Amazon new product research targeting means defining exactly which shoppers you are building a product for, and checking that enough of them search, click and buy, before you place a purchase order. It combines search term data, competitor review analysis and small advertising tests, so the product, listing and price are aimed at people who already convert in that niche.
Many failed launches come from research that answered “is there demand?” but never “whose demand?”. A niche can show strong monthly searches while being dominated by buyers who want a feature you do not offer.
Who exactly is the buyer, and how do you find out?
Start with search terms, not product ideas. Shoppers describe their needs in the search bar. Use Amazon’s own data where you can: Brand Analytics Top Search Terms (if you have a registered brand) and Product Opportunity Explorer in Seller Central, which groups search terms into niches and shows search volume, click share of top products and return rates.
Read the search terms in a niche as a list of intentions. “Yoga mat thick for knees” and “travel yoga mat foldable” are two different buyers. One is older or recovering from injury; the other travels and cares about weight. They want different products, different images and different prices.
- Group search terms by the problem they describe, not by the product noun
- Note which modifiers appear often: size, material, age, use case, gift
- Check Product Opportunity Explorer for search volume growth over 90 and 360 days
- Look at the return rate column; high returns often mean the products miss an expectation
Mining competitor reviews for targeting clues
Reviews are the most honest market research on Amazon. Pick the five to ten top-selling ASINs for your chosen search term cluster and read the three- and four-star reviews first. Five-star reviews say little; one-star reviews often describe shipping problems. The middle ratings describe what nearly worked.
Write each complaint and each praise into a sheet and count them. Tools like Helium 10 Review Insights or simple export-and-tag in a spreadsheet both work. You are looking for patterns that define a sub-audience.
For example, if many reviewers of a pet water fountain say “my older cat will not drink from it because it is too noisy”, the targeting insight is “owners of anxious or senior cats who need a quiet pump”. That becomes your product spec, your main image claim and your keyword focus.
How big does the audience need to be?
There is no universal number, because it depends on price and margin. A simple way to size it is to estimate monthly units for the top ten listings in your cluster using a tool such as Jungle Scout or Keepa sales rank history, then ask what share you could realistically capture in year one.
Here is an example calculation, clearly labelled as an illustration: if the top ten listings in a sub-niche sell roughly 6,000 units a month in total and you can realistically capture 5 percent within twelve months, you are planning around 300 units a month. Multiply by your expected margin per unit after fees and ad costs. If that number does not justify the inventory risk, the audience is too small or the margin is too thin.
Can you test targeting before buying inventory?
Yes, partly. Several methods let you check interest without full stock.
- Run Sponsored Products on a similar existing ASIN you already sell, targeting the new search terms, to see click-through rate
- Build a landing page with the planned images and run a small Meta or Google ads test to measure clicks and sign-ups
- Order a small first batch (for example 200 to 500 units) instead of a full container
- Use Amazon Vine after launch to get early reviews that tell you if the targeting matched expectations
None of these is perfect, but each reduces the chance of ordering 3,000 units of a product aimed at the wrong person.
Translating targeting into the listing and ads
Once you know the buyer, every element of the listing should speak to them. The title leads with the term that matches their main need. The main image shows the differentiating feature if Amazon’s image rules allow. Bullets answer the objections you found in competitor reviews, in the order of how often they appeared.
On the advertising side, launch with exact and phrase match campaigns on the search terms from your cluster, plus product targeting on the competitor ASINs whose reviews you mined. Those shoppers have already shown they want the product category and may be open to an alternative that fixes the complaint.
Signals your targeting is off
- High click-through rate but low conversion on your core terms
- Returns citing “not as expected” or “too small/large”
- Search Query Performance shows purchases coming from terms you did not plan for
Should you target a broad niche or a narrow one?
Narrow targeting usually wins for new sellers. A broad term like “water bottle” is dominated by brands with thousands of reviews and large ad budgets, and your listing will struggle to earn impressions. A narrower cluster, such as “water bottle with time markers for kids”, has less volume but fewer strong competitors, and shoppers there are clearer about what they want.
The risk of going too narrow is a ceiling on sales. Plan an expansion path from the start: a core ASIN aimed at the narrow buyer, then variations or adjacent products that reach the next group. Product Opportunity Explorer’s related niches view helps map that path.
- Check the review count of the top five listings; very high counts mean a long climb
- Check how many of the top results are Amazon’s own brands or large national brands
- Check price spread; a cluster where everyone sits at the same price leaves little room
- Check search volume trend; a narrow niche that is shrinking is not worth entering
Common research mistakes that skew targeting
Relying on a single tool’s estimates is the most frequent mistake. Sales estimates from third-party tools differ, sometimes by a wide margin. Cross-check with Keepa sales rank history and with Amazon’s own Product Opportunity Explorer before trusting any one number.
Another mistake is reading reviews from only one competitor. A single listing’s complaints may come from its own supplier problem. Patterns that repeat across several listings are the real market signal. Finally, do not ignore seasonality; a garden product researched in May will look far bigger than it is in November.
Where to start this week
Pick one product idea and open Product Opportunity Explorer. Export the search terms in the niche and group them into three to five buyer types. Then read fifty three- and four-star reviews from the top competitors and tag each one.
At the end you should be able to write a single sentence: “This product is for [buyer] who needs [benefit] because current options [problem].” If you cannot write that sentence with evidence from the data, keep researching before placing an order.
Keep reading: Amazon New Product Research · Amazon