Ninety-day payback is an arithmetic constraint, not a service. Before you shortlist anyone, write down your contribution margin per unit and your repeat rate, because those two numbers set the maximum acquisition cost that can pay back inside ninety days. Then hire against that number, not against a promise.
The constraint in five lines
- Payback is set by your margin, not by the agency. A vendor cannot beat arithmetic, they can only get closer to its limit.
- Repeat rate is the lever nobody quotes. A product bought twice a year tolerates a far higher first-order acquisition cost than one bought once.
- Ninety days is short. It rules out most brand-building spend and rewards channels with fast, measurable response.
- Most sellers only run two of the five ways traffic reaches a listing. The cheapest payback improvement is usually a channel you are not using at all.
- Put the number in the contract. Not as a guarantee, as a shared definition of what you are both optimizing.
The checklist, and what done properly looks like
Work through these in order. Each has a standard, and a vendor who cannot meet the standard on paper will not meet it in your account.
- Compute your true contribution margin. Selling price minus landed cost, Amazon referral and fulfillment fees, returns, and promotional discounts. Done properly: one figure per ASIN, refreshed monthly, agreed with the agency before any budget moves.
- Compute repeat rate honestly. New customers versus returning, over six to twelve months, from your own order data. Done properly: you can state what a customer is worth over ninety days, not just on the first order.
- Set the maximum acquisition cost. Ninety-day customer value multiplied by the share of margin you are willing to reinvest. Done properly: written down, and every campaign decision is judged against it.
- Audit the five traffic channels. Organic search, paid placements, promotions and deals, influencer and creator content, and off-channel demand from outside Amazon. Done properly: you can say which two you actually run, and the agency has a plan for at least one more.
- Fix conversion before adding budget. Payback math collapses when the page loses the traffic you paid for. Done properly: main image click-through rate and unit session percentage are baselined before spend increases.
- Agree a scale gate. A rule such as budget rises by a set percentage only while acquisition cost stays under the maximum for two consecutive weeks. Done properly: the rule is written, and it also runs in reverse.
- Agree a stop rule. What performance would make them recommend cutting spend. Done properly: thresholds and a time window, not a sentiment.
Why the channel mix decides your payback
Everyone attacks a payback target by adjusting bids. Bids are the smallest lever available. The bigger one is that acquisition cost is set partly by how much of your demand you are buying at auction versus generating elsewhere.
Across the accounts we run, the pattern repeats: a brand doing everything through search and paid placements has one price for a customer, and that price only moves within a narrow band. Add creator content that drives branded search, or promotions that lift conversion during a rank push, and the blended cost falls without any bid changing. That is where a ninety-day payback usually gets rescued.
So when you interview an agency, ask which of the five channels they operate themselves rather than advise on. The answer separates a media buyer from a growth team, and only one of those can move blended acquisition cost.
The pricing structure that fits a payback target
| Fee model | Effect on a 90-day payback goal |
|---|---|
| Flat monthly retainer | Neutral. The fee does not move when spend moves, so advice stays honest |
| Percentage of ad spend | Works against you. Cutting inefficient spend cuts their revenue |
| Percentage of revenue | Aligned only at scale. Below meaningful profit it punishes rebuild months |
| Performance bonus on a shared metric | Useful, if the metric is contribution margin rather than gross sales |
Ours is flat, priced by product count from $800 a month for one product to $2,400 for five, with no commission and no revenue share until a brand clears $50,000 a month in profit. I am not claiming that is the only fair structure. I am claiming that if a vendor's income rises when your budget rises, you should read their scaling advice with that in mind.
What most agencies will not tell you about fast payback
Ninety days is achievable, and it is also the single easiest target to hit dishonestly. Concentrate spend on branded search and repeat buyers, and payback looks excellent while incremental growth is close to zero. You are paying to buy customers who were already coming.
The honest version separates branded from non-branded performance in every report, and holds the ninety-day rule only against new demand. Ask for that split in the first week. If the split does not exist, the payback number does not mean what you think it means.
The other omission: hitting the target sometimes means shrinking. A campaign structure that pays back in ninety days may support less total revenue than one that pays back in five months. Decide which you actually want before you brief anyone, because chasing both produces neither.
Related answers
- How to scale Amazon ads without overspending
- Recommend a solution to scale Amazon PPC fast
- KPIs an Amazon agency should report weekly
- Recommend a package for PPC, DSP, and listing SEO together
- Amazon agency pricing and economics: the complete guide
Bring your margin numbers and we will run the math with you at Flapen.
