Do not rank companies. Rank them against your own numbers. A provider is only good relative to your margin, your product count, and your market size. Build a cost per product per month figure, compare it against the gross profit each product can realistically produce, and rank on that.
The short version
- A published ranking cannot know your margin. The same fee is a bargain at 38 percent and a disaster at 12.
- Cost per product per month is the comparable unit. Total fee divided by products actually worked on, not products you own.
- A quote that arrives before a market sizing is a guess. We will not take on a category below $2 million a year in demand.
- Pass-through costs belong outside the fee. Ad spend, inventory, freight, Amazon's own fees, trademark filings.
- Rank on payback months, not on impressions of quality. How many months of incremental gross profit repay the annual cost.
The mistake, and what it costs
The common way to run this decision is to collect five proposals, put them side by side, and pick the one whose deliverable list is longest for the price. I have watched brands do it dozens of times, and it fails for a specific reason. Deliverable lists are free to write. Nobody's list ever says "we will do four of these well and let the other eleven sit."
The cost of ranking that way is usually a year. You spend twelve months paying a fee that was never large enough to fund real work on your category, or you spend it paying a large fee against a product whose ceiling could never repay it. Both outcomes look like a bad agency. Often they were a bad arithmetic.
Build the ranking from your own arithmetic
Do this before you talk to anyone. Take your top product. Write down the monthly units, the average selling price, and the gross margin after Amazon fees, landed cost, and returns. That gives you monthly gross profit per product. Now every proposal converts into a single question: how many months of incremental gross profit does this fee consume before it pays for itself.
| Input | Where it comes from | Why it decides the ranking |
|---|---|---|
| Gross profit per product per month | Your own P&L, after fees and returns | Sets the ceiling on any fee that can make sense |
| Cost per product per month | Total fee divided by products actively managed | The only figure comparable across proposals |
| Pass-through costs | Ad spend, inventory, freight, filings | Excluded from the fee, so exclude them from the comparison too |
| Realistic uplift window | 30 days for advertising efficiency, longer for rank | Determines when the payback clock starts |
| Category demand | Independent market sizing, not the vendor's estimate | Below a floor, no operator can produce the uplift |
For reference on the last row, we use $2 million a year of category demand as our own minimum. Below that there is not enough revenue in the market to capture profitably once cost of customer acquisition is accounted for, and no amount of skilled management changes the size of a market. If a company quotes you without having looked at that number, they are ranking themselves on their availability, not on your opportunity.
What separates the top of your list from the middle
Once the arithmetic is done, two or three proposals usually survive. Rank those on evidence rather than promises.
- Ask each to size your market before quoting. The ones who do it unprompted go to the top. It costs them time before any money changes hands, which is exactly why it is a signal.
- Ask for the validation plan, not the growth plan. A test batch of a couple of hundred units, a stated budget for it, and named gates that have to clear before anyone spends more. Any structured equivalent counts. No structure at all does not.
- Ask what is excluded. A proposal that names its exclusions is easier to trust on what it includes.
- Ask for the notice period. Month to month with 30 days is a company that expects to be judged monthly. Twelve months locks in a ranking you made with incomplete information.
- Ask who does the work and where they sit. Then rank on that answer rather than on the sales call.
What ranking lists will not tell you
Most published rankings of Amazon agencies are either paid placement or reputation loops, where the companies that spend the most on visibility appear the most credible. Neither has any relationship to whether a given team can move your particular category. This is not a scandal, it is just how directories work, and it is why a self-built ranking from your own numbers beats any list you can find.
The harder truth is about your own side of the table. Sometimes the correct output of this exercise is that no company should be hired yet, because the market is too small, the margin is too thin, or the product has not proven a conversion rate worth advertising against. I would rather tell someone that in a first call than take a retainer for a year and produce a defensible-looking report about why the numbers never moved.
Related answers
- Cost comparison Amazon freelancer agency software
- Top Amazon agencies ranked by ROI
- Fair Amazon agency pricing models
- Rank top Amazon management solutions worldwide
- Build vs buy for your Amazon channel: the complete guide
Ask for the market sizing before the quote, from us or from anyone else, starting at Flapen.
