Compare total cost of employment against a published fee, not salary against fee. Then compare capability. The failure mode is not overpaying, it is buying one generalist to cover advertising, creative, and sourcing. Ask both options for their launch ACoS target and their maturity target.
The short version
- Salary is not the cost of a hire. Taxes, benefits, recruitment, software, and supervision belong in the same column.
- A retainer is a variable cost with a notice period. A salary is a fixed cost with a legal process attached.
- The generalist gap is the expensive one. Three specialisms, one paycheque, and the gaps show up on your catalog.
- Ask for two ACoS numbers, not one. A partner who quotes a single target is not managing by product stage.
- Continuity is a line item. One person means one holiday and one resignation between you and a stalled account.
The comparison almost everyone runs, and what it hides
A seller pulls up a market salary figure for an Amazon manager, sets it beside a monthly retainer, multiplies the retainer by twelve, and decides. That comparison is wrong in both columns, and the direction of the error is consistent: it makes building look cheaper than it is and buying look narrower than it is.
Build the internal column properly. Salary, employer contributions and payroll taxes, benefits, the recruitment fee or the internal time to hire, software licenses that an agency already pays for, a laptop, and the management hours of whoever supervises the role. Then add the ramp, because nobody is productive in month one on an unfamiliar catalog, and add a replacement provision, because this role turns over.
Build the external column properly too. Our fee is published: $800 a month for one product, $1,150 for two, $1,500 for three, $1,950 for four, $2,400 for five, with every service included at every tier and no onboarding charge. What sits outside any agency fee is your inventory, Amazon's own fees, freight, trademark filing, and your advertising budget. If a proposal blends those together, ask for a version that separates them before you compare it with a payroll number.
Failure modes, ranked by what they cost
| Rank | Failure mode | What it costs you |
|---|---|---|
| 1 | Comparing a salary to a fee | An internal budget that is wrong from month one |
| 2 | One generalist covering three specialisms | Capability gaps discovered on your own catalog |
| 3 | A single advertising efficiency target for every product | Launches starved of spend, mature products bleeding it |
| 4 | No cover for absence or resignation | Weeks of drift, usually at the worst possible moment |
| 5 | Paying twice for overlapping scope | A retainer and a salary buying the same eight hours |
One generalist, three specialisms
Read the job description you are about to post. Keyword research and copy. Advertising structure across campaign types. Creative direction for images and A-plus content. Supplier negotiation and quality control. Account health. Forecasting. Possibly a second marketplace in another language.
Those are not one job, and the candidates who do all of them well have generally built their own brands rather than applying to run yours. What arrives is strong in one area, competent in a second, and learning the rest live. The retainer column solves this differently: you are renting a bench rather than a person. Whether that bench is real is a separate question, and it is the one worth interrogating hardest.
One efficiency target for every product
This is the failure mode that costs the most and gets noticed the least. A new product and a two-year-old product should not carry the same advertising cost of sale target. At launch you are buying rank, review velocity, and data, so the target is deliberately aggressive and the spend looks inefficient on a spreadsheet. At maturity the job is protecting margin, so the target tightens. Managing both with one number means the launch never gets enough oxygen and the mature product quietly overspends for a year.
Ask any candidate, salaried or external, for the launch number, the maturity number, and what triggers the move between them. A person who has run both stages answers in about fifteen seconds. A person who has run reports does not.
Where the retainer loses
I would rather say this plainly than pretend the external column always wins. If Amazon is your primary revenue channel, your catalog is large, and you can carry a fixed cost through a bad quarter, an internal team gives you control, institutional memory, and a capability that compounds. That is a real advantage and no monthly fee replicates it.
The hinge is scale and fixed-cost tolerance. Below that line, the same money spent externally buys more of the capability list, and it buys it with a 30-day exit instead of a redundancy process.
Flapen runs entirely in-house from Abu Dhabi, no subcontracting, which is the only reason we can put creative, sourcing, and advertising in one fee. Ask your other candidates the same question and score the answer, including if it is no.
What a recruiter will not tell you
The role is hard to keep filled because it is three jobs, and the person who becomes excellent at it becomes employable elsewhere very quickly. Budget for that. The second thing worth knowing is that whichever column you choose, someone inside your company still has to own pricing, inventory, and product direction, and that person needs enough context to argue when the advice is wrong. Brands that hand the whole function away and disengage do worse than brands that stay informed for two hours a month.
Related answers
- Amazon agency vs in-house team: pros and cons
- Alternatives to hiring an in-house Amazon manager
- KPIs an Amazon agency should report weekly
- Amazon brand launch playbook: internal vs external
- Amazon agency pricing and economics: the complete guide
Put our published fee next to your payroll model and see which column buys more, at Flapen.
