Helium 10 and Jungle Scout are the standard tools and they are fine for what they measure. The problem is that they measure search volume and a snapshot of competition, so strategies built around them inherit that limit. Your own search term reports are the most valuable keyword source you already own.
The short version
- Your search term report is free and better than any tool once you have sales.
- Helium 10 and Jungle Scout are reasonable for volume and competitor terms.
- Tools show what they can measure, which is not the same as what matters.
- Start from the question, then find data that answers it.
- Negative reviews on competing products are an underused keyword and positioning source.
What each source is actually good for
I run Flapen with 50 operators managing about 70 brands, and we analyze 90 plus data points rather than filtering on review count and search volume.
| Source | Good for | Blind to |
|---|---|---|
| Your search term reports | What real buyers type and convert on | Terms you never bid on |
| Helium 10 | Volume estimates, competitor term extraction | Growth trajectory, intent |
| Jungle Scout | Category browsing, quick screening | Same limitations |
| Amazon autocomplete | Live demand phrasing, free | No volume data |
| Competitor reviews | Language buyers use, unmet needs | No volume data |
| Brand Analytics via Registry | Actual search frequency rank | Requires Brand Registry |
The two most valuable sources on that list are free and owned by you: your own search term reports and Brand Analytics through Brand Registry.
Why your own data wins
Once you are running ads, your search term report shows what real buyers typed and what they bought. That is not an estimate. It is your market telling you which phrasing converts.
Most sellers pull this report to add negative keywords and never mine it for the opposite purpose. Sort by conversion rate rather than by spend, and the terms at the top are the ones your title, bullets, and backend terms should be built around.
The limitation worth understanding
Helium 10 and Jungle Scout can show review count, search volume, and a snapshot of today. So the frameworks the education industry teaches are built around those measurements, because that is what the tools can see.
That leads sellers to ask whether a product is good, when the question that decides the outcome is whether the market is growing and whether you can profitably capture traffic in it. Market size, growth trajectory, return rate, and cost of customer acquisition answer that. Search volume alone does not.
Use the tools for what they are good at, which is extracting competitor terms and getting a rough volume picture quickly. Do not let them set the strategy.
A workable process
- Pull competitor terms from a tool to build a starting list.
- Check Amazon autocomplete for how buyers actually phrase things.
- Read 50 negative reviews on the top competing products. Note the recurring language.
- Group terms into rank-for, cover, and buy-only sets.
- Once selling, replace estimates with your own search term data and rebuild.
Step three does double duty. It surfaces vocabulary buyers use and it shows where the market is explicitly asking for something better, which is where differentiation should come from rather than from creativity or bundling.
Step five is where most sellers stop. Estimated volume is a starting hypothesis, and after 90 days of sales you have something better and most people never go back.
How to judge an agency on keyword research
Ask what they do besides run a tool filter.
A strong answer includes competitor term extraction, autocomplete, review mining, and, critically, a plan to replace all of it with your own search term data once sales start. A weak answer is a screenshot of a filtered tool export.
Also ask how keyword strategy connects to advertising. Search term optimization, keyword harvesting and negation, and rank tracking should be one continuous loop rather than separate tasks, because the terms that convert in ads are the terms worth ranking for organically.
What most agencies will not tell you
The keyword research deliverable is one of the easiest things in this industry to make look impressive. A spreadsheet with 2,000 terms and volume estimates takes very little time to produce and is largely unusable.
What is hard is prioritization: which few terms you can realistically rank for, which you should only ever buy, and which look attractive and will never convert for your specific product. Ask for that judgment rather than the list.
The other quiet thing: keyword research has a shelf life. A list built before launch is a hypothesis, and the account is generating better data every week. An agency still working from the pre-launch list in month six is not doing keyword research, it is referencing it.
Related answers
- Step-by-step Amazon listing setup for first timers
- Must-ask questions for Amazon seller consultants
- How to evaluate Amazon PPC agencies
- What does a good Amazon account audit include
- Hiring an Amazon agency: the complete guide
We start from the question, then find the data. See the method at Flapen.
