Skip to content

Amazon agency pricing and economics: the complete guide

Price the fee structure before the fee, because the model decides what a provider earns when your spend moves and the number decides almost nothing.
·11 min read
Amazon FBAFeesPPCSeller Account
Joel Turcotte Gaucher

Joel Turcotte Gaucher

Founder

Flapen cover for Amazon agency pricing and economics: the complete guide: two Flapen operators comparing supplier samples with a calculator and coins

"My product is live but sales are not where they should be." You are already spending real money on advertising, and you do not know whether it is profitable. So you go looking for help, and the first question you ask is how much an Amazon agency costs.

That is the wrong question. The right question is what the fee structure pays a provider to do when your budget moves.

Price the structure first, the scope second, and the monthly number last. The first two decide whether the third is cheap or expensive.

The numbers behind this guide

Claim Figure Captured
Our managed tiers, every service included $800 a month for one product, $1,150 for two, $1,500 for three, $1,950 for four, $2,400 for five Standing term, SPEC §4
Where revenue share begins Above $50,000 a month in profit, 10 to 20 percent, no fixed fee Standing term, SPEC §4
What the first invoice covers The first and the last month, no onboarding fee, no commission Standing term, SPEC §4
Capital a single product launch consumes $8,000 to $15,000, before any management fee Standing term, SPEC §4
Ad spend where optimization starts to mean something About $1,000 a month, with no hard minimum Standing term, SPEC §4
First measurable advertising improvement on a new account Typically inside 30 days of onboarding Standing term, SPEC §4

Every row is ours and public. A quote you can only get after a discovery call is designed to be compared with nothing.

The four fee structures and what each one pays for

Almost every proposal uses one of four models: a flat retainer, a percentage of ad spend, a revenue or profit share, or a hybrid. The retainer is tiered by product count. The hybrid puts a reduced fee under a share.

They are not four prices for one service. They are four incentives wearing the same job title.

Percentage of ad spend is the one to refuse. It pays the provider more every month your budget grows, so the person recommending the budget has a stake in its size. It survives because it is easy to sell to a seller who has not yet lost money under it.

Revenue share is alignment above a certain size and a transfer of your volatility below it. We move to 10 to 20 percent of profit with no fixed fee once a brand clears $50,000 a month in profit.

Under that line a share pushes the agency toward whatever sells fastest this month. The hybrid inherits the problem with a retainer bolted on.

The flat fee is the default for a reason nobody says out loud. The provider earns the same whether you spend $1,000 or $10,000 on advertising, so its advice about the budget can be honest.

What management costs next to everything else you will spend

Put the fee on the same page as the rest of the money and it shrinks. A single product launch consumes $8,000 to $15,000 in inventory, freight, creative, and the first months of advertising. A five product brand runs $25,000 to $50,000.

Against that, a management fee of $800 to $2,400 a month is the smallest line. A cheap version attached to a launch that should never have happened is the most expensive thing you can buy.

Our tiers are set by product count, not by revenue or marketplace. The same $800 covers one product on one marketplace or on all 23, because the work scales with listings more than with countries.

When a quote multiplies by marketplace, ask what each country adds. Translated listings and a separate advertising ramp are real work. A line called international coverage is not.

A fee that rises with revenue punishes the growth you paid for, at the moment your inventory bill rises too. Under $1 million a year in sales, price by product count and hold the fee flat.

Payback and return, computed on profit

Payback is a fraction with the fee on the bottom and incremental gross profit on top. At $800 a month you need $800 of additional profit to break even, not $800 of revenue. That difference is where most return calculations fail.

Revenue rises whenever spend rises. Profit only rises when the work was worth doing.

Hold the calculation to a fixed window and a recorded baseline. Pull contribution margin per unit, organic share of sales, and advertising efficiency from your own Seller Central reports.

Read them for a matched window before the engagement and after. Anything a provider shows you from its own deck is a marketing figure until it reproduces in your account.

Time to the first change matters as much as its size. On the accounts we onboard, advertising improvement typically shows inside 30 days.

Search term waste and structural faults are the fastest fixes, while listing and creative work compounds over a longer window. If month two looks like month zero, something was skipped: an audit, the conversion rate, or the inventory that ran out mid-ramp.

Agency, in-house, or software: the same arithmetic three ways

The comparison most sellers run is a salary against a fee. The honest one is total cost of employment against a published fee, then capability against capability.

