Etsy's Own Search Data: The 3 Numbers That Decide a Keyword

For years the honest answer to “how many people actually search for this on Etsy?” was that nobody outside Etsy knew. Every number you had ever seen was a model. That answer expired in August 2026. Etsy now publishes its own search data inside Shop Manager, free, to every seller — including purchase-based fields drawn from records Etsy holds rather than a third-party model.

Written by Mouhcine El Aboudi at SellerAIReady. Last updated 7 September 2026.

What Etsy actually changed

Etsy released a tool called Marketplace Insights that sits inside the seller dashboard and returns, for a search term you type, the number of searches, the number of competing listings, a conversion-rate band, and a median purchase price range. It also compares your own listing prices against that market median.

That is a structural change, not a feature announcement. Until now, every keyword number circulating in the Etsy seller world came from outside the marketplace: sampled scrapes, autocomplete mining, clickstream panels, and statistical models trained on those inputs. Some of it was good work. None of it could see purchases. Etsy can see purchases, because Etsy processes them.

The practical consequence is that the keyword question changes shape. It stops being “which term has the biggest number?” and becomes “which term does Etsy’s own record say people buy on?” Those two questions frequently have opposite answers, and the rest of this article is mostly about how far apart they can be.

One housekeeping note first, because this site holds itself to it. An earlier article here stated that Etsy publishes no search volume data of its own. That was accurate when it was researched and false by the time it was read. It has been corrected in place, and the reasoning is set out in the piece on whether Etsy is too saturated in 2026.

Where to find it, and what it costs

The path is Shop Manager → Stats → Marketplace insights. There is no separate signup, no paid tier, and no minimum sales requirement. The panel was confirmed present and fully functional in a Morocco-based shop with fifteen lifetime visits and zero orders, which is about as small as a live shop gets. If you have a shop, you have this.

There is one real constraint: lookups are metered. The panel states the allowance in its own words: fifteen terms per week. It is a rolling weekly allowance rather than a lifetime total, confirmed by re-reading the same panel on 4 September 2026 and finding it refilled. Plan a week at a time, and pair your terms so that every lookup answers a comparison rather than a curiosity.

The first number: conversion rate

This is the one that matters, and the one that has never existed outside Etsy. For each search term, Etsy labels a conversion band running from Very low to Very high. It is describing how often that particular search ends in somebody buying something.

No external tool can produce this figure honestly. Conversion requires visibility of completed transactions tied to the search that produced them. Scrapers see listings. Panels see clicks. Only the marketplace sees the checkout. Everything else is inference dressed as measurement.

The band itself repays a closer reading, because Very high is not automatically the term you want and Very low is not automatically fatal. What the label is really measuring, and the trap sitting inside the top band, is unpacked in Etsy conversion rate labels explained.

The second number: searches against competing listings

Volume on its own tells you how loud a room is, not whether anyone can hear you in it. The number worth reading is the ratio of searches to competing listings, and Etsy now hands you both halves of that fraction for the same term at the same moment.

A term with a large search count and a larger listing count is a crowded room. A term with a modest search count and very few competing listings is a conversation. The second one is where a new shop can actually be found, and it is the mechanical reason that broadening a keyword to chase volume so reliably backfires: broad terms are precisely where the hundred million existing listings already sit.

This is also the cleanest rebuttal available to the saturation argument. Demand on Etsy did not collapse. Supply expanded. Those are different problems with opposite solutions, and the ratio is how you tell which one you are actually facing on any given term.

That fraction deserves a threshold rather than a glance. Run across twenty seller terms it produces a four-tier screen that separates a workable niche from a queue, set out in the searches-per-listing competition ratio.

The third number: median purchase price range

For each term Etsy also publishes a median purchase price range, and it displays your own prices beside it. For the term badge reel, that range was USD 9.00 to USD 11.00.

Read carefully, this is a pricing instrument, not a curiosity. It is the difference between an asking price and a selling price. Scraped listing prices tell you what sellers hope to get. A median purchase price range tells you what buyers actually paid. If your price sits far outside that band, you now know it before you launch, not three months later when the listing has views and no orders.

It is also a free monitoring check worth repeating at every checkpoint. Your prices against the market median answers a question you cannot answer from inside your own shop: whether your price sits inside the band buyers are actually paying in. If your listings are getting attention but nothing is converting, this belongs next to the diagnostic sequence in the breakdown of Etsy listing views with no sales.

