Yes, I'm talking about Snap again. People accuse me of loving to talk about Snap. They're right.

When I am right about things, I tend to be right about things, so pay attention.

Here's a thing I keep hearing about in planning meetings, usually delivered with the confidence of someone who stopped learning in 2019: "Snap's interesting, but it's not scaled enough for our media mix." WHAAATT?

I smile when I read it now. Because that sentence tells me everything about how they're going to lose money this quarter.

Let me be direct: I'm not here to sell you on Snap. I don’t have any reason to sell anything, but my newsletter, and occasionally, stuff my kids need you to buy so they can get the free scooter.

I'm here to tell you that Snap is already doing something to you. (and not in a weird way), whether you buy media there or not. It's grading your entire stack. It's the test you didn't know you were taking.

Everyone in this industry loves to talk about measurement. MMM. MTA. Incrementality. They say the words like incantations, like the syllables themselves will ward off the CFO's questions. But here's what nobody wants to admit out loud: most of your measurement architecture is built to confirm decisions you already made. It's not a map. It's a mirror.

Snap is the thing that breaks the mirror.

The industry treats Snap like clearance inventory. Lower CPMs. Younger demo. Nice-to-have when there's budget left over. This framing is so ubiquitous that it feels like gravity—just the way things are.

It's also completely backwards.

Snap is the only scaled environment where you can observe the full post-impression P&L in a single place. Not reach. Not frequency. Not the comforting abstraction of "impressions served to a targetable audience." The actual economics of what happens after someone sees your ad:

Attention — not viewability theater where "50% of pixels visible for one second" counts as a win. Actual eyes-on-screen duration. The thing you thought you were buying everywhere else.

Session depth — what they did next. And next. And next. The behavioral exhaust that tells you whether your ad was a speed bump or a spark.

Incrementality — the hardest word in advertising, the one that makes attribution vendors sweat through their quarter-zips. Snap lets you measure it. Cleanly. Repeatedly.

Marginal ROAS — the curve nobody wants to look at because it tells you exactly where your money started dying. Not the blended average that flatters your deck. The actual slope.

That's not a media channel. That's a laboratory.

And the experiment it's running has results most brands aren't ready to read.

Here's a concept I want you to sit with: post-impression P&L.

Not what you paid to show someone an ad. What happened to your money after the impression fired. Did attention accrue? Did behavior change? Did the conversion you're claiming actually trace back to a moment of genuine engagement, or did your attribution model just credit the last thing that touched the pixel?

Most media planning stops at the invoice. Post-impression P&L starts there.

And Snap—almost by accident of its architecture—is the cleanest live implementation of this we have. Full-screen by default. Sound-on by default. A user behavior pattern that's immersive rather than interstitial. When someone sees your ad on Snap, they saw your ad. That sentence is not true everywhere you're spending money, and you know it.

This is what I mean when I say Snap is your control group. It's the environment where the inputs are cleanest, the measurement is tightest, and the signal actually means something. When you run the same creative on Snap and then compare it to the murky impression fog of other platforms, you're not comparing channels.

You're exposing what the other channels have been hiding.

If you have a smaller-share channel—let's call it, hypothetically, a platform where everyone posts disappearing photos—that consistently produces higher ROAS and lower CPA than your budget anchors... that's not a curiosity. That's not a "test and learn" footnote for slide 47.

That's a measurement indictment.

It means the channels eating 70% of your budget aren't outperforming. They're just older. More familiar. More deeply wired into your org chart, your agency's staffing model, and fifteen years of muscle memory that nobody wants to unlearn.

The performance gap isn't about Snap being cheap. The gap is telling you that everything else might be overpriced.

Let that sit for a second.

Your "scale" platforms aren't delivering more value. They're delivering more volume at lower quality, and your measurement stack is structured to make that look acceptable. Snap is the control that exposes the grade inflation everywhere else.

So here's the cliffhanger, and I'll be blunt about it:

If Snap is the control group—and the data increasingly suggests it is—then the performance gap you're seeing elsewhere isn't inefficiency. It's not "optimization opportunity." It's not something your agency is going to A/B test their way out of next quarter.

It's a tax. A misallocation tax. And you're paying it every single quarter to platforms that have been grading their own homework for a decade.

The question isn't whether Snap deserves a bigger share of your budget. The question is whether you're ready to look at what the control group is telling you about everything else.

Most people aren't. Looking is expensive. It means admitting the model is wrong. It means uncomfortable conversations with partners who've built their business on the old math.

But the brands that look first? They're going to eat the ones who don't.

This is where the free ride ends.

In Part II, available exclusively to ADOTAT+ members, we go deeper:

→ The Attentive-Second Economy — Why the shift from impressions to attention-weighted buying changes everything about channel mix, and how Snap accidentally built the infrastructure for it before anyone else was paying attention.

→ The CFO Conversation You're Not Having — How to translate post-impression P&L into language that survives a board meeting. Including the specific metrics that make finance teams stop asking about reach.

→ The Reallocation Math — What happens when you actually run the experiment. Real numbers from brands that moved budget and what the control group revealed about their legacy allocations.

→ Why Your Agency Doesn't Want You to Read This — The structural reasons the people managing your money have negative incentive to surface these findings. And what to do about it.

The control group doesn't lie. It just tells you things you weren't ready to hear.

Because "you're about to be wrong" isn't a threat. It's a window. And windows close.