Amazon Search Frequency Rank vs Search Suggestions 2026

Type “kitchen c” into the Amazon search bar and you’ll get a short list of what shoppers type most after those letters. It costs nothing and needs no account. Then there’s Search Frequency Rank, the number Amazon publishes inside Brand Analytics that tells you a term is the 1,412th most searched phrase in the whole marketplace. One of those is free and instant. The other needs a registered brand and arrives a week late. We manage stores doing $200k a month for paying clients, and we use both every week, for two completely different jobs. Here’s which one to trust, and when.

What Amazon search suggestions really are

The dropdown under the Amazon search bar is an autocomplete list. You start typing, Amazon guesses the rest, and it guesses based on what large numbers of shoppers have actually typed recently in that marketplace. That last part is what makes it useful. It isn’t a list of products Amazon wants to sell you. It’s a list of phrases real buyers have used.

We call the way we mine it the A-Z search method, and it’s one of the first things we teach a new client. You take a broad one-word category, something like “kitchen,” “pet,” or “office,” then add a single letter after it. Search “kitchen a,” then “kitchen b,” then “kitchen c,” and read what Amazon fills in each time. Twenty-six passes over one category gives you a long list of phrases that shoppers are typing in volume right now.

A real example from our own research sessions: “kitchen c” returns “kitchen curtains” near the top, which is too generic and too competitive for a new seller, so we move down the list. A few rows lower sits “kitchen cabinet organizer.” That one has a clear product shape, a clear buyer, and it’s specific enough to build a listing around. It became a candidate, and we ran it through Helium 10 Xray the same afternoon.

What Xray said matters just as much as what the dropdown said. Average monthly revenue in that niche was strong and the average selling price sat around $28, which is a good window for a private label product. Then we hit the number that killed it: average review count above 5,000. That market is settled. A new seller walking into it spends a year buying attention from brands with a five-year head start.

That result is the whole point of the method, and beginners usually read it backwards. Here’s what the search suggestion bar can and cannot do:

1. It tells you a phrase is popular enough that Amazon bothers to suggest it.
2. It tells you the exact wording shoppers use, including the messy phrasing no keyword tool would invent.
3. It surfaces new and seasonal demand quickly, because it reflects recent typing behaviour.
4. It does not tell you how popular. Position 1 and position 8 in that dropdown could be ten times apart or nearly identical.
5. It does not tell you whether the market is winnable, who owns it, or what a click costs.
6. It is influenced by your own recent searching, so a dropdown on your logged-in account is not a clean sample.

The A-Z method isn’t built to find a perfect product on the first pass. It’s built to generate 10, 20, or 30 candidates fast so you have something real to validate. Treat it as an idea machine, not an answer.

One more thing the dropdown does better than any paid tool: it moves fast. Because it reflects what people have been typing recently, a seasonal phrase or a new product trend shows up there while the reported data is still catching up. If a category suddenly starts filling with a modifier nobody used last year, autocomplete shows it before your keyword tool has a volume estimate for it. That’s a genuine edge, and it’s the reason we still run the A-Z pass on established accounts, not just new ones.

What Search Frequency Rank really is

Search Frequency Rank, or SFR, lives in Brand Analytics. In Seller Central you open the main menu, go to Brands, then Brand Analytics, make sure the Search Analytics tab is selected, and open the Amazon Search Terms report. Every search term in that report carries a rank. Rank 1 is the single most searched phrase in that marketplace for that period. Rank 300,000 is a phrase almost nobody types.

The first thing to understand is that SFR is a rank, not a count. Amazon does not publish how many searches rank 1 received, and it never has. Every “search volume” number you see in Helium 10, Jungle Scout, or any other tool is that tool’s own estimate, usually modelled from SFR plus its own panel data. The tools are open about this if you read their documentation. SFR is the primary source; the volume figure is a derived guess sitting on top of it.

That matters for how you read the gaps. Because SFR is ordinal, the distance between ranks is not even. The difference between rank 200 and rank 1,200 is enormous in real search traffic. The difference between rank 180,000 and rank 181,000 is close to nothing. Sellers who sort a keyword list by SFR and draw a line at “top 50,000” are usually drawing it somewhere arbitrary, because the curve is steep at the top and almost flat in the tail.

