Amazon competitor analysis is the practice of working out why shoppers choose another product over
yours, using evidence rather than impression. Done well it produces two or three decisions. Done the
usual way it produces a spreadsheet nobody reopens.
The mistake almost everyone makes is choosing the wrong competitors. The brands you think about are
rarely the ones taking your sales. Your real competitors are the ASINs that appear alongside you on
the specific search terms where you lose share, and that list is frequently unrecognizable to the
person running the business.
Amazon tells you who they are, free, if you have Brand Registry. Start there rather than with a tool.
Step 1: identify the actual competitors
Three sources, in order of usefulness.
Search Query Performance in Brand Analytics. Shows your funnel and your share per search term. Find
the terms where your share drops and note them. Our guide to
Amazon Brand Analytics covers
reading the report properly.
Top Search Terms, also in Brand Analytics, which names the top clicked ASINs per term. This is the
competitor list. Not the brands you admire. The ASINs winning clicks on the queries you care about.
The search results themselves, in a private window, for your main terms. What ranks above you
organically, and what is advertising against you.
I’d spend most of the effort here. In our experience an analysis aimed at the wrong three competitors is
worse than none, because it produces confident changes in the wrong direction.
The three questions that make the whole thing worthwhile
Before you open anything, decide what you are trying to learn. In our experience an analysis without a
question attached produces a document, and an analysis with one produces a change.
Question 1: why are we losing this specific search term? The narrowest and most useful framing. It
has a checkable answer and a specific fix, and it is the one I’d start with on any account.
Question 2: what does the winner on this term have that we do not? Not “what are they doing well”,
which invites a list. What do they have. A badge, a lower price, 4,000 more reviews, a bundle, faster
delivery. Concrete things you can either acquire or work around.
Question 3: what would we have to accept to compete here? The question nobody asks, and the one that
prevents wasted years. Sometimes the honest answer is that competing on a term means margin you cannot
sustain, and the correct decision is to concede it and win somewhere narrower.
I’ve found the third question saves more money than the first two make. Conceding a term deliberately is
a strategy. Losing it slowly while spending on it is not.
Where the free data beats the paid tools
Worth being direct about, because this category is heavily sold.
Brand Analytics gives you share data no third-party tool has, because it comes from Amazon’s own
records rather than from inference. Nothing you can buy tells you your click share on a specific query.
That single advantage makes the free reports more valuable than most subscriptions for this particular
job.
Where paid tools genuinely help: tracking rank over time without manual checking, monitoring
competitor price changes, and finding keywords you have no data on because you do not rank for them yet.
All real, all narrower than the marketing suggests.
Where they mislead: revenue estimates, inferred ad spend, and anything presented as a competitor’s
actual numbers. I’d use tools for monitoring and Amazon’s own data for deciding.
Step 2: measure six things, not twenty
The temptation is to build a matrix of everything, scoring 20 attributes across 10 competitors until the spreadsheet is impressive and unusable. I’d resist it. These six move decisions and the rest pad the document.
Price, including delivery. The number the shopper actually compares. A cheaper item with slower
delivery is not cheaper to a Prime buyer.
Main image. The single largest driver of click share in search results. Compare yours side by side at
thumbnail size, which is how it is actually seen.
Review count and rating. Both matter, and count matters more than sellers expect at low volumes: the gap between 40 reviews and 400 changes click behavior far more than the gap between 4.3 and 4.5 stars does.
Read the negative reviews specifically, because a competitor’s one-star reviews are a list of things
your product could do better and your listing could say.
Title and first bullet. What they lead with tells you what they believe the buying decision turns
on. Frequently they are right and you have led with something else.
Delivery promise and buy box. Who holds it, and whether they are FBA. Losing to a competitor with
faster delivery is a fulfillment finding rather than a listing one.
Variation structure. How many options sit under one parent. A competitor with 8 children on one
listing accumulates reviews across all of them, which is a structural advantage rather than a marketing
one, and it is invisible unless you look.
That is it. Everything else is interesting and does not change what you do on Monday.
Step 3: read their reviews properly
The highest-value hour in the whole exercise, and the one most often skipped because it is tedious.
Sort their reviews to the negative ones and read 50 of them. Budget 45 minutes. Not the summary, not the AI-generated review highlights. The actual text.
