Most Amazon sellers approach keyword research the same way they'd approach Google SEO — find high-volume terms, stuff them in the title and bullets, repeat. Amazon's algorithm is different. It rewards sales velocity on a keyword, not keyword density. The research process has to reflect that.
After managing keyword strategies across 30+ brands and $25M+ in Amazon revenue, here's the process I use every time — from initial seed keyword discovery to final PPC campaign assignment.
Step 1: Seed Keyword Discovery
Start with what your product actually is — not what you wish it ranked for. Enter your most direct, literal product description into Amazon's search bar and study the autocomplete suggestions. These are real, high-frequency search terms Amazon surfaces because shoppers type them repeatedly.
Record every autocomplete variant. These are your seed keywords — the core terms around which everything else is built. For a silicone baking mat, seeds might include: "silicone baking mat," "non-stick baking mat," "silicone mat for baking," "reusable baking mat."
Tools for seed discovery: Amazon search autocomplete (free), Helium 10 Magnet, Data Dive, Jungle Scout's Keyword Scout.
What to look for: Terms where the first page of results closely matches your product. If page 1 is dominated by a different product type, that keyword won't convert well even if you rank.
Step 2: Competitor Reverse ASIN Research
The fastest way to find high-converting keywords is to find the terms your best competitors are already converting on. Use a tool like Helium 10 Cerebro or Data Dive to run a reverse ASIN lookup on the top 3–5 products in your category.
This gives you the keywords those products rank for organically and the estimated search volume for each. More importantly, it tells you which keywords are driving their sales — not just traffic.
Filter for keywords where:
- Your competitor ranks in positions 1–10 organically (proof the keyword converts for that product type)
- Search volume is above 1,000/month
- The keyword is genuinely relevant to your specific product
Step 3: Segment Keywords Into Tiers
Not all keywords deserve the same treatment. I segment every keyword list into three tiers before assigning them to listings or campaigns:
Step 4: Validate With PPC Before Committing to Organic
This is the step most sellers skip — and it's why they end up with listing optimizations that don't move organic rank. Before you build your listing around a keyword, validate that it actually converts for your product using PPC data.
Run exact match campaigns on your top 15–20 Tier 1 and Tier 2 keywords for 2–3 weeks. Pull the data: which keywords delivered orders at an acceptable ACoS? Those are your real converters. The keywords with clicks but no orders — regardless of their search volume — are not your keywords.
The rule: If a keyword gets 15+ clicks and zero orders, it's not converting for your product. Remove it from campaigns, deprioritize it in listing copy, and move on.
High search volume + no orders = the intent doesn't match your product, your price, or your listing. More traffic won't fix it.
Step 5: Build Your Master Keyword List
What a Healthy Keyword Master List Looks Like
After running through the process above, a typical product ends up with:
- 3–5 Tier 1 keywords — the foundation of your listing and launch PPC strategy
- 15–25 Tier 2 keywords — supporting terms across your bullets, backend, and phrase match campaigns
- 50–100 Tier 3 keywords — long-tail terms in backend and auto campaigns for organic coverage
- Negative keyword list — irrelevant terms to exclude from PPC, built from the same research process
This isn't a one-time exercise. Refresh your keyword research every 90 days. Trends shift, competitors enter, and new search terms emerge. The brands that compound in Amazon rank are the ones that treat keyword research as an ongoing process, not a launch task.
Common Keyword Research Mistakes
- Targeting keywords where page 1 doesn't match your product. If you search your keyword and the results are a different product type, shoppers searching that term don't want what you're selling.
- Optimizing for volume instead of intent. "yoga" has enormous search volume. "yoga mat non-slip extra thick for hot yoga" has less volume but converts 10x better for the right product.
- Never updating the list. A keyword list built at launch is outdated within 6 months. Competitors, seasonality, and platform changes shift which terms convert.
- Treating Amazon search like Google search. Amazon shoppers have purchase intent. The keywords are shorter, more product-specific, and convert faster than informational queries. Research methods designed for Google underperform on Amazon.