The report gives you more than the rank. For each search term you also get the three most-clicked ASINs, each with its click share and its conversion share. Click share is the slice of all clicks on that search term that went to that product. Conversion share is the slice of all purchases. Those two columns are where the real reading happens, and we’ll come back to them, because a term where one ASIN holds a huge conversion share is a very different opportunity from a term where three products split it evenly.

Two limits worth knowing before you build a strategy on it. SFR requires Brand Registry, so a seller who hasn’t filed a trademark yet cannot see it at all. And the data arrives after the period closes, on weekly, monthly, quarterly, and annual views, so it’s a rear-view mirror. Accurate, but describing a week that already happened.

There’s a practical consequence to that gating which nobody warns beginners about. Brand Registry needs a trademark, and a trademark takes months to move through the USPTO, so the seller who most needs demand data to pick a product is the one guaranteed not to have it. That’s not a flaw in the report. It’s just the order things happen in, and it’s why the free methods in this post carry the whole load for your first launch and the paid data only arrives in time to help with your second.

It also explains a confusion we field almost every month. A client compares the volume number their keyword tool shows against a competitor’s tool and finds the two disagree by thousands of searches, then asks which one is right. Neither, in the sense they mean. Both are models. The rank underneath them is the same public figure, so if two tools disagree on volume but agree on ordering, the ordering is the part you can trust.

The difference that decides which one to use

Search suggestions answer “what do people type?” SFR answers “how much does the market care?” Those are different questions, and most of the arguing about which tool is better comes from people using one to answer the other’s question.

Think about where you are. If you have no product, no brand, and no registered trademark, SFR is not available to you, so the comparison is theoretical. Your entire keyword world is search suggestions, competitor titles, and whatever a keyword tool estimates. That’s fine. Every launch we’ve run started there.

If you already sell and you’re registered, the picture flips. Now you have a rank for every term, click share for the products beating you, and conversion share telling you whether those clicks turned into money. At that point, going back to guessing from autocomplete for a ranking decision would be strange.

Here’s the split we use internally, and it’s close to a rule:

• Generating candidates: search suggestions win. They’re free, instant, and they surface phrasing a tool would never produce.
• Confirming a main keyword: both, and neither alone. You need the SERP check described below.
• Prioritising which terms to chase: SFR wins, because rank is an actual measurement of relative demand.
• Deciding PPC budget allocation: SFR plus the click and conversion share columns, never autocomplete.
• Catching a new trend: search suggestions win on speed, because SFR reports on a closed period.
• Judging whether a keyword is worth the fight: neither. That takes competitor data, which is the Xray step.

Neither number tells you whether you can win. Both tell you demand exists. We learned that expensively on a client account we managed for a full year. The product was a wood-therapy massage tool, and the keyword “lymphatic drainage massage tool” grew well the whole time. Demand was real, the term ranked, our ads ran on it. The product still failed, because the item was generic and five competitors sold the identical thing $3 cheaper. Keyword demand was never the problem. A keyword tells you people are shopping. It says nothing about whether they’d pick you.

The workflow we actually run

Every client keyword file we build follows the same order. Suggestions first because they cost nothing, then validation, then SFR to prioritise. Running it in the other direction means paying for tool credits to sort a list you haven’t earned yet.

Step 1: harvest with the A-Z method. Broad category word plus one letter, twenty-six passes, write down everything that looks like a product a person would buy. Half an hour, no cost. You’ll finish with 30 to 60 raw phrases.

Step 2: read competitor titles. Open the top listings in the niche and look at the very beginning of each title. If a top product opens with “Drawer divider for kitchen,” that phrase is a candidate. Now look down the page. If several competitors all lead with the same core term, say “drawer dividers,” you’ve probably found the main keyword for that product.

Step 3: search your candidate and check the results page. This is the step beginners skip and it’s the one that saves the money. Take the term and search it on Amazon as a buyer. Does the results page fill with the exact product design you plan to sell? If you’re building adjustable bamboo dividers and the page comes back with plastic trays, wire baskets, and drawer liners, that is not your main keyword no matter how high it ranks. Go back to the list, take the next most relevant term, and repeat until the page matches your product.

