Google Ads AI Max Campaign Experiments

Key Takeaways

It hasn't been very long since Google released AI Max for search campaigns. And depending on your account, you might have only recently got access to it. If you want an overview about AIAX for search, you can check out the video at the top of the screen right now. But today, I want to talk about a way that you can test it in a little lower stakes than just rolling it out to a campaign. Google's doing a bit better job of making it easier for advertisers to test their new features. And for AI Max for search, they've created a campaign experiment so you can test and see how the functionality would work for your existing search campaign. So in this video, I want to walk you through how to set up a campaign experiment for AI Max for search and see if it might work well in your account. Just like any other campaign experiment in Google Ads, we need to go over to the main menu on the left and under the campaigns heading, we need to go to experiments. Once we're here, depending on which view you want to use, you just need to click the blue plus button to create a new campaign experiment. And then if you have the option to test AIAX, this is what the first page is going to look like. If you do not have campaign experiments for AIAX, the original selection is going to look like this where you see the specific types of experiments. But as you can remember in the other view, we now have a question of what do you want to test? whether it's features and settings, assets, campaign types, or custom. For this option, we want to use campaign features and settings. And here you can see there's lots of different options of what you can test, but we're going to choose AIAX for search campaigns. Then click continue. To test AIAX in an experiment, you have to select a search campaign that you want to test if it would work better in AIAX. And it needs to be a campaign that is not currently in an experiment. But also, this campaign has to be active. I've already tried to put this video together with a paused campaign since we don't actually want to run any campaigns in our placeholder account, but it wouldn't let me set up the campaign experiment. So, right now, we have a live campaign. It's only got a dollar a day budget and no keywords, so shouldn't actually do anything. But once you have the list open of all of your available campaigns, you just need to choose the one that you want to test AIAX for. And then you will quickly be brought to the final page of setting up a campaign experiment for AIAX. This view looks a little bit different than the other views that we have in the account because we're not setting up lots of different custom pieces. There's not a lot of customization. We really just need to get it set up and turned on. So, if we scroll down here, you can see that we have the control arm in this light gray box that is around here. That's going to be the original campaign that we have. If you decide that you want to change it, you can click the pencil button right here, choose a different campaign. You'll see that the budget is set here. Traffic split is set to 50%. There is no customization of the traffic. You can't change that. And this campaign has AIAX off, which is what it currently is set at, and that makes the most sense. In this next box down here below, you can see we have the treatment arm or the second half of our AIAX search campaign test. It has all the same information plus a little more. campaign is the original campaign. Budget is going to be the shared budget that we have selected up above. So that $1 would technically be split across both of these. And the traffic split still 50%. Then AIAX is set to on. Down here below, you can see this little box that they always have open for you right away, which is great. It shows that text customization and final URL expansion are on. Now, if you're paying attention to the news and you're familiar with how AIAX works in most search campaigns, or if you watched our other video that has the better overview of it, you'll know that typically you get to turn these on and off for an AIAX search campaign. But in this campaign experiment, they are on and they are both on. You do not have the option to opt out of them. They are just there. So, let's just do a quick recap of what text customization and final URL expansion are. Text customization is a setting that allows Google to help generate additional assets or headlines and descriptions that can be used in combination with the assets that you provide in your ad copy. So basically, Google is going to help write some variants of your headlines and descriptions for you based on other assets that they have found with your responsive search ads. Now, what are those other assets? These are going to include your ads unique context, which is an interesting way of framing your ad. It includes your domain, the landing page, existing ads in the same region of your account, and keywords in that ad group. It also says that AI can help generate additional assets if you're using English language assets in your campaigns. So effectively, Google is going to be able to take copy from your other ads, your landing pages, some of your different ad assets, and use those to come up with additional headline and description variants. Now, URL expansion is an option that lets you give Google the ability to send users to additional final URLs on your website if it thinks that it might match better to your query and might have a better chance of seeing performance from the user who conducted the search. Effectively, Google can change your landing page if it thinks there's a better page available on your site. Now, typically, you can turn this on or off and leave text customization separate. But if you want to use final URL expansion, you cannot have text customization turned off in a regular AIAX for search campaign. Typically, you do still have the option to have exclusions and inclusions for URLs here, but that's getting into a little bit of a different video. So, let's hop back into our experiment. And now if we keep scrolling down in the campaign setup, you can see here we have the experiment dates. Right now it's going to default to the start date. The end date is going to be October 13th. If I want to try to extend this, I can only extend it to November 10th. I'm not going to do the math, but I'm going to take a guess that that's about 84 days. And that's the longest that Google Ads will let you run a campaign experiment on any different feature. They usually start you more in like the mid50s range. So I just added a little bit more. Now, it's probably running for about 84 days, but as you can see in there, it'll only show how far out you can select, and everything else will be grayed out. As Google says over here on the side, for best results, you probably want this to run anywhere from 4 to 6 weeks, and that's probably right. 