Amazon Keyword Research: How to Find Profitable Keywords with the Search Term Report
Last updated: April 13, 2026
Reading time: approx. 14 minutes
Most Amazon sellers start their keyword research in external tools: Helium 10, Jungle Scout, Cerebro. In doing so, they overlook the best data source they already have — and pay for every single day: the Search Term Report.
The Search Term Report shows you the real search terms that real buyers type into Amazon before they click your ads. No estimates, no projections — just real click and purchase data from your own campaigns. In this guide we show you how to analyze the STR systematically, identify profitable keywords, eliminate budget drains and set up a weekly workflow that takes your Amazon keyword research to a new level.
Table of contents
Why the Search Term Report is your most important keyword tool
External keyword tools like Helium 10 or Jungle Scout estimate search volumes based on algorithms and samples. They're useful for a first overview, but they don't show you what works for your product. The Search Term Report, on the other hand, is based on your real advertising data:
What the STR delivers that no external tool can
- Real search terms: Not keywords you target, but the terms customers actually type in.
- Your conversion data: Which search term led to sales for your product — not for the market average.
- Actual CPCs: What you really paid, not what the tool estimates.
- Negative-keyword candidates: Search terms that generate clicks but no conversions — your budget drains.
In short: external tools show you what the market is searching for. The Search Term Report shows you what works for you. Both matter, but the STR data is the foundation for every optimization decision.
Where to find and download the Search Term Report
You'll find the Search Term Report in the Amazon Advertising Console under Reports → Sponsored ads reports → Search term report.
Step 1: Create the report
Go to “Reports” → “Create report” → report type: “Search term.” Choose the time frame (recommended: last 30 days for the weekly analysis, last 60 days for the monthly analysis).
Step 2: Pick the right time frame
Too short (7 days) gives you too little data. Too long (90+ days) waters down trends. The sweet spot: 30 days for operational optimization, 14 days for quick checks at high spend.
Step 3: Download as CSV
Download the report as a CSV and open it in Excel or Google Sheets. There you can filter, sort and run the analysis we describe in the next section.
Pro tip: Set up a recurring report (Reports → Scheduled reports) that automatically emails you the STR every week. That way you'll never forget the analysis.
The key columns and what they mean
The STR contains many columns, but for keyword research only a handful are truly decisive:
Important: Amazon doesn't show the ACoS directly in the STR. You have to calculate it yourself: Spend ÷ Sales × 100. Add a column in your spreadsheet that does this automatically. Without ACoS per search term, the analysis is blind.
Identify profitable keywords: the 4-step process
This is where Amazon keyword research gets concrete. Follow these four steps to filter the truly profitable search terms out of the sea of STR data.
Step 1: Sort by spend (descending)
Your budget concentrates on a few search terms. The top 20 terms by spend often make up 60–80% of your total spend. Start your analysis there — that's where the biggest leverage is.
Filter: Spend > $5 in the chosen time frame (anything below that has too little data).
Step 2: Calculate the ACoS per search term
Compare each search term's ACoS to your break-even ACoS. This creates three categories:
Profitable keywords: ACoS below break-even → scale and isolate
Borderline keywords: ACoS near break-even → adjust bid, watch closely
Unprofitable keywords: ACoS well above break-even → lower the bid or exclude as a negative
Step 3: Check the conversion rate
A search term with high spend and zero conversions is a clear candidate for the negative keywords list. But even the conversion rate among terms with conversions varies widely:
Benchmarks: CR > 10% = excellent, 5–10% = good, 2–5% = acceptable, < 2% with many clicks = check whether the keyword really fits your product.
Step 4: Isolate the winners
Search terms with at least 2–3 conversions and an ACoS below your target are candidates for Search Term Isolation: move them into their own campaign as exact-match keywords and exclude them as negative exact in the source campaign. That way every top performer gets its own bid and budget.
Worked example: STR analysis for a coffee grinder
| Search term | Clicks | Spend | Sales | ACoS | Action |
|---|---|---|---|---|---|
| electric coffee grinder conical burr | 87 | $52 | $380 | 13.7% | Isolate + scale |
| electric coffee grinder | 214 | $171 | $620 | 27.6% | Isolate, optimize bid |
| coffee grinder | 312 | $280 | $710 | 39.4% | Lower bid, watch |
| espresso grind hand mill | 45 | $38 | $0 | ∞ | Negative exact |
| manual wooden coffee grinder | 62 | $49 | $0 | ∞ | Negative exact |
Result: 2 profitable keywords for isolation, 1 borderline keyword to optimize, 2 budget drains for the negative list. In 15 minutes of work, you've laid the groundwork for a noticeably better campaign.
Spot budget drains: negative keywords from the STR
At least as important as finding the winners is excluding the losers. Every search term with significant spend and zero conversions costs you money that could be invested in profitable keywords.
When does a search term become a negative keyword?
- Exclude immediately: Search terms that obviously don't fit your product (e.g., “wooden hand mill” when you sell electric grinders).
