Google has said Performance Max is the future. Full stop. That's the direction the algorithm is being built around, the direction account managers are being trained around, and the direction spend is being pushed.
Now search "PMax leads" on Reddit or Quora.
The threads are a graveyard. Wasted spend. Junk conversions. Forms filled out by bots. Account managers shrugging. Lead businesses turning PMax off and going back to Search.
Most people read those threads and conclude PMax doesn't work for leads. I read those threads and concluded something different: if everyone's getting turned off it, competition must be lower. The auction must be softer. The cheap bids must be on the other side.
I was right. But there's a catch.
Why most people's PMax fails
PMax is an algorithm. Algorithms learn from data. If the data you feed PMax is bad, the algorithm optimises against bad data and finds you more bad leads. Quickly, and expensively.
For lead businesses, "bad data" usually means one of three things:
- Junk conversions. You're tracking form fills or WhatsApp clicks rather than actual qualified leads. PMax happily optimises for whoever clicks WhatsApp. That includes your competitors, recruiters, and bots.
- No value signal. Every conversion looks the same to the algorithm. A tyre kicker is worth the same as a $50k customer. PMax has no way to bid up for the latter.
- Consent-broken conversions. Your conversions technically fire but the consent flags aren't passed through. Google's docs are quiet about it, but those conversions can be used for measurement and cannot be used for audience building. So PMax learns nothing about who your best customers look like.
Fix none of these, and PMax burns your money. Fix all three, and PMax becomes the best channel you have.
What "clean" data actually means
For PMax to work on leads, the conversion data needs to be three things at once.
Clean. The thing you're sending Google as a conversion needs to be a real, qualified lead. Not a click. Not a form fill. Not a WhatsApp tap. A response, a conversation, ideally an enquiry that you'd recognise as a lead in your CRM.
Consented. The conversion is fired with the consent flags Google needs in order to use it for audience building, not just measurement. This is a tiny technical detail that breaks 90% of the lead-business PMax setups I've looked at.
Value-weighted. Each conversion carries a value that reflects how much that lead is actually worth to you. Could be a flat heuristic (enquiry = $200, booked call = $1,000, closed = $20,000). Could be back-fed from your CRM. Either way, PMax now has a gradient to climb, not just a binary signal.
That's the recipe. Clean. Consented. Value-weighted. Three knobs, all of which most people get wrong, all of which Tagless is built to get right by default.
Why this is the opportunity right now
There's a brief window in any new ad channel where the people who can make it work pay much less than they should. Search had that window in 2005. Facebook had it around 2014. PMax for leads is having it now, and the reason it's having it is that the technical bar to do it properly is high enough that most lead businesses give up and go back to Search.
Search is fine. Search will keep working. But Search is also where everyone else is, which means the auction is dense and the CPCs reflect that.
The cheap bids are facing the other way while everyone's facing one direction.
That's the whole opportunity in one line. Everyone is crowded into Search because PMax burns them on junk. The people who solve the conversion definition (clean, consented, value-weighted) get to bid on PMax with very little competition, on Google's stated future of digital advertising.
All you need to claim it is conversion data PMax can actually learn from. That's what Tagless gives you, by default, in 2 minutes of setup.