# Putting Google Ads AI Max’s automated ad copy to the test > Source: > Published: 2026-07-28 14:00:00+00:00 [PPC](https://searchengineland.com/library/ppc) » # Putting Google Ads AI Max’s automated ad copy to the test ## See where AI-generated assets helped ecommerce, B2B, and B2C campaigns — and where human control still delivered better results. One of [AI Max](https://searchengineland.com/ai-max-for-search-everything-you-need-to-know-462923)’s capabilities is creating assets for you. This can take the strain off the PPC team by reducing the need to customize ads for every single ad group. We wanted to quantify how well the text customization feature worked for various companies, so we ran tests with three different companies to understand how much we should or shouldn’t be using text customization. As we’re a software company that helps companies manage their PPC accounts, not the agency doing the work, we worked with the companies to guide them on how to set up and analyze the tests. We’ll start by examining the process we used to help the companies set up the tests so you can run this experiment yourself, and then we’ll examine the test results. ## Text customization and messaging restrictions Before you run this test, you need to understand the features involved. First, you need to turn on AI Max and the text customization feature. Since this is AI working in the background for you, it can tailor the assets in every single ad group to the keywords in that group. While this sounds nice, the assets can sometimes be used to run promotions or advertise products and services you don’t offer. To help guide the system, you should apply messaging restrictions when using auto-created assets. Messaging restrictions can help guide the system on what the ads should and shouldn’t say. In addition, you can give it rules around your brand guidelines. The overall process for creating good messaging restrictions is fairly simple and only takes an hour or two: - Use a prompt in Gemini to create your initial assets. - Then use a prompt to make the ads overly promotional, create promises you don’t like, etc. Essentially, you’re having the system write ads you don’t approve of and want to ensure Google doesn’t create assets like these on your behalf. - Create messaging restrictions to stop the assets you don’t like from being generated. - Use a prompt to create new, highly promotional ads with your messaging restrictions until all created assets fit your company’s messaging guidelines. *Dig deeper: Is your account ready for Google AI Max? A pre-test checklist* [ See exactly how your competitors win. Analyze your competitors ](https://www.semrush.com/analytics/traffic/paid-search?utm_campaign=ic_sel_0101ppc&utm_source=searchengineland.com&utm_medium=overlay&onboarding=off) Uncover the keywords, ads, landing pages, and strategies driving your competitors’ paid search success—and find your next opportunity to outperform them. ## Choosing the campaigns to test We wanted to see how these assets performed across different types of companies and different levels of campaign optimization. Therefore, we first chose three business types: - Ecommerce. - B2B lead gen. - B2C lead gen. In each account, we only wanted campaigns that met a specific set of criteria: - Did not use brand keywords. - Spent at least $20,000 per month. - Had at least 100 ad groups. Most companies have some campaigns that are their best performers, where the team spends a lot of time optimizing the campaigns. Then you have your campaigns that are long-tail, do well in aggregate, but receive less attention. To understand how well text customization was going to perform, we used two campaigns that were highly watched and two that were somewhat neglected from each account. We also wanted to see how the new assets performed, so we chose campaigns that didn’t rely heavily on pinning, which excluded many of the highest-spending campaigns, since most large and enterprise companies extensively use pinning in their top campaigns. Finally, we wanted to test only the assets, not URL expansion, so none of the campaigns used the final URL expansion feature. ## Asset review As you run these tests, you’ll want to monitor the auto-created assets and remove any that don’t align with your brand messaging or offers. When looking for AI-generated assets, it’s essential that you change the default filters to include the ad, as this filter isn’t chosen by default. The companies in the test monitored these assets as they were created and removed them before they received many impressions. Ignoring the B2B results (more on that later), approximately 19% of the auto-created assets were removed. ## The results ### Ecommerce campaigns This ecommerce company sells over 100,000 SKUs, so it has a lot of products, and many users are accustomed to visiting the website and searching again if the landing page doesn’t have the specific product they’re looking for. At first glance, it appeared that both AI Max and text customization were incredibly successful. However, after further analysis, we found that AI Max was poaching impressions, clicks, and conversions from other campaigns and that the overall revenue for the account declined. The company added many search terms as keywords to help Google prioritize the correct ad group and campaign. Then it added more negative keywords and audience lists to slow cannibalization and reran the tests. We observed that AI text customization wasn’t as effective as human management of assets for highly optimized campaigns. However, it was good at assisting the long-tail campaign. *Dig deeper: Why your brand campaign may not be ready for AI Max* ### B2B lead generation When creating [RSA assets](https://searchengineland.com/how-to-assemble-captivating-google-ads-copy-459469) for B2B companies, one of the most important considerations is how to properly prequalify your audience through your asset usage. You want your ads to be unattractive to B2C searchers and appeal to B2B searchers. This company had previously [used pinning](https://searchengineland.com/how-to-assemble-captivating-google-ads-copy-459469) quite extensively across its assets to ensure this qualification. However, it wanted to see how well Google could optimize its accounts, so it removed its pins during this test. The nuance of prequalifying a B2B audience is one that text customization clearly doesn’t understand. The B2B accounts saw their CTRs skyrocket. However, the conversion rates declined significantly since the ads were attracting many B2C searchers. The messaging restrictions included language to ensure the assets prequalified users as part of a B2B audience. While some of the assets met this criterion, the overall ads that were shown to users didn’t properly appeal to B2B buyers. The other tests ran for over a month. However, after three weeks, the results were so poor that this company stopped the tests and went back to pinning its ads and removing the auto-created assets. Within a week, its results returned to their pretest levels. ### B2C lead generation Our last test was with a B2C lead generation company that localizes its ads through geographic ad copy or geographic insertion. Its optimized campaigns had tailored ad copy for the keywords in every single ad group. Its long-tail campaign had a few headline assets in each ad group, tailored to the keywords, but most assets were reused across ad groups. Seeing formulaic ad copy in lower-priority campaigns is quite common, and this is where we were hoping to see AI auto-created assets perform well, since there was a lot of opportunity for better ads. AI Max auto-created assets didn’t disappoint in these low-priority campaigns. While these assets didn’t outperform the assets humans had spent a lot of time testing in their top campaigns, AI performed quite well for the long-tail campaign. [ Every click they win is a customer you lose. See who’s stealing your traffic ](https://www.semrush.com/analytics/traffic/paid-search?utm_campaign=ic_sel_0102ppc&utm_source=searchengineland.com&utm_medium=overlay&onboarding=off) See where competitors are investing, which keywords drive their results, and how to capture more of the market. ## Where AI Max automated assets work best AI is a fantastic tool at your disposal. For ads where you’re spending a lot of time thinking through your messaging, it can be good at helping you generate ideas, but human-created assets still outperform AI-generated assets. If you need specific types of assets, such as prequalifying users for B2B audiences, specific offers, or short-term promotions, then you should take control of the assets yourself and not turn them over to AI. However, auto-created assets shine when you just don’t have enough time to fully optimize your creatives. Using AI to create your assets, assuming you have good messaging restrictions and regularly review these assets, can help your overall performance. We’re still a long way from AI being a turn-on-and-forget setting. It needs babysitting and oversight. However, having AI do the heavy lifting in areas where you don’t have the time to fully optimize, and then spending your time reviewing and tweaking the outcomes, is the best use of AI in ad creation. *Contributing authors are invited to create content for Search Engine Land and are chosen for their expertise and contribution to the search community. Our contributors work under the oversight of the editorial staff and contributions are checked for quality and relevance to our readers. Search Engine Land is owned by Semrush. Contributor was not asked to make any direct or indirect mentions of Semrush. The opinions they express are their own.*