- After 20+ clicks with no conversion: The search term fits thematically but doesn't convert. This points to a mismatch between search intent and product.
- After spend > 2× product price with no conversion: If you've spent more on a search term than the product costs, without a single sale, the chance it ever becomes profitable is low.
Enter the identified search terms as negative exact match in the relevant campaign or ad group. Only use negative phrase match if you're sure that an entire phrase will never be relevant — otherwise you accidentally exclude profitable variants too.
Brand Analytics & Search Query Performance as a complement
The Search Term Report only shows your advertising data. For the complete picture, you also need the organic data from Brand Analytics and Search Query Performance (available to brand-registered sellers).
Brand Analytics: top search terms
Shows the most frequently searched terms on Amazon, including the click share and conversion share of the top 3 products. Use it to find new keyword ideas that don't yet appear in your STR — because you're not targeting them yet.
Search Query Performance (SQP)
Shows, for each search term, your organic and ad-driven impressions, clicks and purchases compared to the total market. The decisive added value: you can see which keywords you already rank well for organically (and may rely less on advertising for) and where you're invisible organically (and absolutely need advertising).
Combination: STR + SQP = full transparency
STR says: “Keyword X has a 15% ACoS, super profitable.”
SQP adds: “For keyword X, you rank organically in position 2 — you're paying for clicks you'd probably get organically anyway.”
Takeaway: Lower the bid, invest the budget in keywords where you're weak organically.
The weekly keyword research workflow
Amazon keyword research is not a one-time event but an ongoing process. Here's the workflow we recommend for using the STR most effectively:
Monday: download the last 7 days of the STR
Quick check: are there new search terms with high spend and zero conversions? → Add them as negatives immediately. Are there new search terms with conversions that aren't yet running as their own keywords? → Flag them for isolation.
Wednesday: adjust bids
Based on Monday's analysis: raise bids for profitable keywords, lower them for borderline ones. Move new exact-match keywords from the STR into isolated campaigns.
Friday: monthly comparison (every 4 weeks)
Download the 30-day STR. Compare it to the previous month: which keywords have risen or crashed? Are there seasonal trends? Are your budget layers (brand, generic, competitor) in balance?
This workflow takes about 30–60 minutes per week and makes the difference between reactive optimizing and a proactive keyword strategy.
Work faster: Amazon data right inside the tool
The manual workflow — download the STR, open it in Excel, filter, calculate ACoS — works, but it costs 30–60 minutes every week. With HORAiZON ONE this step disappears: the tool is connected directly to the Amazon Advertising API and automatically pulls search-term data, keyword performance and campaign metrics into a single central dashboard.
Instead of exporting and reconciling data manually, you see at a glance which search terms are profitable, which should be excluded as negatives and where isolation makes sense — including automatic ACoS calculation per search term and campaign. That turns the weekly 60-minute workflow into a 5-minute review.
The 5 most common keyword research mistakes
1. Relying only on external tools
Helium 10 & co. deliver estimates. Your STR delivers facts. Use external tools for the initial research, but steer your campaigns exclusively based on your own data.
2. Never analyzing the STR
An alarming number of sellers let their auto campaigns run for months without ever looking at the Search Term Report. That's like throwing money into a mailbox and hoping something good comes back.
3. Making decisions too early
A search term with 5 clicks and 0 conversions isn't automatically bad — it simply has too little data. Wait for at least 20 clicks before you exclude a keyword as a negative, otherwise you eliminate potential winners.
4. Using negative phrase instead of negative exact
Negative phrase excludes all search terms that contain the phrase. If you set “hand mill” as a negative phrase, you also exclude “electric hand mill” — perhaps your most profitable keyword. Use negative exact match by default.
5. Ignoring the halo effect
Some keywords have a high ACoS but massively boost organic sales. Check the TACoS in parallel: if it falls even though the ACoS of individual keywords is high, the halo effect is working in your favor.
Conclusion: from real searchers to profitable keywords
The best Amazon keyword research isn't based on guesses or external estimates — it's based on the real search data of your customers. The Search Term Report, complemented by Brand Analytics and Search Query Performance, gives you everything you need to find profitable keywords, eliminate budget drains and systematically improve your campaigns.
Your action plan
Today: Download the last 30 days of the STR and add an ACoS column.
This week: Analyze the top 20 search terms by spend, isolate the winners, enter budget drains as negatives.
Starting next week: Launch the weekly workflow (Monday: STR check, Wednesday: adjust bids).
Starting next month: Use SQP data as a complement and compare organic vs. paid performance.
Anyone who follows this process consistently will see a noticeably better campaign performance after 4–8 weeks: a lower ACoS, less wasted budget and a growing pool of proven exact-match keywords that can be scaled in a targeted way.
Frequently asked questions
Identify and steer keywords automatically?
With HORAiZON, AI analyzes your search terms and builds the relevant keywords directly into your listing. More in the Amazon SEO Tool.
Optimize 1 product for free now