Most of Your Revenue Has Nothing to Do With Your Advertising

I sat on a call a few weeks ago where a marketing team walked through their quarterly performance, vendor by vendor. Paid search had a number. Social had a number. The affiliate platform, the retargeting partner, the CTV vendor — everyone came with a figure and a slide. By the time they got to the end of the deck, the vendors had collectively claimed more revenue than the company had actually done that quarter.

Nobody in the room said anything about it. That's the part that stays with me. Not that the numbers didn't add up — they never do — but that everyone had quietly made peace with it. The reports were treated as roughly true, each one accurate in its own frame, and the contradiction was filed under "attribution is messy."

It isn't messy. It's answering a different question than the one being asked, and the gap between those two questions is where most bad budget decisions get made.

Here's the uncomfortable place any honest measurement conversation has to begin: most of your revenue would have happened anyway.

Turn off paid search tomorrow. Turn off social, CTV, display, every campaign currently running. Your revenue does not go to zero. For an established brand with existing customers, organic traffic, stores, an email list, word of mouth, and years of accumulated equity, the large majority of sales still happen. That floor is your baseline, and if you don't know roughly where it sits, every conversation about channel performance is starting from the wrong number.

I'd go further: the baseline is the most important figure in your measurement stack and it is almost never on the dashboard. Platforms have no commercial reason to show it to you. Showing you the baseline is showing you how much of "your" contribution was never yours.

What every platform report is genuinely good at is identifying revenue connected to advertising activity. Someone saw an ad, or clicked one, and later bought something. That connection is real and the platform is not lying about it. But connected and caused are different words, and the entire discipline lives in that gap.

A customer searches your brand name, clicks the paid ad at the top, and buys. Paid search reports the sale. Something else convinced that person — a friend, a previous purchase, a CTV spot they saw three weeks ago — and they were arriving with intent already formed. The ad was a toll booth on a road they were already driving down. Another customer sees a retargeting ad on Tuesday and purchases Wednesday; retargeting reports the sale, though that person had already been on the site and already knew the brand. Each of these reports can be factually accurate about the interaction while being completely wrong about the impact.

The question I hear most often on client calls is "what did this channel drive." I think it's the wrong question, and I think asking it is the single most expensive habit in marketing measurement. It invites you to find every customer who touched advertising and assign their revenue somewhere. The better question — the only one worth planning against — is what changed because this channel existed. Those sound similar. They produce opposite answers often enough that a team can spend a year optimizing confidently in the wrong direction.

When teams first see a large baseline, the reflex is to treat it as a failure — revenue the system couldn't attribute, a gap to be closed with better tracking. That reflex is wrong and it's worth correcting hard.

The baseline is real economic activity with real causes. People buy because they already know you, because they've bought before, because someone recommended you, because they need the thing you sell and they typed your URL directly into the browser. They're responding to brand building that happened long before the measurement window you're staring at. That revenue isn't missing. It's just not yours to claim this quarter.

None of that makes advertising irrelevant. It makes incrementality the only thing worth measuring. Advertising doesn't need credit for business that was already coming. It needs to prove it changed the outcome.

Once the baseline is set, the channels sitting closest to the transaction tend to get smaller. They have a structural advantage in conventional reporting: they're positioned exactly where existing demand becomes visible. They see the customer at the finish line, and standing at the finish line is not the same as having caused the race.

Channels operating earlier look weaker for the mirror-image reason. Their effect surfaces elsewhere — as branded search volume, as a direct visit four days later, as improved efficiency in a lower-funnel channel that then claims the sale. Measure each channel inside its own reporting environment and those relationships vanish entirely. You end up optimizing claims rather than outcomes, and the channels best at claiming get funded.

Ask your analytics team one question: how much of our business would we expect to happen without our current paid media?

Do not accept "we can't know that." Nobody observes a counterfactual directly — that's precisely why it gets modeled rather than tracked. Get an estimate, argue about it, refine it. A rough baseline you can defend beats a precise attribution number you can't.

Then put every paid channel above that line and make each one explain change. Not touches. Not conversions associated with an impression. Not revenue that happened to occur after a click. Change.

Budget conversations get much shorter after that. The question stops being which platform reported the best return, and becomes which investments actually moved the business past what it was already going to do.

Never give advertising credit for revenue that would have happened without it.

Talk soon,
Jeff Greenfield
CEO Provalytics

Jeff is the co-founder and CEO of Provalytics. He'll mention it occasionally. He'll also tell you when it's not the right tool.