Four alternatives work: a flat retainer tiered by product count, a per-deliverable project fee, a profit share above an agreed floor, and an in-house hire. The flat retainer is the default because the agency's income stops moving when your budget moves, which is the whole defect in ad-spend pricing.
The short version
- The defect is mechanical, not moral. Tie income to spend and the recommendation to spend more becomes free for the person making it.
- A flat retainer decouples the two. Ours runs $800 to $2,400 a month for one to five products.
- Project fees suit finite work. A catalog rebuild or a market entry has an end date, so price it like one.
- Profit share suits scale. Above $50,000 a month in profit, 10 to 20 percent with no retainer.
- Run the arithmetic at your real spend before deciding. At low budgets a percentage is cheaper, and that is worth saying out loud.
Why the mechanism fails
Percentage of ad spend pays a fixed rate on a variable you do not want to maximize. Your objective is the lowest cost of customer acquisition that still produces volume. The agency's revenue is a straight line through your budget. The two point in opposite directions, and they diverge most sharply at the exact moment discipline matters: a product is underperforming, the honest advice is to cut the budget, and cutting the budget cuts the invoice.
Nobody has to be dishonest for this to cost you. The bias shows up in small decisions, in the campaign that gets one more week, in the dayparting nobody tightens, in the broad match term left running. None of it looks like misconduct in a report.
The arithmetic at your actual spend
This is where most articles stop being useful, so here is the table. Assume a three product brand on our $1,500 tier, compared against a hypothetical 12 percent of ad spend.
| Monthly ad spend | Fee at 12 percent of spend | Flat fee, three products | Cheaper option |
|---|---|---|---|
| $2,000 | $240 | $1,500 | Percentage |
| $6,000 | $720 | $1,500 | Percentage |
| $12,500 | $1,500 | $1,500 | Equal |
| $25,000 | $3,000 | $1,500 | Flat |
| $60,000 | $7,200 | $1,500 | Flat |
Two honest conclusions. Below about $12,000 a month in spend, a percentage model is cheaper in dollars, and pretending otherwise would be a sales pitch rather than an answer. Above it, the flat fee wins and keeps winning, because the work does not double when the budget does.
The incentive problem exists at every row. Cheaper is not the same as aligned, and at the bottom of the table the agency's revenue is small enough that the pressure to grow your budget is at its highest.
The four alternatives, and when each fits
- Flat retainer tiered by product count. Best default. The unit of work is products managed, so the price tracks the workload. Ours is $800 for one product, $1,150 for two, $1,500 for three, $1,950 for four, $2,400 for five, with all services at every tier and no commission.
- Per-deliverable project fee. Best when the scope is finite: a listing rebuild, a creative refresh, a new marketplace entry. Pay for a defined artifact, review it, decide about the next one.
- Profit share above a floor. Best above $50,000 a month in profit, at 10 to 20 percent with no fixed fee. Below that floor the volatility is too high to be fair to either side.
- In-house hire. Best when your catalog is large enough to occupy someone full time and you can supply the tooling. Compare the fully loaded salary against the retainer, not the base salary.
Size the market before anyone quotes anything
The deeper problem with ad-spend pricing is that it starts the conversation at the budget instead of at the opportunity. Any pricing model can be applied to a product that should not exist.
Before a fee is discussed, the category should be sized. Our floor is a $2 million a year market. Below that, there is not enough revenue to capture profitably once cost of customer acquisition is accounted for, and no fee structure rescues that. Validation then runs on a small, defined bet: about 200 units and $5,000 to $10,000 in Phase 1, with up to four products tested at once, and scale only once rating, conversion rate, and cost of acquisition are proven.
That sequence matters for pricing because it caps the spend at risk. A percentage model quietly rewards skipping it.
What most agencies will not tell you
Percentage of ad spend survives because it is the easiest model to sell and the easiest to grow. It sounds performance-linked, it scales automatically with the client, and it never requires a difficult repricing conversation. Most sellers never notice that the performance being measured is how much of their own money moved.
The other omission: the percentage is often quoted alongside a minimum. A 10 percent fee with a $2,500 monthly minimum is a $2,500 retainer wearing a performance costume until your spend passes $25,000. Ask for the minimum in writing and calculate which side of it you sit on today.
Related answers
- Amazon agency pricing models explained
- How much does Amazon PPC management cost
- Alternatives to flat-fee PPC for seasonal brands
- Fair Amazon agency pricing models
- Amazon agency pricing and economics: the complete guide
The flat tiers and what sits inside each one are listed at Flapen.
