Every Amazon seller with Brand Registry has access to the Search Query Performance report. Almost none of them read it correctly.
The report’s top-line search volume number gets treated as “the market” — plugged into a spreadsheet, called addressable demand, used to justify a PPC budget. That number is usually wrong. Not fraudulent, just uncorrected: buried inside it is a much smaller set of queries that represent real, buyable demand for your product, sitting next to a much larger amount of noise that doesn’t.
We built a process specifically to separate the two before any metric gets trusted. We call it Query Safari — a hunt through a large, noisy dataset for the queries actually worth chasing.
The problem with raw SQP volume
Amazon’s Search Query Performance data reports total search volume for every query that surfaced your ASIN’s impressions. That total mixes together queries with genuinely different meanings:
- Someone searching for exactly what you sell
- Someone searching for something adjacent — a related use case, a complementary product
- Someone searching for something your product has no real connection to, that Amazon’s broad-match logic pulled you into anyway
- Someone searching for a product category you don’t yet have proof you can win
Treat all four the same, and your “addressable market” number is fiction. We’ve seen accounts where the real, defensible addressable market — after sorting — was roughly 12% of the raw total volume the export reported. That’s not a rounding difference. That’s the difference between a market-sizing slide that’s directionally true and one that’s actively misleading a client.
The four-pool sort
Before we read a single CTR or CVR number, every query gets sorted into one of four pools:
| Pool | What it means | What we do with it |
|---|---|---|
| Core | Direct match to what the product actually is | Primary optimization target — CTR/CVR benchmarking against market only matters here |
| Adjacent | Related use case, complementary product | Expansion candidate — test it, don’t defend it yet |
| Irrelevant | No real buyer-intent connection | Excluded entirely — usually a sign of broad-match overreach |
| Aspirational | Looks attractive by volume, but no proven conversion evidence yet | Tracked, not chased — a future opportunity, not a current one |
Real addressable market = Core + Adjacent only. Everything else gets set aside before any percentage, any benchmark, any budget decision touches the data.
Case study: the hunt that found 21 rare catches
An international multi-ASIN seller came to us questioning whether their PPC spend matched their actual market opportunity. The raw SQP export suggested a large market. After the sort:
- Real addressable market: 101,223 searches — a small fraction of what the raw total implied
- Weighted impression share across that real market: ~2%
- 21 queries were already beating market-average CTR with strong CVR — proven appeal, still under-captured
Those 21 queries are the actual payoff of the hunt: rare, already-validated, and mostly invisible until the noise gets cleared away. The account’s PPC strategy shifted around scaling those specific queries and building a dedicated campaign for the strongest secondary demand cluster — not around the inflated top-line number that started the conversation.
Why “more volume” isn’t the goal
The instinct when you see a query with big raw volume is to chase it. Query Safari exists specifically to interrupt that instinct. A query with 50,000 raw monthly searches that turns out to be 90% Irrelevant/Aspirational is worth less than a Core query with 2,000 searches and proven CTR/CVR. Budget follows the sort, not the headline number.
This connects directly to keyword-level decisions downstream — once a query is confirmed Core or Adjacent, it becomes a candidate for further classification through Leak Sealer, which decides whether that keyword is worth scaling or cutting. Query Safari’s job ends at is this demand real.
FAQ
Isn’t Amazon’s SQP data already accurate?
The data itself is accurate — it’s the interpretation that goes wrong. Amazon reports what queries surfaced your listing; it doesn’t tell you which of those queries represent real, winnable demand for your specific product. That sorting is the seller’s or agency’s job, not the report’s.
How much smaller is the “real” market usually?
It varies by category and how aggressive the account’s broad-match targeting has been, but we’ve seen real addressable market land anywhere from roughly 10–30% of the raw reported total. The 12% figure is from one specific account, not a universal ratio — every account needs its own sort.
What’s the difference between Adjacent and Aspirational?
Adjacent means the query reflects a genuinely related use case or complementary product — expansion territory. Aspirational means the query looks appealing by volume, but the brand has no evidence yet that it can convert on it. Adjacent gets tested; Aspirational gets tracked and revisited later, not chased with budget now.
Does this apply the same way on every Amazon marketplace?
The four-pool logic is marketplace-independent — what changes is the underlying SQP export’s column structure, which varies by locale (US vs. UK vs. FR exports use different headers for the same underlying fields). The sort itself doesn’t change; the mapping to get there does.
Fix Your Ecom is a boutique Amazon PPC and growth agency. Query Safari is one of twelve named processes in our operating system — The Teardown, our full account audit, runs it as a standard step. Book the Account Teardown to see what’s really in your search query data.