Two things follow from that median. The first is what the field can and cannot tell you, worked through term by term in what Etsy’s median purchase price actually means. The second is what survives the platform’s cut once you have priced into that band, which is the arithmetic in Etsy fees in 2026 and what you actually keep.

The case that settles the argument

Two real lookups, taken minutes apart, are worth more than any amount of theory.

• badge reel — 215,600 searches, 273,600 competing listings, conversion labelled Very low, median purchase price USD 9.00 to USD 11.00.
• custom badge reel — 8,100 searches, conversion labelled Very high.

The specific term carries roughly twenty-six times less traffic, and it is the one that sells. More competing listings than searches on the broad term is the tell: that is a market where the supply of sellers has outrun the demand of buyers, and where a new listing arrives into a queue rather than a gap.

Now consider how that pair reads through a volume-only lens. The broad term wins by a factor of twenty-six on search volume, and on nothing else. A seller optimising for that number would choose it, write the title around it, buy ads against it, and conclude after two months of impressions without orders that Etsy is saturated. The data did not fail. The metric being optimised was the wrong one.

One pair is an anecdote. The same procedure was later run across twenty terms in the seller-tools niche, and the distribution it produced — ten Very low bands against a single Very high — is published in full in twenty seller keywords measured with Etsy’s own data.

What your keyword tool was actually giving you

This is not an argument that third-party tools are worthless, and it is worth being precise rather than triumphant about it. All three kinds of source hedge their numbers, Etsy included. Etsy’s figures are drawn from a sample of its own records and are presented in bands and ranges rather than exact counts. Nobody is handing you a ledger.

The real distinction is narrower and more durable than “accurate versus inaccurate”: a sample of the marketplace’s own records is not the same class of evidence as a model built on external proxies. Both are estimates. Only one of them is estimating from inside the building.

Tools still earn their place on the work Etsy’s panel does not do: bulk keyword expansion, competitor and shop-level tracking, tag audits across a whole catalogue, cross-marketplace comparison, and history over time. Metered lookups will never cover a two-hundred-listing audit. If you are weighing a subscription, the comparison is set out in the piece on eRank versus EverBee for Etsy sellers. The wider question of which paid tools are worth their price at all, with every plan checked against the vendor’s own page rather than an affiliate summary, is in the best AI tools for Etsy sellers.

There is a harder question underneath that comparison: how far the estimates themselves can be trusted before you act on them. Sellers who checked extension figures against their own dashboards have published the gaps, and the verified cases plus a ten-minute verification routine are collected in does EverBee actually work.

The honest hierarchy of keyword metrics

Ranked strictly by how close the source sits to the actual purchase, and nothing else:

1. Conversion rate for the term. Only the marketplace can publish it, because only the marketplace sees checkout.
2. Searches against competing listings. Both halves of the fraction, from the same source, for the same term.
3. Median purchase price range. What buyers paid, not what sellers asked.
4. Everything a paid tool sells. Useful, broad, historical, and one step further from the transaction.

Selection order follows from that, and it is permanent: rank candidates by conversion first, then by the ratio of searches to listings, and never by volume alone. Volume is a tiebreaker between two terms that already converted well. It was never the entry criterion, and it is now demonstrably the worst single metric to lead with.

Etsy’s own warning, which is stronger than any volume number

Inside the how-to panel of the tool itself, Etsy states the following, verbatim:

The marketplace is describing a penalty, not offering advice. Stuffing a high-volume term into a listing it does not describe does not merely fail to help; it degrades the relevance signal for the listing and pushes it down. That is the mechanism behind a familiar outcome: a listing tagged with popular terms that collects impressions, converts nothing, and then quietly stops being shown at all.

It also matters more than it used to, because the same accuracy that protects your ranking is what makes a listing legible to machine buyers. Structured, literal, accurate attributes are the substrate that assistants read. That is the whole argument in the guide to what AI shopping SEO actually is, and Etsy has now made the case in its own words.

Keyword accuracy is also what machine buyers read

Etsy’s warning is about its own ranking, but the instruction it gives you happens to be the same instruction that matters when the reader is not a person. An assistant that answers a shopping question does not browse a category page. It reads structured text and returns something it can defend. Literal, accurate wording is what it has to work with, and vague wording chosen for volume is exactly what it cannot use.

