
What should you do before Amazon keyword research?
Confirm the product is worth launching first. Define the exact product and the job it does, check demand on and off Amazon, work out landed cost and margin, map buyer intent, measure how deep the market is, and list the gaps competitors leave open. Keyword research then confirms the opportunity and shapes your listing and PPC. It cannot rescue a product with no margin or no demand.
Key takeaways
- Keyword research is the second step, not the first. A keyword list built for the wrong product, price or audience wastes the listing and the ad budget.
- Check margin before keywords: your break-even ACoS decides which keywords you can afford to rank for.
- Group demand by intent (problem, product type, feature, use case, competitor) before you look at search volume.
- Market depth matters more than total volume: if three listings take most of the sales, plan a narrower entry.
- Competitor 1 to 3 star reviews and customer questions are the cheapest source of your unique selling point and your best long-tail keywords.
Most sellers open a keyword tool on day one. We don’t. Keyword research is the backbone of a listing that ranks and an ad account that pays back, but the list it produces is only as good as the decisions made before it. If the product, price or audience is wrong, a perfect keyword list just helps the wrong listing get seen.
These are the six checks we run before we pull a single keyword for a client launch. They take a day or two, and they decide which keywords you should chase, which you can afford, and which you should ignore.
Why do prerequisites matter for Amazon keyword research?
Your keyword list feeds almost everything that follows: the 75-character title, the Item Highlights field, bullets, backend search terms, A+ content and every PPC campaign. Since 2026 it also feeds how Amazon’s assistant describes your product. When Rufus became Alexa for Shopping, more shoppers started seeing a short list of named products instead of a full page of results, so a vague product position costs more than it used to.
Prerequisites stop three expensive mistakes: ranking for traffic that will never convert, bidding on keywords your margin cannot support, and launching into a market where a few listings already take almost every sale.
1. Define the product and the job it does
Write down exactly what you are selling: size, material, pack count, variations and price point. Then write the job it does in the shopper’s words, not yours. A “1-litre insulated steel bottle” is a product. “Keeps water cold through a 10-hour shift” is the job.
- List five ways a shopper would describe the problem the product solves.
- List the situations they use it in (gym, office, hiking, kids’ school bags).
- Note anything that rules a shopper out, such as size, compatibility or age range.
This list becomes your relevancy filter later. When a keyword tool returns 2,000 phrases, you will score each against these descriptions instead of guessing.
2. Validate demand on and off Amazon
Search volume tells you people type a phrase. Social listening tells you why. Check both.
- On Amazon: type your product type into the search bar and note the autocomplete suggestions, read the customer questions on the top listings, and if you have Brand Registry, check Top Search Terms in Brand Analytics.
- Off Amazon: search Google, YouTube, Reddit and TikTok for the problem, not the product. Look for repeated complaints, comparison videos and “what should I buy” threads.
- Trend direction: a phrase growing month on month is worth more than a bigger phrase that is flat or falling.
If you only find demand on Amazon, the product will live and die by search and PPC. If you also find it on TikTok or YouTube, you have a second traffic source to plan for.
3. Check price and margin before keywords
Price analysis does two jobs: it tells you whether sourcing works, and it sets the ceiling on what you can spend to rank. Work out your landed cost (product, freight, duties and tariffs), then subtract the referral fee (15% in most categories), FBA fees and storage.
What is left before advertising is your break-even ACoS. If it is 20%, broad head terms with high CPCs are probably out of reach at launch, and your list should lean on long-tail and use-case phrases. We walk through this in five questions to answer before scaling Amazon PPC and in why price is a strategy, not a number.
4. Segment the market by buyer intent
Group demand by what the shopper is trying to do. Each group behaves differently in search and in ads, and each belongs in a different part of the listing.
| Intent | Example (insulated bottle) | Where it usually goes |
|---|---|---|
| Product type | insulated water bottle | Title, exact and phrase PPC |
| Feature | leak proof bottle with straw lid | Title or Item Highlights, bullets |
| Use case | water bottle for gym | Bullets, images, PPC ad groups |
| Problem | bottle that keeps water cold all day | Bullets, A+, assistant-style questions |
| Competitor or brand | [competitor] alternative | PPC product and brand targeting only, never backend |
Also note which categories and sub-categories the top listings sit in. Being in the wrong browse node can cost you ranking for the most relevant terms.
5. Measure market depth
Market depth is how many sellers are actually making real sales, not how many listings exist. Look at the top 10 to 20 organic results for your main keyword and estimate monthly sales and reviews for each.
- If sales are spread across many listings, a new entrant with a better offer can win share.
- If two or three listings take most of the sales, go narrower: a sub-niche, a bundle, or a use case they ignore.
- Check review age and velocity, not just totals. Old listings with thousands of reviews can still lose.
Keyword concentration tells the same story. In our competitive keyword audit, 143 of 500 keywords carried 81% of search volume, and a ten-month-old listing outsold an eleven-year-old one on keyword coverage. For crowded markets, see how to launch in a competitive niche.
6. Find the gaps competitors leave open
Read the 1 to 3 star reviews and the customer questions on the top five competitors. Write down every repeated complaint: lids that leak, sizes that are missing, instructions that confuse, packaging that arrives damaged.
Those gaps do two things. They become your unique selling point, if your product truly fixes them, and they give you long-tail phrases that shoppers actually use, which often convert better than the head terms. Make sure the product delivers what you claim. A strong listing on a weak product only speeds up the bad reviews.
The one-page prerequisite checklist
| Check | Question to answer | Pass signal |
|---|---|---|
| Product definition | Can you describe the job in the shopper’s words? | Five problem descriptions written |
| Demand | Is demand visible on and off Amazon? | Autocomplete, questions and social threads agree |
| Margin | What is your break-even ACoS? | Leaves room for launch ad spend |
| Intent | Which intent groups matter most? | Each group mapped to part of the listing |
| Depth | Are sales spread or concentrated? | Several listings with healthy sales, or a clear sub-niche |
| Gaps | What do competitors fail at? | At least two fixable, repeated complaints |
When every row passes, move on to the step-by-step keyword research method, then to listing optimization. If you want us to run the research for you, that is part of our product research and Amazon SEO work.
Frequently asked questions
Do I need keyword research before choosing a product?
Yes, at a light level. A quick look at search volume and competition helps you choose, but full keyword research comes after you have confirmed demand, margin and a clear gap. Otherwise you research keywords for a product you may not launch.
What is a good minimum search volume for Amazon keywords?
There is no fixed number. We often start with a floor of about 300 monthly searches for the main list in larger categories and drop it to around 100 in small niches, then keep lower-volume terms that are highly relevant.
How does margin affect which keywords I target?
Your margin before ad spend is your break-even ACoS. If it is low, expensive head terms will lose money at launch, so focus on long-tail, use-case and problem keywords where CPCs are lower and conversion is higher.
Can I use competitor brand names as keywords?
Only in PPC product or brand targeting. Do not put competitor brand names in your title, bullets or backend search terms. Amazon’s rules prohibit it and it can get the listing suppressed.
Sources
Get the honest version of your numbers
A free account audit from the person who would actually run your account.