Djinny
Newbie
- Jan 11, 2018
- 8
- 5
How I did this:
The source is a full export of a Telegram groups - HTML files, one for every chunk of conversation going back to July 2023. I wrote a parser to extract every message with its sender and text, which got me 28,475 messages from 1,781 unique accounts. From there I ran keyword searches across the full dataset to find how often specific topics came up (bans, ads, conversion rate, support, pricing, etc.), then read through the actual threads those searches surfaced to understand the context. I also pulled everything sent by the most active contributors separately - the top poster had nearly 2,000 messages, the groups admins close to that - and read their threads as a coherent body of advice rather than isolated quotes. No AI summaries. Just reading.
A few caveats worth naming: the group is Russian-speaking, so everything I'm quoting is translated and some nuance gets lost. It's also self-selecting - people who join a private seller community and post actively are not a random sample of Etsy sellers. They tend to be more experienced, more frustrated, or both. Take it accordingly.
Here are my findings.
The algorithm apparently picks 1-2 search queries per day per listing, based on your budget
This came from one guy who runs three shops and has been at this for years. It is also heavily supported by other sellers. His read: Etsy isn't distributing your spend evenly across all your tags. It picks a small number of queries it thinks are relevant, bets on those, and if a competitor bids higher on the same query, you just lose that slot. Without warning or notification. Which is why throwing more money at a working campaign sometimes just makes it worse - the problem wasn't the budget, it was that you got outbid on the one query that was actually converting.
On ads: 20-30 clicks per sale is the benchmark that kept coming up
First page clicks are running about $0.40-0.50 depending on the niche. Do the math and your theoretical cost per sale under good conditions is somewhere around $12-13. If you're spending $50+ per sale, nobody in that group would tell you to increase the budget. They'd tell you to look at your conversion rate.
The scaling is real problem and almost everyone there had experienced it. Find a $25/day budget that works, try to take it to $100, watch Etsy immediately start spreading the spend to junk queries. The move when you scale is to go straight into Search Terms and see what Etsy started bidding on with the extra room. Most people never do this. One seller described it as Etsy going "feral" with the budget the moment you loosen the cap.
Conversion rate is probably more important than your tags
This kept coming up in different forms. The basic idea: high conversion gets you more organic distribution, low conversion gets you buried, and the algorithm doesn't care why your conversion is low. Bad photos, wrong audience, one viral bad review - doesn't matter, the number goes down and so do your shop.
The trick that made sense for cheap digital items ($1-3): put a new listing in ads, wait for exactly 20 clicks, see if there's a sale. If yes, kill the ads. The listing got its conversion signal. Running ads longer on a cheap item just piles up empty clicks that tank the ratio.
936 messages mentioned bans or suspensions. More than any other topic.
And this wasn't people casually mentioning it - a lot of those messages were people in the middle of it, or just out of it, or fearing it. The thing nobody warns you about: multiple sellers said their stats seemed to reset after reinstatement. Conversion dropping from ~1% to 0.4% after coming back, organic rank gone, essentially starting over while being held to the standards of an established shop.
The appeal strategy the group had landed on: don't use template language. Write something personal. Mention it's your main income if it is. The boilerplate appeals appear to route to auto-processing. Whether that's actually true or just survivor bias from people who wrote personal letters and got reinstated, hard to say.
Etsy support is not going to help you
I know this isn't news. But reading 279 messages about it back to back makes it feel more concrete. Someone asked a specific technical question and got back three paragraphs about keyword best practices. Someone opened a complaint about a competitor copying her listings and received tips on how to use tags. The questions get matched to a topic category and the category gets a canned response. Nobody's reading it.
The workaround that got mentioned a few times: frame it as an account security or payment issue rather than a general question. Apparently those categories have different routing.
One thing that became kind of a running theme: cheap products attract bad buyers. Multiple sellers with low-ticket physical items said the dispute and refund rate was noticeably worse than on their higher-priced stuff. Not just margin math - the actual customer behavior changes. Raising prices to filter the customer pool, not to make more money per unit, was something a few people had tried and said worked.
The Star Seller thing: the most experienced guy in the group basically said don't bother chasing it. His logic was that it all comes down to whether buyers leave reviews, and whether they accidentally give you 3 stars instead of 5. Too much random noise to build a strategy around. He said it, got pushed back on, held his position.
The SEO thing that changed how I think about tags
Before using a keyword, search for it on Etsy and actually look at the results. If page one is full of stuff that isn't what you sell, that tag is sending the wrong traffic, running up your click count without converting, and dragging your ratio down. Someone in the group was using "leopard wall decal" on their leopard wall decals. Totally reasonable. Except the top results for that search were leopard-print fabric. Etsy had learned the query meant something adjacent. Five minutes of checking would have caught it.
