How Etsy search actually ranks a listing

<p>Every seller has read that Etsy's algorithm is a black box. It isn't. Etsy publishes most of what it weighs, and the parts it doesn't publish show up plainly once you look at enough listings. What follows is the mechanism in the order it actually runs, with the folklore stripped out.</p> <p>The single most useful thing to understand is that ranking happens in two distinct stages, and almost all bad advice comes from confusing them.</p> <h2>Stage one: are you even eligible?</h2> <p>Before anything is ranked, Etsy has to build a candidate set — the pool of listings that could plausibly answer this query. It does that by matching the shopper's words against four fields, roughly in this order of weight:</p> <ol> <li><strong>Title</strong> — 140 characters</li> <li><strong>Tags</strong> — thirteen slots, twenty characters each</li> <li><strong>Category</strong> — where the listing sits in the taxonomy</li> <li><strong>Attributes</strong> — the structured options you filled in</li> </ol> <p>If your listing does not match on any of those, nothing else in this article applies to you. You are not ranked badly; you are not in the running at all.</p> <p>This is why the most expensive mistake in Etsy SEO is a wasted tag. You get thirteen. A tag nobody types costs you a slot and returns exactly nothing — not a low ranking, nothing.</p> <p>It is worth being blunt about how common this is. Run a handful of your own tags through a <a href="/tools/etsy-related-keywords">related-keyword check</a> and count how many come back with no measurable demand. For most shops the answer is three or four out of thirteen — a quarter of your eligibility surface, spent on phrases nobody searches.</p> <h3>Phrase matching beats word soup</h3> <p>Etsy matches multi-word phrases, not a bag of loose words. A tag reading <em>cat mom shirt</em> is eligible for that phrase. Three separate tags reading <em>cat</em>, <em>mom</em> and <em>shirt</em> are not a substitute — they compete in three enormous, unwinnable pools instead of one specific one.</p> <p>The practical rule: every tag should be a phrase a person would actually type into the search box, in the order they would type it.</p> <h2>Stage two: listing quality score</h2> <p>Once you are in the candidate set, Etsy sorts largely on how well your listing converts the traffic it already receives. Views that end in a purchase or a favourite push you up. Views that bounce push you down.</p> <p>This has a consequence people find counter-intuitive: <strong>a listing with fewer views can outrank one with many more.</strong> Etsy is not ranking popularity, it is ranking the probability that the next shopper buys.</p> <p>You can see the effect in any competitive term. Pull the <a href="/tools/etsy-keyword-listings">listings winning a keyword</a> and read the conversion column rather than the sales column. It is normal to find the top listing converting at 9 or 10 per cent while the one below it converts at 4 per cent on twice the views. The first listing has a photograph and a price that work. The second is riding traffic it is in the process of losing.</p> <h3>Why the first fortnight matters more than the next six months</h3> <p>Etsy gives new listings a temporary visibility boost. It is not generosity — it is the platform buying conversion data it does not have yet. It needs to know whether your listing converts before it can rank it.</p> <p>Spend that boost on a weak photograph and you do not simply miss out on a fortnight of sales. You hand Etsy a poor conversion figure, and the listing then has to climb out of a hole it dug on day one.</p> <p>This is the strongest practical argument against the common advice to "list a lot and see what sticks". Volume is fine. Listing something that is not ready is not, because the cost is not zero — it is a permanent handicap on that listing.</p> <h2>The smaller signals, honestly weighted</h2> <table> <thead><tr><th>Signal</th><th>Real weight</th><th>What to do about it</th></tr></thead> <tbody> <tr><td>Recency</td><td>Small, short-lived</td><td>Not a reason to relist on a schedule. Renewing a listing that already converts throws away its history for a bump measured in hours.</td></tr> <tr><td>Shipping price</td><td>Genuine, market-dependent</td><td>In markets where Etsy promotes free delivery, a listing with paid postage is competing at a disadvantage it cannot photograph its way out of.</td></tr> <tr><td>Review average</td><td>Moderate, slow</td><td>Below about 4.8 at volume usually signals a fulfilment problem rather than a product one.</td></tr> <tr><td>Shop age</td><td>Small</td><td>Nothing to do. It resolves itself.</td></tr> <tr><td>Shopper location</td><td>Real</td><td>Two people searching the same words in different countries see different results. Your rank is not one number.</td></tr> </tbody> </table> <p>That last row deserves emphasis, because it quietly invalidates a whole genre of advice. There is no single position for your listing. When someone tells you they "rank third", they mean third from where they were sitting, on the day they looked, logged in as themselves.</p> <h2>What this means for choosing a keyword</h2> <p>Put the two stages together and the strategy falls out on its own.</p> <p>Stage one says: be eligible for phrases people actually type. Stage two says: be eligible for phrases where you can realistically convert better than the incumbents.</p> <p>Which is why raw search volume is close to useless as a target. Take a genuinely large term like <em>gift for her</em>. On recent data it moves well over 900,000 units in a month — and scores around 22 out of 100 on opportunity, because the demand is already absorbed by shops with years of conversion history. A new listing there is eligible and invisible.</p> <p>A term a tenth of that size, with an opportunity score above 55, is the better bet every time. There is demand, and the incumbents have not locked it up. That trade — smaller pool, winnable position — is the whole game.</p> <p>The <a href="/tools/etsy-keyword-research">keyword research tool</a> gives you both numbers side by side, and the <a href="/tools/etsy-niche-finder">niche finder</a> searches for the combination directly.</p> <h2>Five things that do not work</h2> <p>Worth stating plainly, because each of these still gets repeated in seller groups every week.</p> <ul> <li><strong>Repeating a keyword in every tag.</strong> Etsy does not count occurrences. You have spent thirteen slots on one phrase.</li> <li><strong>Stuffing the title with commas.</strong> A title written for a machine reads badly to a human, and the human is the one whose click decides your quality score.</li> <li><strong>Relisting daily.</strong> A small recency bump, paid for with your accumulated conversion history.</li> <li><strong>Copying a bestseller's tags.</strong> Their tags work because their listing converts. Bolt them onto a listing that does not and you have simply entered a race you lose faster.</li> <li><strong>Chasing the biggest term you can find.</strong> Covered above, and it is the most expensive mistake on this list.</li> </ul> <h2>A working order for a new listing</h2> <ol> <li>Find a phrase with real demand and an opportunity score you can live with.</li> <li>Check who currently ranks for it and what they convert at. If the leaders are all above 8 per cent, pick a different phrase.</li> <li>Write the title for a person, with the phrase near the front.</li> <li>Fill all thirteen tags with phrases that have measurable demand. Verify, do not guess.</li> <li>Put the strongest photograph first. This is the single largest lever on your quality score.</li> <li>Leave it alone for a fortnight, then read what happened.</li> </ol> <p>Step six is the one people skip. Etsy needs two weeks to form an opinion. Editing on day three throws away the experiment before it has produced a result.</p>