Steady traffic with falling orders is a completely different problem from a visibility crash, and almost every piece of Etsy advice online is written for the other one. In August 2026 Etsy published enough about how its search and feed now work to explain part of this. Below: the one number that tells you which problem you have, what Etsy actually said in its own words, and what to do about the listing itself.
When orders stop, the reflex is to go back to keywords — new tags, a new title, a keyword tool subscription. But if your views are holding, the search step already worked. People found you. Keyword tools grade the part you have already passed.
What broke is the step after arrival: someone landed on your listing, or saw it in a grid, and did not buy. That is a different mechanism with different fixes, and it lives in the listing itself — the photo in the grid, the opening of the description, the specifics, the price with shipping attached.
If the problem were visibility, the fix would be search work: titles, tags, being findable at all. That is what most people do next, and if the problem is conversion it changes nothing, which is how sellers end up rewriting two hundred listings and watching the same flat line.
If the problem is conversion, the fix is what the listing says and shows once someone is already looking at it. Those two plans look nothing alike, and you can tell them apart in about fifteen minutes.
Step one
About 15 minutes. Do this first.
You are trying to separate fewer people saw it from the same people saw it and passed. Your conversion rate does that, and Etsy shows it to you.
| What your Stats show | What it means |
|---|---|
| Visits flat or up, conversion rate down | This page. People are finding you and choosing something else. The problem is on the listing, after arrival. |
| Visits down, conversion rate flat or up | A visibility problem, not a listing-quality one, and the better of the two to have. Different guide: why did my Etsy views drop. |
| Both down, starting on the same date you last edited listings | Edits, most likely. Every title change resets that listing's accumulated performance history. Again, the views guide covers it, including why reverting makes it worse. |
| Visits flat, conversion flat, revenue down | Not a listing problem at all. Check your average order value — you may be selling the same number of cheaper items, or a bundle or discount changed. |
Write the numbers down somewhere before you touch a listing. Once you start editing you lose the baseline, and without a baseline you cannot tell later whether anything worked.
Step two
Worth ten minutes, because most of what is circulating is a paraphrase of a paraphrase.
We are not going to tell you what Etsy secretly did. We are going to quote the document. All four passages below are verbatim from Etsy's Q2 2026 shareholder letter, published 5 August 2026 and filed with the SEC. Under each one is what we think it means for a seller, and where we are not sure, we say so.
"At the same time, we continued to improve real-time personalization in Search while evolving our approach to quality, moving beyond surfacing what has historically been most popular or most likely to convert, toward helping buyers discover the best our sellers have to offer for their specific shopping mission."
What it means for you. For years, a listing's own track record was one of the strongest things it had going for it — popular things got shown, which made them more popular. Etsy is saying it is moving beyond ranking primarily on that. If your listing's main advantage was accumulated history rather than what the listing actually says, it has less of that advantage than it did.
The flip side is real: a listing that clearly states what it is, for whom, in what material and size, has more to match on than a listing whose only asset is that it sold well in 2024.
Be careful with this one. Etsy is describing a direction, not announcing a switch that flipped on a date. Nobody outside Etsy can tell you how much weight moved or when. Anyone who gives you a percentage is guessing.
"Underpinning these improvements are significantly richer buyer profiles—now covering more than 65 million buyers and providing three times more signals than earlier this year—with an average of about 30 interests and preferences per buyer."
What it means for you. More of what gets shown depends on the individual shopper. Two people typing the same words are increasingly seeing different results.
The immediate practical consequence: searching for your own listing tells you almost nothing now. You are a logged-in person with a profile full of your own products. Screenshots of "my position" mean less than they used to, and so does a friend's screenshot. Trust your Stats, not the search bar.
The second consequence is the useful one. A profile made of interests and preferences has to be matched against something. Vague listings give it nothing to grab: "wall art" is not an interest, "linocut print" is.
"Improvements to feed freshness, including suppressing over 80% of previously viewed and engaged listings, drove gains in feed favoriting, listing views, and new searches while remaining GMS neutral."
What it means for you. This is the passage being quoted most, usually without the caveats, so read it precisely. It is about the feed — the personalized home and recommendation surface — and not about search results. It means a shopper who has already seen your listing is much less likely to be shown it again there.
If some of your sales came from repeat impressions — the shopper who sees a piece three times over a fortnight and buys on the third — that path is thinner than it was. It is also a reason your views can look normal while orders fall: the letter says the change increased listing views. Views from a first-time impression and views from a fourth reminder do not convert at the same rate.
Be careful with this one. Etsy says the change was "GMS neutral" — neutral for total sales across the whole marketplace. Neutral in aggregate is entirely compatible with some shops losing and others gaining. It is not evidence that your shop was unaffected, and it is not evidence that it was.
"As in the first quarter, average order value was the largest contributor to GMS growth, with higher listing prices being the prevailing driver."
"Our efforts to elevate higher-quality items on- and offsite across search, discovery, and marketing are increasingly contributing to a shift toward higher AOV items."
What it means for you. Etsy says it is deliberately elevating what it judges to be higher-quality items, and that this is showing up as a shift toward higher order values. What Etsy means by "quality" here is not published in any detail, so treat the word as Etsy's, not as a rubric you can score yourself against.
Sellers of low-ticket items — the $6 bookmark, the small print, the add-on trinket — have been reporting those listings going quiet while the rest of the shop holds. That is sellers reporting a pattern, not a measured effect, and we are labeling it as one. If it is happening to you, your own Stats per listing will show it far more reliably than anyone's forum post.
Etsy did not say established shops are being penalized. It did not say handmade is being downranked. It did not announce a "2026 algorithm update" with a date on it. Those framings are circulating alongside the real quotes, which is exactly why the quotes are worth reading in full.
Step three
Ten minutes per listing. Uncomfortable, and the highest-value thing here.
If people are arriving and not buying, the decision is being made in the first few seconds, in two places you have stopped being able to see because you already know what you sell.
Step four
An afternoon, then a month of restraint.
The instinct after a drop is to redo everything. Two reasons not to. Every title edit resets that listing's accumulated performance history — that one is an observed pattern reported by sellers and SEO tools rather than something Etsy publishes, but it is consistent enough to plan around. And if you change everything at once you will never learn which change did anything.
If conversion moves on the five and not the control group, you know what to roll out. If nothing moves on either, the problem is not the listing text and you have saved yourself from rewriting the whole shop for nothing. Both outcomes are worth more than another month of guessing.
Some honesty about the limits, because plenty of quiet months in 2026 have nothing to do with the above:
Do step one, per listing. Knowing whether your conversion rate fell or your visits fell tells you which of two completely different problems you have, and every hour spent before you know that is a coin flip.