Back to all questions
Answered Google Ads Strategy

AI Max is enabled on one of my campaigns and I am fairly sure I did not enable it

M
Marcus Hale Asked Jul 30, 2026
3 25
Checked settings this morning and one of our four search campaigns has AI Max switched on. I didn't turn it on and neither did my colleague. It may have been on for weeks - I have no way to tell from where I'm sitting. Two things I want to establish before I decide anything. First, how do I find out when it was enabled and by whom? If a rep did it during an optimisation call then I would like that on record. Second, what does it change about how I should be reading the last month of reporting? If a chunk of traffic arrived without a keyword then my keyword-level analysis has been describing a subset of the campaign and I didn't know it. Not looking for opinions on whether AI Max is good - happy to form my own. Looking for how to audit what has already happened.

Replies (3)

Accepted answer
K Keyword Ninja Team · Jul 30, 2026
Both are answerable. When and by whom: Google Ads change history records the setting change with a date and the account it came from, and it will distinguish between someone in your account and a change applied by Google. That's the authoritative record and it's worth screenshotting if the answer turns out to be a rep, because these get applied through auto-apply recommendations more often than people realise. What it changes about your reporting - this is the more consequential half, and your instinct is right. Queries on an AI Max campaign can arrive with no keyword attached. Any report grouped by keyword silently omits those, so your keyword-level view of that campaign has been describing part of it. The search terms are the complete picture; the keyword list isn't. To size it: look at the campaign's search terms and compare the ones with a keyword against the ones without. That tells you how much traffic was outside your keyword list and what it cost. One caution on that exercise. Search term reporting only ever covers the portion of spend Google is willing to show - on a typical account that is roughly half of campaign cost, with the rest below the anonymisation threshold. So calculate the AI Max share against your observable search term cost, not against total campaign cost. Divide by campaign cost and every account on earth looks like a disaster, because you are measuring Google's privacy threshold rather than AI Max. One more thing now that you know it's on: negatives need to sit at campaign level on these campaigns. Ad group level doesn't hold, because the query did not come through an ad group in the first place.
J Jenna Krause · Jul 31, 2026
The auto-apply angle is worth everyone checking regardless of whether they have noticed anything (settings, recommendations, auto-apply - it's not where you would look). We found three things enabled on a client account last year that nobody in the room had agreed to. None of them catastrophic. All of them decisions somebody else made about how we spend money.
G Greg Pallister · Jul 31, 2026
"How do I audit what already happened" is exactly the right first question and almost nobody asks it - they go straight to whether the feature is good... The share-of-observable-cost point above is the one I would write on a card. I have already seen someone divide keywordless spend by total campaign cost, conclude that two thirds of their budget had been taken over, and take that number to a client. It was wrong and it was wrong in the direction that makes you look hysterical.

Want to join the conversation?

Log in to post