That is why the keyword decision described above is not only a search-ranking decision. The same field that Etsy penalises you for stuffing is the field a machine reader parses. One accurate term serves both. One inflated term damages both, and it damages them silently, which is the part that makes it expensive.

This site treats that as a separate discipline, and the whole cluster is built on it:

• The definition is linked in the section above. The field-by-field comparison, including which fields the two disciplines disagree about, is in how it differs from ordinary Etsy SEO field by field.
• Writing descriptions an AI shopping assistant can actually parse, and what photo SEO means once an image is being read rather than admired.
• Optimising a listing for Google AI Mode, what ChatGPT Shopping means for a small Etsy shop, and the Universal Commerce Protocol explained without the announcement language.

None of those pieces promise placement, and neither does this one. They describe how the surfaces read a listing, which is a thing you can check for yourself.

A repeatable ten-minute workflow

Do this once per product, before you write a title and before you set a price.

1. Write five to eight candidate terms on paper. Pair each broad term with its specific version, exactly as badge reel pairs with custom badge reel.
2. Open Shop Manager → Stats → Marketplace insights and look up the pairs, not single terms. The comparison is the finding.
3. Discard every term with a low conversion band, no matter how large its search count is. This is the step that will feel wrong, and it is the entire point.
4. Among the survivors, prefer the best ratio of searches to competing listings.
5. Read the median purchase price range for the winning term and price deliberately against it.
6. Write the title, tags and attributes to describe the item literally, in the buyer’s words — the split between those fields is covered in the piece on Etsy tags versus attributes.
7. Log the term, the bands and your price. In four weeks you will want to know what you decided and why.

The workflow above decides a keyword. Deciding whether the product behind it is worth making at all is a wider screen — demand, supply, price, and the two gates most launches quietly fail — laid out step by step in how to validate demand before you create another Etsy product.

What this does not replace

Keyword selection is one input. It is not a shop.

A perfect term will not rescue a first photograph that nobody clicks, and it will not close the gap between a favourite and an order — that gap is price, shipping and trust, in that order. The measurement side has not changed either: the metric that matters is still the one furthest down the funnel that you can honestly observe, which is the argument made in the guide to measuring AI shopping SEO on Etsy.

Trust is the one input on that list a keyword cannot buy. A shop with no review history converts below its own data no matter how well the term was chosen, and the three compliant levers that move a shop off zero are in how to get your first three reviews on Etsy.

What has changed is narrow and real: the one keyword metric that depends on seeing a purchase is now free, first-party, and sitting in your dashboard. We have no data on how many sellers open it, so the honest version of that claim is the smaller one: it is available to you today, and a lookup costs nothing.

What each source actually says

Every figure on this page comes from one of six sources. This is what each one is, when it was read, and what it actually states.

Frequently asked questions

Is Etsy Marketplace Insights free for every seller?

Yes. It sits at Shop Manager → Stats → Marketplace insights, requires no separate subscription, and has no minimum sales threshold. It was confirmed working in a shop with fifteen lifetime visits and zero orders. The one limit is the metered allowance, which the panel states as fifteen terms per week.

How many Marketplace Insights lookups do I get?

The panel states the allowance in its own words: fifteen terms per week. It is a rolling weekly allowance rather than a lifetime total — the same panel was re-read on 4 September 2026 and had refilled. Plan a week of terms before you open the tool rather than searching on impulse.

Does Etsy publish its own search volume data?

Yes, as of August 2026. Etsy publishes search counts, competing listing counts, a conversion band and a median purchase price range for individual search terms, through the free Marketplace Insights allowance where the feature is available. Any guidance written before that date — including an earlier article on this site, now corrected — states the opposite and should be re-checked.

Should I stop paying for keyword tools?

Not automatically. Etsy’s panel is the better source for a decision about one term, because it is first-party and it publishes conversion. Tools remain better for bulk expansion, catalogue-wide tag audits, competitor tracking, history over time, and other marketplaces. Metered lookups cannot audit two hundred listings.

Which number should I choose a keyword by?

Conversion band first. Then the ratio of searches to competing listings. Then the median purchase price range, to set your price. Volume alone comes last, and should only ever break a tie between terms that already convert well. The verified case is badge reel at 215,600 searches with Very low conversion against custom badge reel at 8,100 searches with Very high conversion.