Google Ads is testing something new, and Google hasn't said a word about it.
A beta page called "Asset activity" has started showing up in some advertiser accounts, tucked into the Assets menu between the Assets report and Videos. It delivers weekly AI-generated insight cards about your ad headlines: what's working, what search trends are shifting, how your creative stacks up. It was first spotted Nov. 6 and reported Nov. 9. My read: this is Google quietly extending the same AI-scoring layer it's been building into Performance Max for over a year, one UI section at a time.
What Is the Google Ads Asset Activity Page?#
It's a beta section in the Google Ads console, labeled "Beta," that refreshes weekly with cards sorted into categories such as:
- Featured
- Asset performance
- Asset diversity
- Industry trends
- Competitor performance (seen in at least one account, not all)
Each card carries a headline, a date range, and a "View details" link.
What Actually Happened#
PPC consultant Anthony Higman posted a screenshot on LinkedIn on Nov. 6 showing the page live in his account. Search Engine Journal and Search Engine Roundtable both wrote it up on Nov. 9, and both found the same gap: no help-center article, no entry on Google's own Ads announcements page, nothing from Google confirming the feature exists at all.
One card in Higman's account flagged a rising click-through rate on his top-rated headlines. Another, in a legal-services account, suggested adding "financial recovery" language to attract injured-rider searches, based on a week-over-week industry trend read. A second practitioner, Hana Kobzová, spotted that fifth category, Competitor performance, which didn't show up in Higman's screenshot. That mismatch alone suggests the test isn't rolling out the same way to every account.
Why This Matters, and Who Can Skip It#
If you don't run Google Ads, this changes nothing for you, and you can stop reading right here.
If you do run it and haven't seen this page, that's normal. It's beta, it's inconsistent account to account, and Google hasn't published any eligibility criteria. If you do see it, treat it as a prompt to go check your asset report, not as a verdict.
The cards appear to lean on the same click-through-rate, cost-per-click, and ROAS numbers Google already labels "directional indicators only," because Performance Max groups judge assets together, not one by one. A card telling you a headline is "top-rated" is working off the same imprecise data Google has used for years, just with a friendlier face on it now.
The Honest Trade-Offs#
The upside is real. A weekly nudge toward testing a new headline beats nobody looking at asset performance for months, which is most accounts most of the time. The downside: Google still hasn't explained how the cards get generated, and the in-product disclaimer admits the AI "could give inaccurate information."
We don't know the sample size behind a "top-performing headlines" card, or whether it's reading your account alone or benchmarking against a wider pool. Until that's documented, every card is a hypothesis, not a finding. Google has walked this exact path before, quietly testing AI-generated ad imagery and new PMax brand controls as unannounced betas months ahead of any formal confirmation.
How We're Treating It in Our Own Creative Process#
At WebEpex we write the ad and landing-page copy that feeds campaigns for clients across GCC, European, and US accounts, and this is exactly the layer Google is trying to auto-score. We reviewed Higman's and Kobzová's screenshots the day they circulated and added one line to our creative-testing checklist: any AI-suggested headline, from Google's cards or anywhere else, gets logged and tested at the asset-group level before it touches a client's budget.
It never gets promoted off a card's word alone. That's the same rule we already run against Meta's own AI-generated ad defaults, which shifted again with Meta's API v27 earlier this month. We got burned once trusting a platform's own "this is working" label over our own split-test numbers. We haven't repeated it.
We're not switching on every beta panel a platform ships us. We are keeping a running list of which AI insight layers actually earn their keep across Google and Meta, because whichever one gets this right first tends to shape what the other builds next.
What I'd Tell a Client Asking About This#
Check your Assets menu this week. It costs nothing to look. If the page is there, read the cards, but don't act on them yet. Pull the real numbers for any headline a card calls out and compare it against your other creative in the same asset group, over the same window. If it holds up on your own data, test it properly. If you don't see the page at all, you're not missing anything; most accounts don't have it yet.
Close#
If you want a second pair of eyes on what your Google Ads account is actually telling you versus what an AI card says it's telling you, send me a screenshot and I'll give you a straight read. Two minutes, no pitch: cal.com/webepex/growth-review
