How advertisers should respond to fewer ads showing on Google

nito / Shutterstock.com

After a quiet start to the year with few notable changes in AdWords, the last month has seen a lot of interesting things happening, most notably the change to the layout of ads on the search results pages.

I’ve shared my initial thoughts about this change on my company blog; but now that it’s been live for a few weeks, I wanted to weigh in with ways to optimize accounts that are impacted, including using bid to position rules and Quality Score optimization.

What changed to ads on Google

The week of February 19, Google began removing right-hand side (RHS) ads, reserving that space instead for product listing ads (PLAs) and the Knowledge Panel. This change reduced the maximum number of text ads on the page from 11 to seven.

A change as significant as this one caused some angst amongst advertisers, but the impact seems neutral to positive, according to several pieces of research: iProspect noted that RHS ads had a terrible click-through rate (CTR) anyway; Merkle|RKG analyzed the stats two weeks after the change and found no major impact to the big picture; and others reacted similarly. Google’s own Matt Lawson even said that “[i]n aggregate, this change is neutral for small advertisers.”

Google’s search results pages no longer show text ads on the right side, and one new ad is now shown above the organic results.

I’ve run a few estimates of my own (which I’ll share below) to gauge the impact of the change, and I also think this is a net positive change because it opens up a very large number of new clicks to advertisers.

Regardless of all the studies concluding that the aggregate or average change will be neutral or positive, remember that individual results may vary. None of us are averages. We’re all individuals, so we need to pay close attention to how we’re individually impacted by the change.

This being election season, let me put it like this: If one of the presidential candidates proposed changes to taxation that would reduce taxes, I wouldn’t vote for that candidate without first checking that my individual taxes would also go down — the aggregate of all taxes has no impact on my own bank balance.

The change makes 18% more clicks available

This explanation comes from my earlier blog post, but in case you missed it, I’m including it with some more detail here because this goes to the heart of why I believe this is a positive change.

Let’s consider two facts before we estimate what impact this change may have.

1. In 2011, when I was on the team at Google, we came to the conclusion that “on average, this [bottom ad] placement performed better than side ads in terms of click-through rate in our tests.”
2. As has been widely covered as part of this change, ads shown at the top of the page have a significantly better click-through rate (CTR) than the same ads on the RHS.

iProspect says that the CTR of the top ads is 14 times that of RHS ads, so if the CTR of a RHS ad is the unknown x, then a top ad’s CTR would be 14x.

• RHS ad CTR: x
• Top ad CTR: 14*x  (top ads perform 14 times better than RHS ads)

Now, let’s run some estimates to help determine what the change may mean to advertisers. To simplify the example, we’ll assume there used to always be three ads at the top and 8 on the RHS, and there will now always be four at the top and three at the bottom. Obviously this isn’t the case, but it’s good enough to let us compare some possible numbers before and after the change.

Before the change, advertisers were competing for the clicks from three top ads and 8 RHS ads. This is how many clicks there used to be available for every ad impression:

• 3*14x (for the three top ads) + 8*x (for the eight RHS ads) = 50x clicks (the total clicks of a page)

Remember, x is the CTR of an ad on the right. Let’s say it’s 2%, so that means that for every search, Google expects one click between all of the 11 ads on the page.

If Google just removed the RHS ads, they’d have reduced possible clicks by 16%:

• -8x (for the 8 RHS ads that are gone)/50x (for the total clicks of the page before the change) = -16%
  

But remember they are adding one new ad at the top and that one may also get 14 times the clicks of a RHS ad, so here’s how many clicks we can expect from four ads at the top:

• 4*14x (for the 4 top ads) = 56x clicks

That’s a 12% improvement from before — but we’re not done, because a few of the old RHS ads now move to the bottom, where we expect them to get a better CTR than before. But let’s just assume their CTR stays exactly the same in their new position:

• 4*14x (for the 4 top ads) + 3x (for the 3 bottom ads) = 59x clicks

So with this change, Google may have very well gone from 50x clicks to 59x clicks for every query, an increase of 18%.

• (59x – 50x)/50x = +18%
  

Again, remember that x is the unknown CTR of a RHS ad. If that was 2%, then for every search, Google could now expect 1.18 clicks between all of the seven ads on the page.

What happens when we add in what we’ve seen so far

Mark Ballard and Andy Taylor from RKG reported that top ad CTRs have declined around five percent because there is one additional ad to click on, whereas bottom ads have improved CTR by 60 percent because there are fewer to choose from.

Top ads CTRs have declined five percent by adding one more ad. Bottom ads have increased CTRs by 60 percent by showing fewer ads in a better location below the search results.