Step 4: sort by search volume in a keyword tool. Feed the phrase into Magnet or an equivalent and sort by volume. The highest-volume term is often too broad to be useful, which is why step 3 comes before this one, not after.

Step 5: bring in SFR to rank the shortlist. Once you’re registered and selling, pull the Amazon Search Terms report and find your shortlisted phrases. Now you’re sorting real relative demand instead of an estimate. Sort your terms by rank and split them into the tiers you’ll bid on.

Step 6: check indexing on everything you kept. Take your list to the Helium 10 Index Checker, paste in your ASIN, paste in the keywords, and run it. Ignore most of the columns and read the Cumulative one. A checkmark means Amazon can find you for that phrase. A dash means it cannot, so no matter how hard you bid, organic search will never return you for it. Fix a dash by working the phrase into your bullets naturally or adding it to backend search terms, wait about 15 to 20 minutes for Amazon to refresh, then re-run the check to confirm.

Step 7: re-run the whole file on a cadence. Weekly through month 1, then quarterly. On the vegetable chopper launch we reviewed performance every week for the first 30 days, which is four cycles, adjusting title keywords, bullet phrasing, and image order each time. After that we moved to a review every three months. Once a listing is ranking and converting, weekly tweaking adds noise and risks killing whatever is working.

Where sellers get this wrong

The most common mistake is treating dropdown position as a volume ranking. It roughly correlates with popularity, but Amazon personalises suggestions based on your own recent searches, so after an hour of researching drawer organisers your dropdown is no longer a neutral sample of what America types. It’s a sample of what America types, bent toward what you’ve been typing. Use a clean browser session if you’re going to make decisions from it.

The second mistake is chasing the lowest SFR numbers available. Rank 400 looks like the prize, and it is the prize for someone. A term at rank 400 in Home and Kitchen is a term with established brands, heavy ad spend, and an entrenched top three holding most of the click share. Our short, mid, and long tail rule exists for exactly this. Start research in the mid tail, the two-word phrases like “leather wallet” that carry a product identifier. Launch on mid and long tail only. A shopper typing “wallet” might want a women’s purse, a crypto wallet, or a passport holder, so you’ll buy thousands of clicks and watch conversion fall through the floor. You only move up to the short tail after your mid and long tail campaigns are profitable and you’ve built real sales history.

The third mistake is treating either number as permission to enter a niche. On the client launch that worked best for us, the keyword research was the easy part. The filter that made it profitable was a six-point niche checklist applied before a single unit was sourced: average star rating of top sellers at 4.5 or above, most first-page sellers on FBA rather than FBM, no Amazon or single-brand dominance on the main keyword, average first-page review count under 600, more than 40 sellers sitting under 200 reviews each and still doing $10k a month or more, and fewer than 3 variations on the product. That launch landed a 30% net margin with a 15% ACoS in month 1. The keyword told us nobody would find the product without it. The checklist told us we could win.

The fourth mistake is forgetting to check indexing at all. We’ve audited listings where a client was bidding on a phrase for months that their listing was never indexed for. The ads ran, the clicks came from a different matched term, and the reporting made it look like the keyword was working. That’s a slow, quiet leak.

Search suggestions vs SFR, side by side

Neither is a substitute for the other, and the table below is the version we hand to clients when they ask why we use both.

The column that earns SFR its place is the competitor one. Search suggestions tell you a phrase exists. The Search Terms report tells you which three products currently own the clicks on it and what share of the purchases each one takes. A term where one ASIN holds most of the click share and most of the conversion share is a term with a defending champion. A term where the click share is spread and nobody converts well is a term where shoppers are looking and not finding what they want, which is a much better door to walk through.

The report that answers what SFR can’t

SFR tells you a term is big. It doesn’t tell you how your own listing performs on it. For that, stay inside Brand Analytics and open the Search Query Performance report, which is also Brand Registry only.

That report breaks the shopper journey into four stages for every search query: impressions, clicks, add to carts, and purchases. For each stage you get three columns. Total Count is the whole market on that term across every ASIN. Brand Count is your slice. Brand Share is your slice as a percentage. If a column name is unclear, the small question mark icon at the top of it gives Amazon’s own definition.

Two calculations turn that into decisions. Your conversion rate on a term is your brand purchases divided by your brand clicks. The market’s conversion rate is total purchases divided by total clicks. Compare them and the next move usually picks itself.

