Here's a scenario most Amazon sellers have lived through at least once.
You launch a Sponsored Products campaign. You pick keywords that seem reasonable, broad terms that match your product category, and some that tools say have decent search volume. You set a bid, set a budget, and let it run. A week later, you open the search term report and scroll through it.
There's a keyword driving spend. Lots of spending. You click to see the conversion rate. Zero. Not low, zero. The keyword has sent 80 clicks to your product page and produced no sales. You've paid for 80 people to visit your listing and leave, none of them interested enough to buy.
Now multiply that across 400 keywords. Across 15 campaigns. That's the reality of keyword management without a qualification system.
Profitable keywords are not found in keyword databases. They're earned through a systematic process of discovery, measurement, qualification, and refinement, a process built on your own actual buyer data, not estimated traffic numbers. The best source of profitable keywords for your product is always the same: what Amazon's own shoppers actually typed when they bought from you.
In Hector AI's internal analysis of 2,400+ Sponsored Products campaigns in 2025, keywords harvested from auto campaign search term reports and promoted to exact-match manual campaigns generated 43% lower ACoS than campaigns relying on broad keyword targeting alone (Hector AI Internal Data, Jan–Dec 2025). The gap is that large because the qualification step removes the guesswork.
What Actually Makes an Amazon Keyword Profitable
Before looking for keywords, you need a definition of what you're looking for. 'Profitable' is not the same as 'high-volume' or 'relevant'. A profitable keyword meets a specific financial threshold for your product's margin structure.
The 5-Factor Keyword Profitability Framework
Evaluate every candidate keyword against five criteria. A keyword that clears all five is worth bidding on and scaling. One that fails on cost or conversion is worth excluding, regardless of how relevant it sounds.
The order of priority matters. Conversion rate and ACoS are the output metrics; they tell you whether a keyword is actually working. Volume and competition are input signals that help you find candidates. Relevance is the filter that eliminates false positives before you spend on them.
Why High Search Volume Is the Wrong Starting Point
Most sellers begin keyword research by looking for the highest-volume terms in their category. This is understandable; more searches mean more potential impressions. But volume is a measure of market size, not conversion potential.
High-volume terms in competitive categories are also the highest-CPC terms. They're the most expensive keywords to bid on, they carry the lowest conversion rates (because the search intent is broader and less specific), and they're where the most overfunded campaigns compete hardest. For most sellers below a dominant market position, high-volume broad terms generate impressions, clicks, and cost, not profit.
Long-tail, specific keywords, 'stainless steel travel mug with lid 16oz' rather than 'travel mug', convert at 2–5x the rate of broad category terms, with significantly lower CPCs, because the shopper knows exactly what they want. These are the keywords that make a campaign profitable. And the fastest way to find them is through your own data, not a keyword database.
Using Your ASIN to Discover Keyword Opportunities
Your ASIN is the starting point for keyword discovery because it anchors the research to your actual product, not a category average or a competitor's catalogue. Three data sources are available, and each reveals different keyword opportunities.
Auto Campaigns as Keyword Discovery Engines
Auto campaigns are one of the most reliable sources of keyword discovery because they use actual shopper search behavior against your listing. The search terms they generate are based on real interactions with your product rather than estimated search-volume data or third-party projections.
When setting up an auto campaign, Amazon provides four targeting expressions: close match, loose match, substitutes, and complements. Each targets a different type of shopper intent and can reveal distinct keyword opportunities.
Reverse ASIN Lookup: Finding Keywords Your Competitors Rank For
Reverse ASIN lookup is a research technique that identifies which search terms a specific product ASIN is ranking organically or running paid ads on. Third-party reverse ASIN tools estimate keyword relationships using Amazon search result and ranking data, helping advertisers identify terms for which competing products appear organically or in sponsored placements.
The strategic use: enter your top 3–5 competitor ASINs and extract the keywords where they're generating the most organic rank or ad impressions. These keywords are confirmed to be relevant to similar products, which means they're likely relevant to yours. They also show you gaps: keywords your competitors rank for organically that you don't yet target in your campaigns.
Search Query Performance (SQP) Data: The Underused Keyword Source
Amazon's Search Query Performance (SQP) report is available in Brand Analytics for Brand Registry sellers. It shows, at the brand and ASIN level, which search queries are driving impressions, clicks, and conversions for your products, along with your share of total clicks and conversions for each query.
SQP data is particularly valuable for two reasons: it shows you queries where you're already converting without running ads (suggesting those keywords would perform even better with bid support), and it shows you queries where competitors are taking most of the click share (indicating where you're losing traffic you could be competing for). SQP provides aggregated search-query-level performance insights, including impressions, clicks, and purchase share, helping brands understand how they perform relative to the broader market for specific search queries.
The Auto-to-Manual Keyword Harvesting Process
Keyword harvesting is the systematic process of moving proven search terms from your auto campaign's data into targeted manual campaigns. It's the core operational loop that separates profitable Amazon PPC from expensive guesswork.
Reading Your Search Term Report for Conversion Signals
The search term report is accessible from Campaign Manager → Reports → Search Term Report. Download it for a 14–30 day window. Sort by spend, high to low. You're looking for two distinct patterns:
• High spend + zero or low conversions, these are the terms bleeding your budget. Add them as negative keywords immediately.
• Any conversions + ACoS below your target threshold, these are your profitable keywords. Harvest keywords after they demonstrate consistent conversion efficiency and sufficient click volume relative to your category economics.
Don't filter by spend level alone. A keyword that spent $8 and produced two conversions at 14% ACoS is a significant finding; it needs a higher bid and a dedicated campaign, not to be buried in an auto campaign's undifferentiated spend pool.
