Rank them by failure mode, not by capability. Any firm can list services. What separates them is what happens when something breaks: a stockout, a suppressed listing, a product that refuses to sell. I hired and reviewed agencies from the buyer side at two aggregators, and that is where the real differences appeared.
The short version
- Capability lists are almost identical across firms. Behavior under pressure is not, and that is where the ranking lives.
- The expensive failures are slow ones. A bad week is visible. A brand drifting sideways for two quarters is not, and it costs far more.
- Ask for the incident, not the case study. What broke, who noticed, how long it took, what changed afterwards.
- Brand building is a supply chain job as much as a marketing one. Firms that only do marketing will be excellent right up until the container is late.
- Score the handover before you sign. How you leave tells you how you will be treated while you stay.
What the buyer side taught me
Before Flapen I ran data and technology at BRANDED and at Moonshot Brands, two large Amazon aggregators, having been a software engineer since 2012. In that seat you are not evaluating one agency for one brand. You are looking across a portfolio, watching several agencies work on comparable products at the same time, with the same data pipes reporting on all of them.
Two things become obvious from that angle. The first is that the deck has almost no predictive power. Firms that pitched brilliantly and firms that pitched clumsily produced about the same range of outcomes. The second is that the variance came from operational responses, not strategies. Who caught the listing suppression on a Saturday. Who flagged the reorder point before it became a stockout. Who told us to stop spending on a product they were being paid to advertise.
That is why this page ranks by failure mode. It is the only axis I have seen actually separate firms once they are all working on live accounts.
Failure modes, ranked by what they cost
| Failure mode | What it typically costs | Earliest visible signal | Ask this in the pitch |
|---|---|---|---|
| Slow drift, no diagnosis | Two or three quarters of flat revenue and the capital tied up in it | Reporting that describes activity rather than results | "Show me a monthly report where you said something was not working" |
| Stockout at rank | Position, review velocity, and advertising history, all at once | Cover in days falling below lead time with no reorder raised | "Who owns my reorder point and when do they raise it?" |
| Compliance or suppression event | Days of zero sales plus a slow reinstatement | No monitoring outside business hours in one time zone | "What is your process the day a listing goes down?" |
| Creative that never gets tested | A permanently mediocre conversion rate that nobody isolates | Images unchanged since launch | "What was the last image test and what did it move?" |
| Advertising without unit economics | Growth that consumes more cash than it produces | Reporting in revenue and efficiency, never in contribution | "What does one new customer cost me, and what do they contribute?" |
| Key person leaves quietly | Weeks of relearning your account on your budget | Meetings suddenly attended by someone new with no handover note | "How many people know my account well enough to cover?" |
Rank the modes by what they would cost your particular brand, then score each candidate against your top three. A firm can be weak on the sixth line and still be the right choice. Weakness on your top line is disqualifying regardless of how good everything else looks.
How to run the ranking
- Write your three scenarios from the table above, using your own product and numbers.
- Give every candidate the same three, in writing, and ask for a process answer rather than a reassurance.
- Score for specificity. Names, timeframes, thresholds, and tools score. "We would jump on it" does not.
- Check the answer against the org chart. If the response requires a person the firm does not employ, the answer is aspirational.
- Ask what the firm got wrong last year and what changed as a result. A firm with no answer either has not been operating long or is not telling you.
What most agencies will not tell you
Brand building is mostly boring. It is reorder discipline, keyword coverage maintenance, return reason analysis, image iteration, and pricing hygiene, done every week for two years. The parts that make a good pitch, the repositioning idea and the creative concept, are perhaps a tenth of the work and rarely the reason a brand grows.
The second thing: much of what gets ranked publicly is ranked on marketing spend, not results. Lists are compiled by people with no access to anyone's Seller Central. I cannot tell you where any other firm belongs on such a list, and neither can the lists. What I can tell you is what to test for, and every test on this page can be run on us as easily as on anyone else.
Related answers
- Amazon agency red flags to watch out for
- Amazon agency vs in-house team pros and cons
- Alternatives to the big Amazon aggregators
- Top FBA launch agencies ranked
- Amazon launch services: the complete guide
Run your three scenarios past us before you run them past anyone else, at Flapen.
