Effective CPC via brand-suppression holdouts

Nobody has ever turned your brand campaign off. Not for a day, not for one region, not even to see what would happen. It is the cheapest line in the account, it converts, and it renews every month without an argument.

Here is the catch that makes it worth measuring. Some of those clicks would have arrived free, through the organic listing sitting directly below the ad, and the invoice has no column that tells you which ones.

Effective CPC is the price after that correction: what a paid brand click costs once you subtract the clicks organic would have caught anyway. It is usually a multiple of the sticker price. And you can measure it.

How do you measure it?

You turn some of it off on purpose. The design borrows from the holdout testing article and points it at money. A holdout is a fair test: change one slice, leave a comparable slice alone, read the difference. The season everybody lived through cancels out.

So you dial brand spend down in a controlled slice: a region, a match-type segment, a scheduled window. The rest of the account keeps running as the comparison, and you watch what organic recovers in the slice you suppressed.

If organic catches most of the lost paid clicks, most of that spend was buying traffic you already owned, and the effective price of the genuinely incremental clicks is the whole budget divided by the few that were real. If organic catches little, the spend is defending something.

The slice design carries the honesty. A suppressed region compared against its own history and against the untouched regions absorbs seasonality and campaign noise. What is left in the gap is your paid channel's real contribution.

The two bars below make the point. The short one is what the invoice says, the long one what the click cost after organic recovery. Those lengths are illustrative. The gap between them is the argument.

What the number actually changes

It turns a vague cannibalization debate into a price tag. Brand defense at a knowable cost can be worth every cent when a competitor bids your name, or when an AI Overview, the AI-written block Google puts on top of a results page, crowds your listing.

The same spend with high organic recovery and no competitive pressure is a subscription to your own traffic, and neither case is obvious from inside the account.

Either way the renewal conversation changes shape. It stops being about whether brand spend feels wasteful and becomes about whether this measured price is worth this measured protection, and defense priced at its effective cost sits beside every other protective spend the business already funds.

The test has honest limits

Suppression windows share the weather with everything else. Seasonality, a promotion, a competitor launching mid-test. The comparison slice absorbs most of it, and the significance check catches the rest: a statistical test of whether the change beats ordinary week-to-week wobble, at the bar every other claim has to clear.

Recovery rates drift too. A test from last year describes last year's results page, and AI Overviews have been rearranging exactly the pages where brand terms live. Re-run the measurement when the page changes shape, and date the number when you quote it.

One more boundary. Suppression tests read traffic and conversions in your named goals, the goals your team defined in your analytics tool by name, a Purchase or a Demo request. Finance can audit the readout without a glossary.

Overlap findings feed the test queue

Your test candidates come from the overlap check: keywords where you pay while ranking at the top organically. The article on paying for keywords you already rank for covers that worklist. The highest-spend overlaps pay back fastest.

The tier policy then consumes the result. A defend-tier keyword with a measured effective CPC gets a budget somebody can stand behind, and the tiered paid and organic operating policy article shows where it sits.

Sequencing matters more than volume. Test the biggest overlap first, bank that finding, and price the keyword properly. Then let each result decide whether the next test is worth running, because two or three well-chosen suppressions usually reprice most of a brand budget.

See also: the overlap check that builds the queue →

Measured beats modeled

Plenty of tools will model your cannibalization from industry curves. A suppression test measures yours: your brand, your page, your buyers, this quarter. Where the number came from is what separates a debate from a decision.

One habit completes it. Report effective CPC beside sticker CPC, with the test date attached to both. The gap between the two numbers is the entire argument, made in one row.

Your paid team tends to become the test's biggest advocate once the first one lands. A budget with a measurement behind it survives planning season. A budget defended with conviction survives until the first hard question.