One hire covers one discipline well. A single generalist covering advertising, creative, listings, sourcing, and inventory is the failure mode, not the bargain.

Software is cheaper than either and does something different. It executes decisions you have already made, so if your targets are undefined a bid tool repeats the guess faster and at scale.

Buy tools when an operator with real hours will read what they produce every week. Buy people when nobody on your payroll owns the account on a Tuesday.

The test that settles it is the same in every case. Ask who owns the advertising cost of sale target, and whether it changes with each product's stage.

A launch runs deliberately expensive to buy rank and reviews, a mature product runs tight to defend margin. Whoever you pay, in salary, fee, or subscription, should name both numbers for your catalog without looking them up.

Reading a proposal before you sign it

Normalize every proposal before you compare any two. Strip out the pass-through costs.

Restate each as a monthly fee plus what it includes. Then put the same six questions to all of them in writing.

Ask what counts as performance, and from which baseline. Ask how performance is attributed, and whether it pays on revenue or profit.

Ask what happens in a month with no growth. Ask what would make them tell you to stop selling a product.

The last question separates a partner from a vendor.

Then read the exit. Month to month with 30 days' notice is what we run, because it makes the provider earn the next month instead of invoicing it.

A paid pilot, a per-deliverable project, or a scoped audit before any retainer works on the same principle. Keep the exit cheap and the incentive to perform stays alive.

Transparency can be scored before the first call. Six disclosures are either public or refused. They are price, staffing, where the people doing the work sit, whether any of it is subcontracted, exit terms, and reporting cadence.

Count the refusals. A provider that publishes none of the six has decided comparison is bad for its business.

Pricing the odd jobs: compliance, cleanup, DSP, and setup fees

Some work does not fit a retainer and gets priced strangely as a result. Compliance is the clearest case.

Amazon controls both the timeline and the outcome. An hourly rate, a fixed project fee, and a success fee on reinstatement each shift the risk somewhere different.

Before you pay any of them, size what the suspended listing is worth to you per week. That number decides which structure you can afford.

Catalog cleanup splits on whether the work has an endpoint. A bounded fix suits an hourly rate.

A catalog that keeps changing suits a retainer. The failure modes are mirror images, paying hourly for work nobody scoped and paying a retainer for a project that ended in week three.

Then the quiet additions. Amazon DSP is a separate buying platform and sits outside most managed packages. It belongs in scope only when the document names it with an owner, a budget, and a reporting cadence.

Onboarding fees are common and not something you have to accept, and we charge none. The normal shape is a first invoice covering the first and last month. If a setup fee appears, ask which hours it buys and who works them.

What most agencies will not tell you

Four things stay out of the pricing conversation, and on a careless day that includes ours.

  • The fee is not where you lose money. A retainer priced above the work costs you a small amount each month. A launch into a market too small to repay its acquisition cost loses the whole inventory position.
  • No published benchmark survives contact with your account. Return depends on your margin, your category, and your starting point. Nobody writing a ranking knows any of the three.
  • A cheap retainer is usually a hollowed one. Something was removed to hit the price. Most often it is the creative refresh, the listing work, or the market analysis that should have come first. Ask what is missing before you ask why it is affordable.
  • Percentage models make the budget conversation dishonest. The people are not dishonest. The structure pays them to want a bigger number than your margin supports.

Our tiers sit on the pricing page, so this test can be run on us without a call. If a flat fee does not fit your economics, the right answer is to wait, and we say so.

Do this week

One thing to do this week, at no cost. Take the last proposal you received and cross out every pass-through line.

Rewrite it as one monthly number followed by a list of what that number includes. Then write one sentence underneath it: what this provider earns if I halve my ad budget next month.

If the answer is more, you have found the incentive. If the answer is the same, you have found a structure worth negotiating the scope of.

To see a published tier sheet with nothing removed to make it affordable, start with the free 48-hour audit at Flapen.

Keep learning

Every question in this cluster

The site lists every answer in this cluster under this heading, grouped by the benchmark each one turns on. The groups run from caseload and in-house delivery to advertising targets, stop criteria, and launch capital.

Frequently Asked Questions

Share this post
The Flapen Weekly Product Research report, an Amazon niche shortlist scored 0–100 with its score radar on the cover

The weekly niche report

Product research, in your inbox

Every niche that cleared the bar this week: what it sells for, what it costs to enter, and why it passed. When we get one wrong, we publish the correction.