1 week is not really enough for an experiment. 2 weeks could be a fluke. Personally, I almost always set every campaign experiment to be the full 84 days. And then if I want to turn it off early, I can do that. but it at least gives me the option to have it run for as long as possible and gather as much data as possible before making a decision. Now, speaking of making a decision, this is something I find interesting. There's a checkbox down here that says, "Apply experiment changes to the original campaign if the treatment arm performs best." Now, there's a whole lot going into this tiny little checkbox. First of all, if we set up another campaign experiment someplace else, this is just a custom experiment, but you can see that we've got the length, we've got the experiment split, we have the enable sync, which is different than applying the changes because you'll notice that there is no option to apply the changes based on what performs best. That's something that typically when you're running a campaign experiment and you tell it to end, you would then select whether or not you want those changes to apply or not. This is something that happens at the end of an experiment, not at the beginning. Second, the very beginning of an experiment, especially a custom experiment, is where you choose your metrics that you want. Clicks, impressions, cost, conversion, conversion value, and then you decide if you want them to increase, decrease, or have no significant change. This tells Google what you're trying to expect out of this experiment and why you're running it. But on this new page, there isn't any of that here. None of that shows up. So, this begs the question, what does it mean if it performs better? you haven't told Google what your options are. Depending on the bid strategy you're using, maybe you've chosen target CPA or target rorowaz and Google could maybe infer that one has a better return on ad spend, but you're optimizing for max conversions. Is Google going to consider the one that spends the budget every day or the one that has the better conversions the better performer? Or for rorowaz, is it going to think that it's better if it has a higher rorowaz or is it going to choose a floor for you and think this is good enough? It's still performing the best. Personally, this checkbox, first of all, being automatically checked, but just in general, is terrifying to me. There's no way for me to tell Google what I think it means to perform better. And it's not telling me what it thinks it's going to perform better. I'd almost guarantee you if you opt into that, 99% of the time, it's going to say AIAX is going to perform better. And you know what? It might, but I don't want Google making that choice for me. So, categorically, if you're testing these, please uncheck this box. and then come back multiple times throughout your experiment, but also at the end of your experiment and evaluate for yourself what performs best. Now, lastly, you can give your campaign a name. And this is what's going to show up in your campaigns manager. So, make this something that fits in with your current naming convention, something you'll be able to see, pull it out really easily. And then the last thing is you just come over to schedule, click that, and now it's set up and running, and it's on its way. Again, if you were to ever click in here, you can just easily apply the changes that we have. That would be just fine. Or you can click into the campaign experiment itself. And this is where you're going to see the performance for the control arm. So, your original campaign without AIAX and the new augmented campaign with AIAX turned on. This is the page where I would like you to review performance, decide which performs best, and then if you want to easily apply the experiment, just click that button. Or if you get halfway through your experiment and everything looks terrible and you don't want to use AIAX, you can easily do what I'm going to do and click end experiment. You got to double check it, but yes, end it. Now it's turned off. And I'm going to go turn off our placeholder campaign cuz again, we don't want to run in the first place. But overall, that's how you set up an experiment for AIAX for search in Google Ads. It's super easy, but we lose a lot of customization. And unless you're smart and pay attention to that little checkbox, you lose the ability to determine what is a successful test and what's not. So hopefully this rundown has given you a little more confidence in how to set up this campaign experiment, but also what to look out for. But if you have any additional questions about this setup process or anything else with AIAX for search, leave us a note in the comments below. Thanks for watching our video. We really appreciate it. If you liked it, give us a thumbs up below. If you really liked it, maybe think about subscribing to the Paid Media Pros YouTube channel and you'll get alerted every time a new video drops. If you really really liked it, you can help support the channel by checking out some of the t-shirts that we're wearing on our merch shelf, as well as looking at the super thanks button.