July 30, 2026 | Vol 118
One point of attribution is not a strategy.
Comcast Advertising and Affinity Solutions just cut a deal that ties TV commercials to actual credit card purchases. The setup combines 30 million Comcast households with transaction data from over 100 million consumers, so brands can link on-screen ad exposures to real world purchases.
The headline is that TV is now a performance channel.
It landed now because the pressure got unbearable. Proving ROI is the top barrier to increasing brand investment, and only 17% of US senior decision makers say they feel confident defending linear TV ROI. NIQ reports 74% of CMOs face increased pressure to prove it.
The money side is worse. More than two thirds of senior marketing executives estimate they waste at least 11% of their media budgets because sales tracking signals are too slow or missing entirely. Eighty percent admit they routinely optimize live campaigns without verified purchase data.
Define It
The mechanism is a data clean room. Comcast uploads viewing logs, Affinity uploads merchant transaction feeds, personal information gets stripped and scrambled into random tokens, an algorithm matches households that saw a spot and then bought something inside a set window.
The output is aggregate, not individual. Something like 12,000 exposed households generated $45,000 in revenue.
Three things this unlocks for advertisers: mid-flight optimization instead of waiting for the post buy, third party attribution that counters platform self-grading, and targeting households by actual purchase history.
That last one is the part your competitor is going to like.
The Solution
Ask yourself a better question.
Would you rather have one clean point of attribution, or every part of your marketing pointing the same direction and touching the buyer at every stage of the journey?
That is not a rhetorical question. It is a budget decision, and most operators get it backwards.
KPIs matter. The bottom line has to move or none of this counts. But declaring that a piece of your advertising is your best performer or your worst performer because it happened to be the last touch is a recipe for disaster. Last touch does not measure contribution. It measures proximity to the transaction.
Digital has been running that con on itself for years. Retargeting is where platform-reported returns are most inflated, because those customers were already in market and would have converted at a high base rate whether the ad showed up or not. The platform claims the credit anyway. Now the same logic is being fitted to television with a cleaner interface and a bigger dataset.
The fix is not a better single number. It is alignment.
Point everything the same direction. Same offer, same message, same promise across TV, digital, search, social, and whatever your people say when the phone rings. Consistency across touches is what compounds. A buyer who sees your spot, then finds your review, then hits a landing page that matches, converts because of the sequence, not because of one link in it.
Never let one test become policy. Markets shift. A single result is a data point.
The goal was never to find out which ad gets the credit. It is to build a system where the buyer keeps running into you until buying from you feels obvious.
Straight Talk From The Crazy
I like the technology. I do not like the premise it is being sold on.
This is single party attribution wearing a lab coat.
A clean room can prove two events happened. It cannot prove one caused the other. If a customer sees your spot but walks into your store because they live two blocks away or ran out of something, the algorithm credits the TV ad anyway.
The only 100% proof that a commercial made a sale is the customer telling you it did. Everything else is correlation with a nice interface.
So here is the trap. If single party attribution is not actually necessary to run a smart buy, why are you paying a premium for it? And if it is necessary, this does not deliver it. Pick one.
Add sample bias skewing toward high income spenders and you are optimizing toward the customers you were already going to get.
The privacy question is not settled, it is deferred.
This runs on credit card transaction data. The FTC has already warned against corporate privacy washing, and weak clean room constraints can let bad actors combine transaction timestamps and zip codes with outside registries to reverse engineer tokens and unmask households.
Now look at who supposedly consented. Pew Research found 56% of Americans click agree without reading, 67% understand little to nothing about how their data is used, and 73% feel they have lost all control of it.
That is not consent. That is a signature on a form nobody read.
The bottom line. This is a useful new tool in the measurement stack. It is not the answer to TV attribution, and anyone selling it to you as the answer is selling you certainty that does not exist yet.
Use it as one signal. Never as the verdict.
Let's talk
If you are getting pitched on attribution right now and the math is not adding up, hit reply and tell me what they are claiming. I will tell you what I would ask them.
No pitch. No pressure. I just like talking shop.
Until next time - Expect to Win!
JIM
#xp2win
Source: EMARKETER, "Comcast and Affinity deal makes TV a trackable performance channel," June 9, 2026 (budget waste and optimization data originally from Affinity Solutions 2026; ROI confidence from EMARKETER Jan 2026; CMO pressure from NIQ; consumer data from Pew Research Center)
