-
Ai Max sucks for tech
Posted by on March 26, 2026 at 8:48 pmI experimented with Ai Max on a few campaigns for three quarters, and the results were disappointing. While the top of the funnel generated inflated leads, the deeper analysis revealed very poor results.
Despite running these campaigns with a robust negative keyword list, I was surprised by the lack of seriousness in the intent matching process. We were matching with tools and software that were not aligned with our offerings. Broad keywords like “AI” were used. Additionally, I limited Ai Max to only intent matching and didn’t to include the landing page variable.
This performance was compared to our previous results, which had accepted leads and a pipeline. However, this experiment was too challenging to maintain, and it was overall ineffective for B2B tech or at least for me. The weekly negative keyword list tasks were cumbersome because eliminating one broad theme led to the emergence of another.
I am curious to hear how others have fared with Ai Max, as I suspect some individuals in the e-commerce industry may have achieved success with it.
-
1 Reply
-
Yep.
AI Max maximizes the broadness of your targeting. Using negative keywords with it is like trying to nail slime to the wall.
Unless you sell everything related to {word_on_your_website} half or better of your clicks will be irrelevant/low intent.
-
I think it can work just like P-Max can work but you need a pretty high volume of conversions to help drive clicks from the right prospects. And of course lead quality will always be lower when people are coming in from upper funnel terms.
-
The quality issue you’re describing is real and it compounds downstream in a way agencies often don’t flag to clients.
Even when AI Max surfaces a genuinely good lead — someone who did match your ICP — there’s a second problem waiting: how fast the client’s team actually responds. We’ve run controlled tests across B2B and service businesses and the average response time is 2+ hours, often much longer.
The lead quality debate becomes moot if the follow-up window closes before anyone picks up the phone. Harvard data puts first-response within 5 mins at 100x higher connect rate vs. an hour later.
So for agencies managing these clients: fixing targeting is one lever, but pairing it with a tight speed-to-lead process is the other half. Otherwise you’re optimising the top of the funnel while the bottom leaks.
-
Ai max sucks so far. Period. Try again in a year or so 🤷
-
this is a pretty common outcome for B2B with AI Max — the intent threshold just isn’t tight enough for considered purchases.
but there’s a downstream issue that doesn’t get discussed enough: when broad campaigns generate inflated lead volume, the sales team’s response pattern changes. they start triaging by submission time instead of intent signals, which means the rare genuinely qualified lead gets the same 4-hour callback as the irrelevant one.
the negative keyword grind you mentioned is real. but the conversion gap often lives *after* the click — not in the campaign settings. better matching helps, but so does having a signal for which of those inflated leads is actually worth calling first.
-
Same, i just turned it off, only shitty search terms. Whatever good search term we got ended up being like our phrase match.
-
If you have adopted broad match and pmax it’s not doing much except potentially giving you slop AI copy.
-
For the people thought the “AI” along with the “MAX” was going to work well I have a bridge to sell you
-
AI Max = DSA 2.0
DSA = weak and almost depricated
Log in to reply.