You are looking for three things:
Recurring complaints, which are your product opportunity if yours does not share them, and your
warning if it does.
Language shoppers use, which is frequently not the language in anyone’s listing. This is the best
free keyword and copy research available and it costs nothing but attention.
Expectations the listing created and the product failed. These tell you what to promise carefully in
your own copy, since the same gap will produce the same complaints for you.
When I have done this properly, it has changed the product roadmap more often than the listing. In our
experience that is the point at which competitor analysis stops being a marketing exercise.
Step 4: decide what you are actually competing on
The part that turns analysis into strategy, and the part usually missing.
You cannot win on everything, and trying produces a product that is second best at each. Pick one:
Price. Viable if your costs genuinely support it. Ruinous if you are matching somebody with better
manufacturing economics, and I’d check that before starting a price war you cannot finish.
Quality or specification. Needs to be visible in images and stated in the first bullet, or it does
not exist as far as the shopper is concerned.
A specific use case. Frequently the strongest position for a smaller brand, and in my experience the fastest to show results, often within 2 to 3 months. Being obviously right
for a narrower buyer beats being adequate for everyone.
Bundle or configuration. Changing what one purchase includes sidesteps direct comparison entirely.
Service and trust. Slower to build, harder to copy, and it shows up in reviews rather than in the
listing.
What to ignore
Worth saying, because competitor analysis attracts a lot of activity that feels productive.
Estimated revenue figures from tools. They are modeled from rank and they carry wide error bars.
Useful for ranking competitors roughly, misleading to the dollar, and I’d never make a purchasing
decision on one.
Their advertising spend. You cannot see it. Anything presenting it as fact is inferring, and in our experience the inference is usually wrong by a wide margin.
Their whole catalog. You compete ASIN against ASIN on specific queries, not brand against brand.
Copying their listing. It optimizes you into a slightly worse version of them. Read their reviews
instead, since that is where the gap they have not closed is described by their own customers.
A cadence that survives contact with a normal week
Quarterly is enough for most catalogs, and monthly for a fast-moving category. When I have set this up for clients, quarterly held and monthly quietly lapsed.
- Pull your losing search terms from Search Query Performance.
- Identify the top 3 ASINs on each from Top Search Terms.
- Score them on the six measures. One row each, roughly 30 minutes for 9 ASINs.
- Read 50 negative reviews across the set.
- Write down 3 decisions. Not 20 observations.
- Diarize a check on whether those decisions moved your share.
Step 5 is where this usually fails. An analysis that ends in observations ends. One that ends in
three decisions with a date attached becomes a habit.
If you would rather have this run properly on a schedule than intended quarterly, that sits inside our
account management service. We are an
Amazon Ads partner and an Amazon SPN Verified Partner.
FAQ
How do I find my real competitors on Amazon?
Through Brand Analytics rather than intuition. Search Query Performance shows the terms where your share
drops, and Top Search Terms names the ASINs winning clicks on those terms. That ASIN list, not the brands
you think about, is your competitor set.
What should I measure in an Amazon competitor analysis?
Six things: price including delivery, main image at thumbnail size, review count and rating, title and
first bullet, delivery promise and buy box holder, and variation structure. Everything beyond those is
interesting without changing decisions.
Why read competitors’ negative reviews?
Because they list what their product fails to do, in the words shoppers actually use. That gives you a
product opportunity, free copy and keyword research, and a warning about expectations that a listing can
create and a product then fails to meet.
Are competitor revenue estimates from tools accurate?
No. They are modeled from sales rank and carry wide error bars, so they are useful for ranking
competitors roughly and misleading to the dollar. No purchasing decision should rest on one.
Should I copy a competitor’s listing?
No. Copying optimizes you into a slightly worse version of them. Reading their reviews is more valuable,
because that is where their own customers describe the gap they have not closed.
How often should I do competitor analysis?
Quarterly for most catalogs and monthly in a fast-moving category. The frequency matters less than
finishing with three written decisions and a date to check them, rather than with a page of
observations.
Last updated: August 31, 2026. Brand Analytics report names and availability change and differ by
marketplace, and third-party estimates of competitor revenue are modeled rather than measured. Seller
Central carries what is enabled for your account.
