Ashwin Balakrishnan: 3 tests you can run to optimize RSA performance

Transcript: 3 tests you can run to optimize RSA performance

[auto-generated transcript]

My name is Ashwin I'm from Optimyzr and welcome to my PPC zone presentation on how to build and test winning responsive search ads. I'm really excited to talk today about RSAs, why you should be paying attention to them. As well as sharing some tests and some data that we pulled from Optimyzrs product data as.

So let's get into it. So response to search ads of course, are the new default. When you create a search campaign in Google, this is the option that you're gonna start with. They have the same minimum requirements as expanded text ads, which you can no longer create three headlines to descriptions and you can have up to three ads, three RSA per ad group. 

The difference of course with RSA is that you can now go up to 15 headlines and four descriptions, and then Google will mix those up in different combinations to find the best ad for a particular search query.

Now, one of the reasons that Google is moving towards more automated solutions like responsive search ads is so that anybody can open up Google ads and start advertising quickly. In this process though, admirable some of the inputs and recommendations from Google are not necessarily the best and ad strength is one that you can disregard right off the bat.

So in this example, that nav Hopkin shared with us some time ago, you can see that the quality of the ads did not really say anything, but Google's only feedback is that you're not using enough keywords. So the detection of quality is simply not present ad strength while a good guideline is not something that you should aim to improve.

Rather use your own judgment when determining the quality of. What you should instead focus on are structure and creative. And along with first party data, these two are probably gonna be what really allows you to get the most out of a more Google automated version of Google ads. So with ETA's account structure was pretty straightforward.

You'd have the campaign, multiple ad groups within the campaign and each ad group targeting a group of keywords, and then you'd have multiple expanded text ads within those ad groups. Each of those competing with one another to see which ad can serve most frequently and get the best performance with response to search ads up until the ad group level is largely the same principle.

But when you get to the ad level, RSAs don't just compete against other ads in the ad group. They compete against versions of themselves. So RSA one, depending on the number of headlines can have a certain number of combinations that present. So not only is it compete against the other two RSAs in your ad group, it's also competing against all the other combinations that Google thinks it might be a good fit for.

So in order to find your best RSAs and build the best structure you can. There are three tests that we recommend running. The first of these is pinning. So another feature of RSAs is the ability to tell Google that certain headlines and description should here in specific positions. So you can pin one field to a specific position.

You can say headline one should always show in the first position. You can then pin multiple fields to the same position. You can say your first three headline. You can pin all of those to first position and then Google will make sure that any one of those always serves as headline. Number one. And the third thing you can do is pin all of your elements that keep in mind.

If you're pinning fully, you can only show three headlines in two descriptions, no matter what. So the first three headlines and the first two descriptions that you pin will be the only ones that Google shows as part of your ad. A Google says that pinning is going to reduce your reach and is not recommended.

Yes. And no, our data shows that even though unpinned Ads have decent performance, when you pin some elements, though, the CTR takes a small dip. The conversion rate does increase considerably and for fully pinned ads both CTR and conversion rate do go up and our interpretation straightforward.

When advertisers have more control over how their ads show, the results tend to be better, which isn't surprising. The next test, probably the most important as a copywriter, I would definitely prioritize this above everything else is your messaging. So when you have.  when you had ETA's, it was easy to go in and kind of build a campaign on the slide, with RSAs it becomes much more important that you plan your structure out beforehand. So you know what you're trying to achieve. You wanna have good structure from top down, which means that you have a clear campaign goal, and then you build things out. So there's no overlap between keywords and ad groups, but you also wanna make sure that it makes sense from the bottom up, that you're coming back to that same goal and that your campaigns aren't cannibalizing each other or eating into each other's traffic. So messaging is, is one of the things that you can use to continually test the structure of your RSA campaigns. You can test calls to action, you can test value propositions, you can test variations in messaging.

And if you're gonna be changing things across campaigns, then you can use ad variations in Google ads to make that happen pretty quickly and easily.  We found from our data, that sweet spot is one to two RSAs per ad group. This is where you do get a good increase your performance. It does come at a small cost increase, but when you get to three RSAs, you don't really see the increase in performance that the cost increase justifies.

And of course, the final test that you wanna run is maximizing your impressions. When it comes to ETA's, it was pretty straightforward. If the keyword didn't work, it didn't work. You had to try something new, but an RSA with buried copy, as long as you're not fully pinning or, or making it too restrictive on Google especially paired with broadcast.

This allows Google to then create new ads on the fly, qualify you for more options and really be able to show your ad to more people. And that's really what RSA are about, is maximizing the number of impressions that you can potentially get. If you're gonna run this experiment, we highly recommend having more text more text clearly means more impressions.

And if, if that's what you're trying to see, then we, we recommend maxing out on your text. So ultimately, why does this all matter? The bottom line is Google likes RSA. We don't know why, whether it's because of internal product decision or because RSAs tend to lend to better qualities score, but they do get more than double the impressions of ETAs.

But it's not just Google, even customers before RSAs. We see in our data that the RSA ad groups tend to get more than 50% more conversions than the ones, which is ETA's. And as you implement some of these changes and tests, we highly recommend that you change the way you look at success with an expanded text, that it was pretty straightforward.

You look at conversion rate, you look at which ad performed better, ultimately, and that was your winner with RSAs because of their nature, the increase in impress. Frequently offsets the slight dip that you'll see in conversion rate. So monitor not only the number of conversions that you get, but also look at the total value of those conversions.

Measure them up against your target CPR, target ROAS. If you have one or your total ad spend, you may discover that even though your conversion rate dips, you're ultimately running more profitable ads. I hope you have a lot of luck with the RSAs they're the only ads that we can create for search campaigns going forward.

So definitely important piece of Google Ads that is here to stay. Thank you so much for your time and thanks Jyll for having me!