• High conversion, low impression share: your listing converts better than the market but few shoppers see you. Bid harder on that term. You’re leaving money on the table.
• Conversion well below market average: more ad spend makes the leak bigger. Fix the page first, meaning images, copy, and price, then come back to bidding.
• Brand share at or near 0% on a relevant term: shoppers use that phrase to find products like yours and never see you. Add it to backend search terms, confirm it indexes, then open a campaign on it.

This is also where the click and conversion share columns connect to something physical. On a sports and fitness launch we ran, the client bundled the core product with a hand band, a wrist band, a resistance band, and a small carry bag, and put all four items in the main image rather than burying them in the title. Priced 20% above the category average. Month-1 ACoS came in around 25% and TACoS around 13%, profitable from the first 30 days, and the listing held first page from launch. A shopper scanning the results page saw four items in our thumbnail against one in everyone else’s and clicked ours at a higher price. Keyword work got us onto the page. The main image won the click share once we were there.

When keyword work is worth outsourcing

Everything above is doable on your own. The A-Z method costs nothing, Brand Analytics is free once you’re registered, and a Helium 10 subscription is a normal cost of doing business. What it costs is judgement, and judgement is the part that takes a few launches to build.

Across 47 client onboardings since 2022, the pattern we see is consistent. Beginners can gather keywords fine. Where it goes wrong is the relevancy call in step 3, reading the results page and deciding honestly whether it matches the product they intend to sell, and the discipline to walk away from a niche that passes the keyword test and fails the competition test. Our Product Research service runs the full cycle, including the niche checklist, supplier vetting, and a listing outline, so the keyword file arrives attached to a product we’d be willing to launch ourselves. If you’d rather build the skill, the same sequence is laid out step by step in our product research process and the keyword mechanics go deeper in our keyword research walkthrough.

Frequently asked questions

What is Amazon Search Frequency Rank?

Search Frequency Rank is the number Amazon assigns each search term inside the Brand Analytics Amazon Search Terms report, showing how popular that term is relative to every other search on the marketplace. Rank 1 is the most searched phrase. It’s a ranking, not a search count, and Amazon has never published the underlying volume.

Is a lower or higher Search Frequency Rank better?

Lower is more searched. Rank 500 gets far more searches than rank 50,000. Whether lower is better for you is a separate question, because low-rank terms carry the heaviest competition and the highest cost per click. Most new launches do better on mid and long tail phrases further down the rank list.

Can I see Search Frequency Rank without Brand Registry?

No. The Amazon Search Terms report is Brand Registry only, so you need an approved trademark and an enrolled brand. Until then, use search suggestions, competitor titles, and third-party tool estimates, and remember those estimates are models built on top of SFR rather than a direct read of it.

Are Amazon search suggestions ranked by search volume?

Roughly, and not reliably. The dropdown reflects what shoppers commonly type, so the top entries are generally more popular than the bottom ones, but Amazon doesn’t publish the ordering logic and it personalises results based on your own recent searches. Use a clean browser session, and treat the list as a source of candidates rather than a ranking.

Should I use search suggestions or a keyword tool?

Both, in that order. Suggestions give you the raw phrases shoppers use, including wording no tool would generate. A keyword tool then puts estimated volume against those phrases so you can sort them. Skipping the suggestion step means your keyword list only contains what the tool already knew about.

How often should I check Search Frequency Rank?

Weekly for the first month after launch, then quarterly. During launch you’re watching whether your terms are moving and whether your click share is building. After that, quarterly catches seasonal shifts without tempting you to tweak a listing that’s already working.

The bottom line

Search suggestions and Search Frequency Rank aren’t competing tools, and picking one is the wrong move. Suggestions are where keyword work starts, free and open to anyone, and they’re the fastest way to turn a blank page into 30 real candidates. SFR is where keyword work gets serious, once you have a brand, because a rank is a measurement and an estimate is not. Run suggestions to generate, run the results-page check to filter, run SFR to prioritise, and run the Index Checker to confirm Amazon can actually find you. Neither number will tell you whether you can win a niche. That’s a competitor question, and it gets answered with the review counts, the star ratings, and the seller mix on page 1, before you buy a single unit.