It sounds aligned and usually is not, because the metric chosen is almost always ad revenue or ROAS, and both can be inflated by bidding on your own brand name. If you use this model, define performance as profit after ad spend, agree the baseline first, and cap the payout.
The short version
- ROAS is the wrong target. It counts sales the ad was credited for, not sales the ad created.
- Branded search is the loophole. Bidding on your own name buys clean numbers and adds nothing.
- Incrementality is the real question. What would have sold anyway, and how does anyone know.
- A cap protects both sides. Uncapped upside on a seasonal spike is a bill nobody planned for.
- Ask who is actually in the account. The pricing model matters less than whether the operator sits inside the firm.
The mistake that makes this model look good
The common error is judging a PPC engagement on the advertising numbers alone. Ads are the easiest part of the account to make look excellent and the hardest part to make matter. Push the budget toward branded terms and exact match keywords you already dominate, and ROAS climbs, ACoS drops, and the campaign dashboard turns green. Total sales barely move, because those customers were arriving anyway.
Under a pay per performance agreement that behavior is not a temptation, it is the design. The payout is calculated from the metric, so the metric is what gets managed. This is why the definition of performance is the entire negotiation and the percentage is a detail.
There is a second version of the same mistake. If your conversion rate is low, no amount of ad spend fixes it, and a performance deal on advertising quietly pays somebody to keep buying clicks for a page that does not convert. The money goes to traffic because traffic is what the contract measures.
Three ways to buy PPC management, compared
| Model | What you pay for | Fails when | Decision rule |
|---|---|---|---|
| Flat management fee | Time and judgment | You want the provider to carry downside | Default choice under about $50,000 monthly profit |
| Percentage of ad spend | A share of your budget | Always. Their income rises with your spend | Avoid |
| Pay per performance | A defined result | The metric is revenue or ROAS rather than profit | Only with a profit base, an agreed baseline and a cap |
The decision rule in one line: if you cannot write down, in a sentence, what number triggers the payment and where that number comes from, the model is not ready to sign.
If you are going to do it, structure it like this
- Define the base as profit after ad spend, not revenue, not ROAS, not attributed sales.
- Agree the baseline from the prior twelve months, taken from one named report, before any work starts.
- Exclude branded search from the measured set, or the loophole stays open.
- Set the window at a full quarter. A month is noise, especially on a seasonal catalog.
- Cap the payout at a multiple of what a flat fee would have cost, so an unexpected spike does not produce an unmanageable invoice.
- Write the stop clause. Define what happens if the right advice is to cut spend, and make sure giving it does not cost them money.
Six clauses. Any provider who resists all six is telling you the model was the pitch, not the plan.
The question that outranks the fee model
Ask who does the work and where they sit. PPC management is often the layer where work gets passed along, and you can be paying a performance rate to a firm that is paying someone else a flat one. Flapen does not subcontract anything, 100 percent of the work is in house, including the sourcing studio in Guangzhou and the creative studio in Dubai, and our own engineering team builds the advertising tooling we run on. I am not claiming that is the only workable arrangement. I am saying you should know the answer before you agree to any fee structure, because a performance contract with an unnamed subcontractor behind it is a contract with nobody.
What a performance PPC proposal will not tell you
That advertising is often not the constraint. On a large share of the accounts we audit, the ad account is competent and the listing is the problem: a primary image losing the click, a price that no longer matches the category, a variation family splitting review count. A performance deal on ads takes a fee for optimizing the one layer that was already fine.
The second thing left unsaid: performance pricing shifts risk in name, but the media budget is still yours. You fund every click. If the quarter goes badly you have paid for the traffic and they have earned less. Genuine risk sharing would touch the budget, and almost no proposal does.
Related answers
- Alternatives to pay for performance Amazon management
- How to evaluate Amazon PPC agencies
- Are performance bonuses worth it on Amazon
- What to ask about fee structure vs ad spend
- Hiring an Amazon agency: the complete guide
If you want to know whether ads are really your constraint, request the audit at Flapen.
