Amazon PPC Keyword Research, Done Properly

What this actually is, and what it isn't

Keyword research for Amazon PPC is the process of building a list of search terms worth bidding on, ranked by how likely they are to convert at a price you can afford. That's different from SEO keyword research, which cares about ranking organically over months. PPC keyword research cares about today's auction: what does this term cost per click, and does it turn into a sale often enough to justify that cost.

The confusion between the two is where most of the wasted spend starts. A term can be excellent for your listing's backend search terms and terrible as a PPC target, because it's high-volume but low-intent — lots of browsers, few buyers. Amazon PPC keyword research has to separate those two questions from the start.

The four real sources — and what each one is actually telling you

There are only four places keyword ideas legitimately come from. Everything else is a repackaging of one of these.

• Your own Search Term Report. This is the only source that shows what actually happened on your account — real clicks, real spend, real orders, against the exact phrase a shopper typed. It's the highest-trust source because it's not modeled, it's observed. Its limit: it only shows terms you've already been shown for, so it can't surface new opportunities on its own.
• Reverse ASIN lookup. Pull the search terms a competitor's listing ranks and gets found for, using a tool connected to Amazon Brand Analytics. Good for finding gaps in your own targeting fast. The data is still an estimate layered on top of Amazon's own aggregated numbers, so treat volume figures as directional, not exact.
• Amazon's own autocomplete. Type a seed term into the Amazon search bar and read what it suggests. This is real shopper behavior, live, for free — no modeled estimate involved. Its weakness is scale: you can't bulk-export the whole category this way, and it tells you nothing about cost or conversion.
• Amazon Brand Analytics top search terms. If you have brand registry, this shows the highest-volume terms for your category with the ASINs currently winning them. Useful for prioritizing which terms are worth the fight before you spend a dollar testing.

Third-party tools like Helium 10 or Jungle Scout build modeled search-volume estimates on top of these sources — that's genuinely useful for scale and speed, and worth paying for if you're doing this manually across a large catalog. Just remember it's a model of demand, not a receipt of it. Your own Search Term Report is the receipt.

A worked example: turning a search term report into decisions

Say you've been running broad match on "stainless steel water bottle" for three weeks and pull the Search Term Report. You'll typically see a spread like this: some terms with clicks and no orders, some with a handful of orders at a reasonable cost, and a long tail of one-click terms that haven't spent enough to mean anything yet.

For example — a term that racked up 40 clicks at a $0.75 average CPC and zero orders has cost you $30 for nothing. That's usually enough data to negate it with confidence. A term with 12 clicks and 3 orders at the same CPC has cost you $9 per sale on a product with real margin — that's a promote, worth breaking into its own exact-match campaign with a higher bid, because you now know it converts. A term with 4 clicks and 0 orders hasn't told you anything yet either way; negating it this early is a guess, not a decision.

The rule underneath all of it: don't act on a term until it's spent enough to be meaningful relative to your typical cost per sale. Acting too early on too little data is the single most common way sellers turn a keyword list into a mess of tiny, unreadable campaigns.

Match type changes what you're actually testing

Broad and phrase match are discovery tools — they exist to surface search terms you didn't think of. Exact match is a harvesting tool — it exists to control cost and bid aggressively on a term you've already proven converts. Running exact match as your only research method means you'll only ever find variations of what you already knew to type in.

The stages below are the same regardless of which tool or spreadsheet you use to manage them.

Where sellers get this wrong — including mistakes worth admitting

The most expensive mistake is treating search volume as relevance. A high-volume term that isn't specific to your product will burn spend on clicks from shoppers who wanted something else — the classic case is a broad category word that looks attractive on a report but describes a different product entirely.

• Negating on too little data. Covered above, and worth repeating because it's the single most common error we see when reviewing accounts that were previously self-managed.
• Never revisiting the list. A keyword list built at launch and never rebuilt goes stale within a season. Search behavior shifts, competitors move in on your terms, and last year's golden keyword becomes this year's break-even one.
• Ignoring internal cannibalization. Running the same term on broad, phrase, and exact simultaneously without harvesting negatives between them means you're bidding against yourself, and Amazon's auction will happily take the money.
• Copying a competitor's front-end terms verbatim. What ranks a competitor organically isn't automatically a good PPC bid target — their conversion rate at that price point may be nothing like yours.

When the research says something you don't like

Sometimes the honest answer is that a keyword you believed in isn't working, and the fix isn't a better bid — it's dropping the term. If a term has real spend behind it and still isn't converting, more budget won't fix it; the term is telling you the shopper's intent doesn't match your product, no matter how relevant it looked on paper. Pull it, and put the budget behind a term that's already proven itself in your own data.

This is also where a rule worth stealing comes in, whichever way you manage the account: every keyword change should carry the evidence behind it, a measurement plan for how you'll know it worked, and a rollback trigger for when it doesn't. That discipline is exactly what we run inside Dr. PPC — the evidence, the measurement plan, and the rollback trigger travel with every proposed change, whether a human clicks approve or the system runs it inside agreed guardrails.

Dr. PPC is priced at $300 a month plus 3% of ad spend, capped, month-to-month, with the first 30 days free — and Orbit, the full software suite behind the keyword and search-term tracking, is included at no extra charge rather than sold as a separate subscription. It's built and run by Full Circle, which has managed more than $500M in revenue across 100+ brands, for sellers who'd rather have the keyword decisions made and evidenced than make every one of them by hand.