19 Jan The End of Expanded Text Ads
It’s that time of year again. The time of year when you get to anticipate the gift that keeps on giving. Unfortunately, I’m not talking about the holiday season that’s right around the corner. I’m talking about the time of year when Google decides they need to make a change to their ad platform. The biggest difference between the holiday season and when Google decides to make changes is that one happens at the same time every year, where the other springs up out of nowhere. It is the gift that you don’t always want to keep on giving.
Recently, Google announced that starting next July, they will be eliminating the Expanded Text Ad option for search campaigns in Google Ads. The only ad type that will be available for search will be their responsive search ads. What’s the difference between the two? Expanded text ads are comprised of three headlines, two descriptions, two paths, and a click-thru URL. Responsive search ads are comprised of up to fifteen headlines, four descriptions, two display paths, and a click-thru URL. An even bigger difference is that responsive ad components are randomly combined by Google. More is better, right? Wrong, or at least not always.
Part of the reason expanded text ads are a better option, at least in this humble SEM specialist’s opinion, is that they offer more control and better data insight. They may have a max of only three headlines and two descriptions, but because of that, it’s easier to tell what part of an ad may or may not be performing. Advertisers also know that headlines one and two will always have a higher chance of appearing in search results so the most prominent information can go there, and then get a good sense of which headlines are successful with the data collected.
Currently, responsive search ads do not break down performance by individual headlines or combinations, and instead, only show the impression share of headline combinations as determined by Google’s machine learning. The data reported on responsive search ads will tell users that a headline or description may not be performing as well as others, but that’s about it. Currently, the system does not show specifically which combinations drove conversions or other significant actions, but just the impression share they accumulated. We may see that an ad had ten conversions this month, and a combination of headlines four and seven had the most impression share, but that does not tell us how many of those conversions that particular combination, if any, accounted for.
One way to try and have control after the switch will be to utilize the ability to pin or fix certain headlines in certain locations to rank your ad combinations. For example, some advertisers may have restrictions as to what can be said and where it can be said, so one may have to fix a headline to a specific position in the responsive ad setup. The problem with this is that Google does not recommend pinning headlines to certain locations and not only do you have to compete with other advertisers, you also are limiting your opportunity within Google’s ad platform. For example, ad A has two pinned headlines while a competitor in ad B has none. Google is most likely going to favor and rank the ad without restrictions because of its flexibility. So, it’s no longer just competing against the content of your competitors’ ads but the setup in Google. It becomes a two-front war!
The hope is that Google will update the reporting for responsive search ads and give more insight as to what works and how it works beyond impression share of combinations. On top of the current lack of insight on performance, there is also an increased workload when it comes to responsive ads versus expanded text ads. Normally, an ad group should launch with two expanded text ads with some headline variation for A/B testing purposes. These ads would call for about six total headlines and four descriptions, but the workload has now more than doubled with the need for fifteen unique headlines as part of a responsive ad. Again, an ad does not have to have all fifteen, but for the most effective ranking possibilities, it should. Remember, ads with more variations will have a better chance of ranking against ones that do not have as many varieties.
At the end of the day, this seems like a better way for Google to make more money, which is frustrating as an advertiser. Their experts say that the updates in their machine learning and AI used in their automation is designed to account for a growing number of search terms they have never seen before. Unfortunately, what it means for advertisers is less control over where and how their money is spent in digital ad campaigns. There’s a time and place for automation, but it is frustrating to have it forced on you. I’m not sure if SEM specialists were on the naughty list this year, but I must imagine this feels like getting a lump of coal this holiday season.