How to Do Amazon Keyword Research: A Step-by-Step Reverse-ASIN Method

There are many tools and shortcuts for Amazon keyword research, and Amazon keeps changing how search and its assistant work. The method below has held up through those changes because it is built on two things that do not change: what real competitors rank for, and how relevant each phrase is to your product.

Before you start, run the six keyword research prerequisites. You need a defined product, a margin you understand and a list of the problems your product solves, or you will not be able to judge relevancy in step 5.

Step 1: Map the search results page

Search your product-type keyword, for example “water bottle”, with all departments selected. Note what Amazon shows: sponsored placements, organic results, video ads and any assistant answers. Then use the category filters on the left of the results page to find the sub-category that matches your product most closely.

The sub-category matters. An insulated steel bottle competes with different listings than a kids’ plastic bottle, even though both show up for “water bottle”.

Step 2: Choose the right competitor ASINs

Inside the right sub-category, use your research tool to load the full results, then sort by price and sales. Select 8 to 10 ASINs that:

• are organic results, not sponsored placements, so you see what Amazon ranks on merit;
• match your price point, product utility and design;
• include both established leaders and newer listings that are growing fast.

Newer fast-growing listings are often the most useful. They show which keywords a listing can win quickly, which is exactly what you need at launch.

Step 3: Run a reverse-ASIN lookup

Run the selected ASINs through a reverse-ASIN tool such as Helium 10 Cerebro, Jungle Scout, SellerSprite or Data Dive. These tools estimate which search terms each ASIN ranks for, organically and in ads, and how much search volume each term has.

Export the results to a spreadsheet. Keep only the columns you need: the phrase, estimated monthly search volume, number of competing products, and the organic rank of each competitor ASIN. Delete the rest so the sheet stays readable.

Step 4: Filter by volume and clean the list

• Set a volume floor. We often start at 300 to 400 monthly searches in large categories and drop to around 100 in small niches.
• Remove competitor brand names from the listing list. Keep them in a separate tab for PPC brand and product targeting.
• Merge duplicates such as plurals and word-order variants, since Amazon usually treats them as the same search.
• Remove anything you cannot claim honestly, such as features your product does not have.

Step 5: Score every keyword for relevancy

Add a relevancy column next to the phrase and score each keyword by hand. This is the step most sellers skip, and the one that makes the biggest difference.

Sort by relevancy first, then by search volume. In our competitive keyword audit, coverage of relevant keywords explained why a newer listing outsold a household name with five times the reviews.

Step 6: Place keywords in the listing and PPC

Also write two or three natural-language questions a shopper might ask, such as “which water bottle stays cold all day”, and make sure your bullets answer them. That helps with optimizing for Alexa for Shopping, which answers conversational queries in the main search bar.

Step 7: Validate with real data

Tool estimates are a starting point. Real data comes from two places:

• PPC search term reports: after two to four weeks of auto and broad campaigns, you can see which customer searches produce sales. Promote winners to exact match and add them to the listing if they are missing. Our Amazon PPC management runs this loop weekly.
• Search Query Performance: in Brand Analytics, Brand Registry owners can see impressions, clicks, add-to-carts and purchases for the top search queries, with your brand’s share at each step. A query where you get impressions but few clicks usually points to the main image, price or title.

Re-run the full research every quarter, and straight away after a category change, a new competitor or a big rank drop. If organic rank falls while PPC holds, read why launch rank is often rented.

Common Amazon keyword research mistakes

• Using sponsored listings as the competitor set, which shows you what people pay for, not what ranks.
• Sorting by search volume only and ending up with broad terms that never convert.
• Stuffing the title, which now risks an AI rewrite under the 75-character rule.
• Repeating title words in the backend, which wastes the 250-byte limit.
• Never revisiting the list after launch.

Once your list is ready, use it in our four listing optimization strategies. If you would rather hand it over, our Amazon SEO service covers research, copy and tracking.