The Exact Workflow: From Auto Discovery to Manual Precision
Setting Match Types Correctly for Newly Harvested Keywords
Match type determines how closely a shopper's search query must match your keyword before your ad triggers. For newly harvested keywords, the correct sequence is:
• Exact match primarily targets the specified keyword and close variants with similar shopper intent, giving advertisers the highest level of control and relevance among Amazon's match types.
• Phrase match captures searches containing the keyword phrase in sequence, along with close variations and additional terms before or after the phrase.
• Broad match can be valuable for discovering new search terms and expanding keyword coverage. However, once a keyword has demonstrated consistent profitability, advertisers typically move it into exact or phrase match campaigns to gain tighter control over bids, search intent, and performance.
A common mistake: moving harvested keywords into phrase or broad match manual campaigns instead of exact. Broad match on a confirmed profitable keyword dilutes the bid across irrelevant variants, the same problem your auto campaign was already solving. An exact match is what makes the harvesting process work.
Negative Keywords, The Hidden Half of Keyword Profitability
Most sellers focus on keyword work on finding more profitable search terms to bid on. The equally important, and often higher-impact, work is identifying the unprofitable terms and eliminating them. Negative keywords are the mechanism.
What Wasted Spend Looks Like in a Search Term Report
Open any Sponsored Products search term report that has been running for four or more weeks without active negative keyword management. Sort by spend. Scroll through the top 20 terms by cost. In most cases, you will find at least 3–5 terms that have spent meaningful budget, $20, $50, sometimes hundreds, and produced zero conversions.
For a brand spending $5,000 per month on Sponsored Products, 23% waste is $1,150 per month, $13,800 per year, funding traffic that was never going to convert. Negative keywords recover this spend and redirect it to terms that do convert.
Building a Negative Keyword Architecture That Compounds Over Time
Negative keyword management is not a one-time cleanup. It's an ongoing process that gets more precise and more valuable with each iteration. Structure it in three tiers:
• Campaign-level negatives, terms that are irrelevant to your entire brand (competitor brand names, unrelated categories, obvious mismatches). Set once, maintain rarely.
• Ad group-level negatives, terms that are relevant to your brand but not to the specific product in that campaign. A hammock brand needs 'camping chair' as a negative on their hammock ad group, but maybe not at the brand level if they also sell chairs.
• Harvested negatives from search term reports, terms that have spent budget without converting in the past 14–30 days. Add these to every harvesting cycle.
Over 12 months of consistent negative keyword management, a well-structured campaign will have eliminated the majority of its wasted spend and concentrated budget on its highest-converting terms. The compounding effect is significant; each harvesting cycle improves the campaign's efficiency ratio, and the improvements persist forward.
Why Manual Keyword Management Fails as Your Catalogue Grows
The manual keyword harvesting process described above works perfectly, until it doesn't. The breaking point is predictable, and almost every growing Amazon brand hits it.
The Volume Problem: What Happens After 500 Keywords
A seller managing 3 ASINs with 2 auto campaigns and 1 manual campaign per ASIN has roughly 150–200 active keywords to review every two weeks. That's manageable; a few hours of focused work produce meaningful optimisation.
Scale to 10 ASINs with a full campaign structure, auto, broad manual, phrase manual, exact manual, and the active keyword count reaches 800–1,200. The search term report for 10 campaigns over 14 days has 10,000–20,000 rows. The harvestable keywords are buried in those rows, alongside thousands of irrelevant terms. The negatives that need adding are scattered across campaigns.
At this scale, the manual harvesting process takes longer than the harvesting cycle itself. A two-week harvest requires 8–12 hours of focused analysis, leaving only 2 weeks before the next cycle begins. Quality degrades. Profitable keywords get missed. Wasted spend accumulates. This is exactly the problem rule based automation solves executing your harvesting logic across every campaign, every day, without manual intervention. The campaign stagnates despite having all the data it needs to improve.
How Automated Harvesting Rules Close the Gap
Automated keyword harvesting rules replicate the exact logic of the manual process, but execute it continuously, across every campaign, without human review cycles.
The rules are defined once: if a search term in an auto campaign generates 1 or more conversions and an ACoS below [target threshold] over a 14-day window, add it as an exact-match keyword in the designated manual campaign and add it as an exact-match negative in the auto campaign. If a search term has spent above [waste threshold] with zero conversions, add it as a negative across all campaigns.
These conditions evaluate every keyword in every campaign every day, not every two weeks. A profitable keyword surfaces in Monday's auto campaign data and is in an exact-match manual campaign by Tuesday. A wasted term is excluded before it doubles its spend.
A Practical Keyword Qualification Scorecard
Before adding any keyword to a manual campaign, whether discovered through auto campaigns, reverse ASIN lookup, or SQP data, score it against this checklist.
Scoring Your Keywords Before You Bid on Them
A keyword that passes all six criteria goes into your exact-match manual campaign with a bid set at your auto campaign's average CPC for that term, increased by 20–30%.
A keyword that passes 4–5 criteria goes into a phrase-match manual campaign for further qualification and data collection. Phrase match can help expand keyword coverage while maintaining more control than broad match.
A keyword that passes 3 or fewer goes back into your auto campaign's discovery pool or gets added as a negative. Don't force marginal keywords into manual campaigns on optimism.
The discipline of using a consistent qualification framework, rather than adding keywords based on intuition or tool recommendations, is what separates sellers whose keyword portfolios improve every month from those whose campaign performance plateaus.