Amazon autocomplete
Amazon's own search bar is a direct window into what real buyers are searching for. Type your seed keyword and systematically add letter by letter after it ("yoga mat a," "yoga mat b," and so on), recording every autocomplete suggestion. Then repeat the process with letters before the keyword ("a yoga mat," "b yoga mat"). Do the same with question formats ("yoga mat for...," "yoga mat with..."). This takes 20–30 minutes per seed keyword and produces a list of buyer-validated phrases pulled directly from Amazon's own search data.
Autocomplete results reflect actual search frequency — Amazon won't suggest a phrase in autocomplete if it doesn't get meaningful search volume. Phrases that appear consistently across multiple starting letters are particularly strong signals. You can also search in incognito mode to avoid personalization effects, and try with different geographic locations via VPN if you sell in multiple marketplaces.
Brand Analytics
Amazon Brand Analytics is free for brand-registered sellers and provides actual search frequency rank data — not estimates from third-party tools, but real numbers from Amazon's index. The Search Terms report shows the top clicked ASINs for any given search term, their click share, and their conversion share. This tells you not just what buyers search, but which products they buy after searching — critical data for evaluating whether a keyword actually converts in your niche.
If you're brand-registered, treat Brand Analytics as your primary volume data source. Enter your seed keywords and competitors' ASINs, review the related search term suggestions, and build your keyword list from that output. The data is more accurate than any third-party estimate because it comes from Amazon's own transaction and search records. Sellers who skip Brand Analytics while paying for third-party tools are often paying for worse data.
Competitor listing analysis
Your top three to five competitors have already done keyword research — you can read most of it in their listings. Pull up the listings that rank first for your primary keyword and read the title, bullets, and description carefully. Write down every keyword phrase you find that you haven't already listed. Pay particular attention to terms in their titles (those are the keywords they've deemed most valuable) and any phrases that appear in multiple competitor listings (consensus across competitors is a strong signal of keyword importance).
For each competitor listing, also look at their customer Q&A and review sections. Buyers often use very specific language to describe what they were looking for when they found the product — language that differs from how sellers describe their own products. A buyer who writes "I needed a water bottle that wouldn't leak in my backpack" is telling you "leak-proof water bottle for backpack" is a real search phrase worth targeting.
Review mining
Read 50–100 reviews across your top competitors and note recurring descriptive phrases. Look for: specific use cases mentioned ("takes this to the gym every day"), problem-solution language ("finally found one that doesn't leak"), attribute descriptions ("the handle fits perfectly in a standard cup holder"), and negative patterns about what other products in the category do wrong that yours doesn't. Each of these is a potential keyword cluster that reflects genuine buyer language.
One underused technique is looking at one-star and two-star reviews of your competitors. These tell you what buyers were hoping to find and didn't. If multiple buyers complain about a feature your product handles well, those complaints reveal search queries from buyers who are still in the market and looking for a better option. Addressing those specific concerns in your listing copy targets a pre-qualified, frustrated audience.
Affordable complements
If you want data-backed volume estimates without the cost of a full Helium 10 subscription, a few affordable tools are worth considering. MerchantWords offers a basic search volume lookup tool at a fraction of the cost of full suites. Keyword Tool Dominator has an Amazon-specific mode that pulls autocomplete suggestions across multiple permutations automatically. DataDive offers ASIN reverse lookup at competitive pricing for sellers who want to find what keywords a specific competitor is indexing for.
Your own PPC campaigns are also a powerful free keyword research tool. Run a broad match or auto-campaign for two to four weeks after launch, then pull the Search Term Report and review every query that generated at least one click. This shows you exactly what real buyers searched before clicking your listing — priceless data that no third-party tool can replicate because it's specific to your actual listing performance and buyer audience.
Putting it together
The complete manual process: start with autocomplete to build an initial list, validate with Brand Analytics if you're brand-registered, expand with competitor listing analysis, deepen with review mining, and confirm with your own PPC search term reports post-launch. At each stage, add unique new terms to your master list and tag them by type (primary, secondary, long-tail) and priority.
The limitation of this process versus a tool like Helium 10 is speed and completeness — a manual review that takes you three hours might take 20 minutes with a paid tool's ASIN reverse-lookup. For sellers with one or two products, the manual approach is entirely viable. For sellers optimizing ten or more ASINs regularly, the time cost of manual research eventually justifies a tool subscription. Make that decision based on your actual product count and how often you're doing research, not on aspirational usage.