What is performance marketing, really?

"Performance marketing" has been diluted to near-meaninglessness. It now describes any agency that runs paid media and produces a monthly report, which is to say, nearly all of them.

We use the term about ourselves, so we owe you a definition specific enough to be falsifiable.

The definition we hold ourselves to

Performance marketing is marketing where the provider's compensation moves with the outcome the client cares about.

That is it. Not "we measure things." Not "we optimize for conversions." Not "we are data-driven." Those describe competent execution, which should be the floor rather than the pitch.

The distinguishing property is exposure. Does the agency's income change materially if the campaign underperforms? If the answer is no, the arrangement is a retainer with good reporting. That can be perfectly fine work, but it is not performance marketing, and calling it that muddies a distinction clients need.

If you cannot lose money when the client loses money, you are not a performance partner. You are a well-instrumented vendor.

What the retainer model does to the work

The retainer is not evil. It is predictable, it funds teams, and clients understand it. But it has three structural consequences that are worth naming.

It rewards activity over outcome. When the invoice is the same regardless, the natural equilibrium is to demonstrate effort. Hence the monthly deck with forty slides of impressions, reach, and engagement: metrics that are real and that nobody would choose if revenue were the measure.

It makes the difficult conversation expensive. The most valuable thing a marketing partner can say is often "stop spending on this." Under a retainer tied to ad spend, that sentence reduces the agency's income. The incentive to say it is negative, and people are people.

It hides the point at which value stops. Plenty of engagements deliver enormous value in months one to four and very little thereafter. The retainer does not notice. Both parties drift.

Why more agencies do not do it

Honest answer: it is harder, riskier, and it requires giving something up.

You have to underwrite the client. If your income depends on their results, you cannot take everyone. Their product has to work, their sales process has to convert, their operations have to handle volume. Performance pricing forces you to say no to a great many prospective clients, which is a revenue problem for the agency and a quality signal for the ones who pass.

You need real attribution, agreed in advance. Ambiguity about what counts is fatal when money hangs on it. What is a qualified lead? What is the attribution window? Who owns a customer who saw an ad, searched the brand, and called? These must be settled before work starts, in writing.

Cash flow gets harder. Outcome-based revenue arrives later and less evenly than a monthly fee. You need a balance sheet that tolerates that, which is a real barrier.

You must be willing to be wrong in public. A retainer lets you explain a bad quarter. Performance pricing prices it.

The four questions

If you are evaluating an agency claiming performance marketing, these separate the real thing from the term:

1. What happens to your invoice if we miss the target? If the honest answer is "nothing", you have your answer.

2. What do you consider a result, and who decides? A partner should have a strong, specific answer (qualified leads by an agreed definition, booked revenue, ROAS at a stated margin) and should be comfortable putting it in the contract.

3. What would make you tell us to stop spending? Anyone who has never said this to a client either has extraordinary luck or an incentive not to.

4. Which clients have you turned down? A firm that takes everyone is not underwriting outcomes. A firm that can describe the last three prospects it declined, and why, is telling you what its model actually is.

What it looks like from the inside

We run as a fractional CMO and performance venture studio. In practice that means:

• We underwrite before we sign. We look at the product, the economics, the sales process, and the operational capacity. If the model does not work, more traffic makes things worse, and we say so.
• The scoreboard is agreed on day one. One number that matters, defined precisely, visible to both sides continuously, not assembled for a monthly meeting.
• We are structurally able to lose. That is uncomfortable, and it is what makes our advice trustworthy. Our recommendation to cut a channel costs us money and we make it anyway, because the alternative is being the vendor who watched.
• We build the tooling. Attribution, reporting, and operations have to be genuinely reliable when money depends on them. It is a large part of why we build our own software rather than assembling dashboards.

The honest caveat

Performance pricing is not right for everyone, and we would rather say so than oversell it.

If you are pre-product-market-fit, the outcome is not yet controllable by marketing, and tying an agency to it produces bad behavior on both sides. If your sales cycle is eighteen months, the feedback loop is too slow. If you cannot instrument your revenue, no attribution model will save the arrangement.

In those cases a good retainer with a competent team is the right answer. What you should not accept is a retainer described as performance marketing, priced as though the agency shares your risk when it does not.

The term should mean something. Ours means we can lose.

If that is the arrangement you want, apply for partnership.