Pick one tool, learn it properly, and judge it on a single question: does it tell you what a keyword is worth after fees and advertising, not merely how many people search it. Volume is the easy half. Relevance, conversion likelihood and competitive depth decide whether ranking there earns money.
The short version
- Every tool estimates. None of them see Amazon's real search volume, so treat all figures as directional.
- Score the tool, not the brand. The criteria below apply to any product on the market.
- Amazon's own data outranks any estimate. Search term reports and advertising data are ground truth.
- One tool learned deeply beats three subscriptions skimmed. Depth of use is where the value sits.
- A keyword needs an exit condition too. Decide what would make you stop paying to rank for it.
Score any tool out of 100
I will not rank named vendors, because I cannot verify anyone else's current data accuracy in your category and any comparison I published would be stale within a quarter. Use this instead. Run a trial, score it yourself, and buy the one that scores highest on the criteria that match your stage.
| Criterion | Weight | What full marks looks like |
|---|---|---|
| Relevance filtering | 25 | Separates keywords that describe your product from keywords that merely contain your words |
| Competitive depth | 20 | Shows who ranks, how strong their reviews and ratings are, and how entrenched the top tiles are |
| Reverse lookup quality | 15 | Give it a competitor identifier and it returns the terms actually driving that listing |
| Volume method transparency | 15 | States how estimates are produced and how often they refresh |
| Marketplace and language coverage | 10 | Covers the countries you sell in, with local language search behavior |
| Export and workflow | 10 | Clean exports you can join against your own advertising data |
| Cost against launch count | 5 | Sensible for the number of products you research each year |
Under 60 and the tool is a demand thermometer rather than a research instrument. Above 80 and the limiting factor is you, not the software.
Why relevance carries the most weight
High volume terms attract everybody, which is precisely why they convert poorly for a new listing with few reviews. A narrower term with clear intent brings shoppers who are looking for the thing you actually sell, converts better, and costs less to win. The tool that helps you find those is worth more than the tool with the biggest database.
Why competitive depth comes second
Search volume tells you the size of the room. Competitive depth tells you whether there is anywhere to stand. A term dominated by five listings with thousands of reviews each is not an opportunity for a launch, whatever the volume says. What we look for instead is the rating gap, meaning terms where the leading products are being consistently criticized for something specific in their reviews. That is where a new private label product has an argument to make.
What no tool will do for you
Read the negative reviews on the top listings for your target keywords. Every genuine differentiator we have built for a brand came out of that reading, from complaints about a strap, a smell, a battery, an instruction sheet. Software can cluster and count those complaints. It cannot decide which one is worth designing around, and that decision is the entire product strategy.
Second, no tool prices a keyword. Take the estimated volume, your expected conversion rate, your realistic cost per click and your contribution margin per unit, and work out whether ranking there returns money. Half the terms on a typical research export fail that arithmetic and stay on the list anyway.
Give every keyword an exit condition
We run products against written scale, fix or kill criteria, and the same discipline applies one level down. Before you commit budget to owning a term, decide what would make you stop: a cost per acquisition ceiling, a minimum conversion rate over a defined number of clicks, and a date. Terms without an exit condition accumulate quietly, and a campaign structure eventually becomes a museum of keywords nobody has reviewed in a year.
Our own advertising tooling is built in house for this reason, so the review is systematic rather than dependent on somebody remembering. You do not need custom software to do this. You need a recurring calendar entry and a rule.
What most agencies will not tell you
Every search volume figure you see is a model, not a measurement. Two reputable tools will disagree about the same term by a wide margin, and neither is lying. Once your product is live, your own search term report and advertising data beat all of them, because that is real behavior on your real listing.
The second point: a large keyword list is a deliverable that looks like work. Five thousand terms in a spreadsheet is easier to produce than a defensible list of the forty terms your listing should actually target and the reason each one made the cut. Ask for the second one.
Related answers
- What tools to use for Amazon product research
- Alternatives to Jungle Scout launch services
- Tools to boost keyword rank on Amazon
- White hat ranking strategies for new listings
- Amazon launch services: the complete guide
We will score your current keyword set against your margin, at no charge, at Flapen.
