do so Never More.

Sometimes a large part of the SEO strategy will revolve around generating traffic through an extensive list of very specific keywords, and sometimes the strategy will be focused around a few critical, large volume keywords.

The same can be true for PPC. Large volume keywords can cost a fortune to advertise on, putting them out of reach for many smaller advertisers, but for some advertisers, they are critical to the success or failure of the campaign.

‘The Hammer And Scalpel’ is one approach to ensuring that you get the most from your advertising budget – The Hammer is SEO, and the Scalpel is PPC.

The principal is simple – you use SEO to target a number of very large keywords (that are out of reach for your PPC budget), whilst you pick up the long tail keywords with the PPC.

Clearly, this won’t work if high rankings are impossible to achieve on the biggest keywords, perhaps if you have a new website, or the incumbent top adverts are immovable advertising colossuses (colossi?!).

But if your website has some history, then page one rankings for these keywords may certainly be achievable. Whilst you focus your SEO efforts on these keywords, PPC can be used to pick up any potential traffic on lower volume keywords.

There are a number of advantages to this approach:

• You will be represented on all of the relevant keywords, unlike an SEO-only strategy.
• You won’t have to worry about your adverts cannibalising traffic from each other.
• Your advertising spend is likely to be far less than a PPC-only strategy, targeting the same keywords.
• You’ll start getting a return on your investment far more quickly than from an SEO-only strategy.
• Your advertising spend may start to fall in the long term, as you achieve top results in SEO.
• Once you have top results in SEO, you can target new keywords, and remove these from the PPC campaign if you want to.
• You will still get a ‘halo effect’ on other keywords in SEO, as a result of improving your website’s visibility in Google on critical terms. These can, again, be removed from the PPC if you want.

Advertisers with deep pockets, or in less competitive or smaller markets, may still be better targeting everything in PPC – this solution certainly isn’t for everyone.

But if you’ve found in the past that PPC is too expensive, then a combined PPC and SEO strategy may be the perfect solution for your needs.

The performance of the Content Network went haywire a few days ago. This isn’t that unusual, as people who use the Content Network will testify. A big website suddenly has a page about your product or service, and your figures can go all over the place.

But as I always do, I looked into it, to see if any action was needed. Indeed, as expected, the spike in traffic was all coming from one website. But it wasn’t a newspaper website or indeed any well-known site. It was a website that I’d never heard of.

I ran a placement report, and discovered that prior to 4 days ago, I’d never had a click from this website. In the last four days, they’d burned through £700 of my client’s money, without a single conversion to show for it.

Remarkably, the clickthrough rate from this website was a staggering 6%, a figure almost unheard of on the Content Network.

I decided to visit this website, and I wasn’t all that impressed. There was one page of text, stuffed with keywords about a single service. There were no less than 10 adverts on this page, along with some ‘learn more’ links that all went through to a free prize draw website.

They had a ‘Contact Us’ page with no contact details, a form, and another 12 adverts, and the Terms and Conditions page (a strange thing to have, since they offered no products or services) was similarly wallpapered with adverts.

So far, they hadn’t really broken any rules or laws, but where were all these clicks coming from? They had no natural or paid search results – even for their own name. And why was the clickthrough rate so high?

It all points towards them clicking on their own adverts, and pocketing the cash. The website in question was about student loans, so it’s likely that they would have no problems at all using different IP addresses, bypassing Google’s detection.

I have of course passed it on to Google’s click fraud team to investigate, and I’m sure they’ll be shut down soon. But there’s nothing to stop anyone else pulling a trick like this – and unless you check your placement reports when strange things happen on the Content Network, you could easily become a victim.

It’s likely that Google will reimburse my client, but will they reimburse every other advertiser on this website? I have no idea, but I would say that to be on the safe side, you definitely want to be the one that catches this, rather than relying on somebody else to spot it.

Conversion Funnels are now available on the majority of accounts, and I’ve decided to revisit my prognostications, in light of the data that we are seeing on our accounts at Epiphany.

The basic principle of conversion funnels is that many users visit your website multiple times before committing to make a purchase, and that reporting conversions based solely on the final click can skew the effectiveness of different keywords.

Keywords that people use when choosing a product or company are frequently different to ones that they use once they are committed to a conversion.

An example of this would be somebody purchasing a toaster, who searches for generic keywords when deciding which toaster to buy, then on more specific terms once they know what specific product they want. Finally, once they have compared a few prices, they search for the website that had the product they wanted at the lowest price.

According to a final-click model, the conversion would be tracked against the brand name, so the value of the previous keywords would be ignored.

Google’s Conversion Funnels don’t cover all touchpoints, only Adwords clicks, but even so, analysing this data could paint a very different picture of the campaign, and could lead to significant changes in bids.

One assumption here is that a large percentage of Googlers visit a website multiple times before converting – clearly, if the vast majority convert on the first visit, then any changes made from understanding their journeys would be small.

This led to the first piece of analysis that I did, establishing how many people converting from an Adwords click had made previous visits to the website through Adwords.

The results were quite surprising – in the majority of cases (14 out of the 20 accounts that I reviewed), over 80% of the conversions were on the first visit via a paid search.

Where less than 1/5 of the conversions can be reassigned to different keywords, the impact is inevitably small – in fact, there were very few cases where the appropriate changes were large enough to be worthwhile.

As I said in my initial study, the keywords whose value is most likely to be overstated by a last-click model are variations on your brand name. And since these are likely to be cheap conversions anyway, you are unlikely to want to reduce your bids on them.

In fact, this was the problem that I discovered – my analysis told me that I’d slightly underestimated the value of clicks from everything except my brand names, but since in many cases we were working to a restrictive budget, there was no benefit to increasing the bids.

Where you are likely to get more valuable results are the accounts where you can clearly identify which are your more generic keywords and which are more specific, perhaps based on your Ad group structure.

I mentioned that in most of the accounts, the first click was the one that generated the conversion, and it was interesting to see which accounts bucked this trend.

With one or two very surprising exceptions, it was retailers of products with high prices (such as furniture) or businesses offering B2B services. In both cases, it’s not surprising that searchers will want to compare a number of different websites before converting.

The business-to-business example was an interesting one, as in neither case did the conversion form any kind of a contract – it was simply a request for information. Perhaps this is an interesting learning then, that people ‘shop around’ even when they are simply looking for information about a company.

Based on my initial work, I don’t think that Conversion Funnels are going to make a major difference to the PPC landscape. The data is very difficult to work with and analyse, requiring extensive manipulation in Excel before you can draw any kind of meaningful conclusions from it, and in the majority of cases, it simply isn’t worthwhile.

If I can offer one piece of advice to anyone thinking about analysing this data on their accounts, it’s this: Check what percentage of conversions are coming through on the first Adwords-driven visit. If it’s more than 75%, then it’s highly unlikely that you are going to make any major changes to your account as a result of the analysis that you are performing.

There is an alternative to performing a complete analysis of their conversion paths, using a report within the Conversion Funnel suite. Google summarise how many last-click conversions each keyword or Ad group generated, and how many ‘assists’ those keywords or Ad groups got. The ratio of these is a reasonable guide to which bids would be increased or reduced as a result of the analysis – a kind of Click Attribution Lite.