If we add these numbers to my estimates, here’s the potential for new clicks that are now available:

• 4*14x*(1–5%) (for the 4 top ads) + 3x*(1+60%) (for the 3 bottom ads) = 59x clicks
• 53.2x + 4.8x = 58x

This still leaves us with a 16% gain in possible clicks:

• (58x – 50x)/50x = +16%

Are there really more clicks available now?

While I argue that the change could make 16 percent more clicks available, the reality is that we probably won’t see this increase right away because Google still controls how many ads show above the results, and they may fine-tune that threshold to keep the auctions relatively steady.

As time goes on, Google may play with this threshold, and we may eventually see the full impact of all the newly available clicks.

How to get the Ad Rank needed to be Top 4 or Top 7

Returning to my point that averages and aggregate numbers don’t matter, what can individual advertisers do to improve their results if they are not happy with the CPCs or Ad Ranks since the change?

Our available levers to change Ad Rank have not changed, so we can either change the bid (which is quick and easy) or try to change our Quality Score (harder to do).

We can also improve our ad extensions like Sitelinks, which are now more important because all ad positions have the space to show all extensions (something that wasn’t the case for RHS ads that had less horizontal screen-estate to play with).

Solution 1: Bid to position

The easiest way to make your ads eligible to appear above the search results is to use a bid-to-position methodology to aim for position 4 or better.

Google has a Flexible Bid Strategy called “target search page location,” where we can choose between aiming for the top of the page or the first page of results. Google will automatically change the bids as needed to try and achieve these positions.

Use Google’s “Target Search Page Location” Flexible Bid Strategy to get your ad to show at the top of the page or on the first page of results.

Alternatively, advertisers can use bid rule engines they write themselves to steer their ads toward the desired average target position.

When using systems like these (like Automated Rules), remember to take into account that the average position shown in AdWords usually includes data from search partners, and for broad match keywords, data from all the various queries.

Try to filter that data down to a narrowly defined set of data points — for example, exact match on Google Search only. If you don’t define it this narrowly, then any change in bids will change the query mix, and that can skew the averages dramatically, making it impossible to compare apples to apples.

When using Google’s Flexible Bid Strategy, there is an option to exclude any keywords with a Quality Score (QS) of 3 or less. This is because bidding more for low QS keywords can be cost-prohibitive, and these keywords would be better served with a QS optimization first.

Solution 2: Improve Quality Score

Whether you’re looking to improve QS because your keywords have a low QS or because you’d rather spend time than money to improve Ad Rank, the good news is that Google’s reports now include the three sub-components of QS, making it easier to identify the best way to improve QS: ad text relevance, expected CTR or landing page experience.

Here’s one way to utilize the new data from Google:

1. Identify keywords that have a “below average” score for ad text or landing page.
2. Check if these keywords are in an ad group where other keywords are “average” or “above average” for the same sub-components of QS.
3. If YES, then move these keywords to new ad groups and try a new ad or landing page based on the currently best-performing variant in the ad group for the selected keyword.
4. If NO (and all keywords in the ad group are “below average”), then rewrite the ad in the ad group or change the landing pages, since all keywords may benefit equally.

In theory, this is pretty straightforward; in practice, this can take some time. AdWords doesn’t have a report that shows both the keyword-level details like QS and the ad text performance segmented by keyword. You can download an ad text report segmented by keyword, but this won’t include QS details.

At Optmyzr, we’ve written a freebie AdWords Script that generates the needed report that merges all the data about keywords and ads. You can download it here. If you need a primer on how AdWords Scripts work and how to add this one to your account, there was an excellent series by Ginny Marvin on this site about a year ago.

Here’s an example of the output of my script:

The AdWords Script generates a Google Sheet with both ad text level data and keyword level attributes like the three components of Quality Score.

Now it’s easy to look for “below average” ad text relevance among keywords in ad groups that have some “average” or better ad text relevance. From there, you can take the segmented CTR data from the existing ads to create a new ad group that eliminates the worst CTR ads and test a few new ad texts.

In the example above, notice that one of the ad texts has a better CTR; that should be the starting point for new experiments where we try to boost our ad text relevance.

Conclusion

While the reduction of ads on the Google search results pages may seem like it’s making fewer clicks available, I believe this change will significantly increase the number of clicks available.

Now it’s up to advertisers to duke it out for who gets their piece of the bigger pie. With some straightforward changes in account management strategies related to bid management and Quality Score optimization, advertisers should find themselves well positioned to benefit from this change.