One person launched a shop with a $1,000/day ad budget
Actual spend was $200-250/day. He ran it like that early on, then tapered down to $50/day once organic was working. The reasoning: you're going to go through an algorithm learning phase no matter what, you can either limp through it at $10/day for six months or pay to compress it. Not something most people can do, but the underlying idea - front-load spend early, taper as organic builds - got general agreement.
There was also a whole thread about card security I wasn't expecting. A bunch of sellers figured out that Thai VPS providers they'd used for hosting had leaked card data. Cards linked to Etsy, Erank, Corjl, various tools started showing fraudulent charges. The fix they landed on was setting a $1/month spend limit on any card used for online tools and only raising it when making a payment. Paranoid sounding until you're the one with a $200 charge from "FACEBK" at 3am.
Anyway. Three years of people talking Etsy shops. The honest summary from the most active voice in the group, translated:
"Almost every shop needs its own vector. Some shops I've been experimenting with for two or three months. Etsy loves changing its strategies. The main thing is: wait, do, wait again."
Sounds about right.
The source is a full export of a Telegram groups - HTML files, one for every chunk of conversation going back to July 2023. I wrote a parser to extract every message with its sender and text, which got me 28,475 messages from 1,781 unique accounts. From there I ran keyword searches across the full dataset to find how often specific topics came up (bans, ads, conversion rate, support, pricing, etc.), then read through the actual threads those searches surfaced to understand the context. I also pulled everything sent by the most active contributors separately - the top poster had nearly 2,000 messages, the groups admins close to that - and read their threads as a coherent body of advice rather than isolated quotes. No AI summaries. Just reading.
A few caveats worth naming: the group is Russian-speaking, so everything I'm quoting is translated and some nuance gets lost. It's also self-selecting - people who join a private seller community and post actively are not a random sample of Etsy sellers. They tend to be more experienced, more frustrated, or both. Take it accordingly.
Here are my findings.
The algorithm apparently picks 1-2 search queries per day per listing, based on your budget
This came from one guy who runs three shops and has been at this for years. It is also heavily supported by other sellers. His read: Etsy isn't distributing your spend evenly across all your tags. It picks a small number of queries it thinks are relevant, bets on those, and if a competitor bids higher on the same query, you just lose that slot. Without warning or notification. Which is why throwing more money at a working campaign sometimes just makes it worse - the problem wasn't the budget, it was that you got outbid on the one query that was actually converting.
On ads: 20-30 clicks per sale is the benchmark that kept coming up
First page clicks are running about $0.40-0.50 depending on the niche. Do the math and your theoretical cost per sale under good conditions is somewhere around $12-13. If you're spending $50+ per sale, nobody in that group would tell you to increase the budget. They'd tell you to look at your conversion rate.
The scaling is real problem and almost everyone there had experienced it. Find a $25/day budget that works, try to take it to $100, watch Etsy immediately start spreading the spend to junk queries. The move when you scale is to go straight into Search Terms and see what Etsy started bidding on with the extra room. Most people never do this. One seller described it as Etsy going "feral" with the budget the moment you loosen the cap.
Conversion rate is probably more important than your tags
This kept coming up in different forms. The basic idea: high conversion gets you more organic distribution, low conversion gets you buried, and the algorithm doesn't care why your conversion is low. Bad photos, wrong audience, one viral bad review - doesn't matter, the number goes down and so do your shop.
The trick that made sense for cheap digital items ($1-3): put a new listing in ads, wait for exactly 20 clicks, see if there's a sale. If yes, kill the ads. The listing got its conversion signal. Running ads longer on a cheap item just piles up empty clicks that tank the ratio.
936 messages mentioned bans or suspensions. More than any other topic.
And this wasn't people casually mentioning it - a lot of those messages were people in the middle of it, or just out of it, or fearing it. The thing nobody warns you about: multiple sellers said their stats seemed to reset after reinstatement. Conversion dropping from ~1% to 0.4% after coming back, organic rank gone, essentially starting over while being held to the standards of an established shop.
The appeal strategy the group had landed on: don't use template language. Write something personal. Mention it's your main income if it is. The boilerplate appeals appear to route to auto-processing. Whether that's actually true or just survivor bias from people who wrote personal letters and got reinstated, hard to say.
Etsy support is not going to help you
I know this isn't news. But reading 279 messages about it back to back makes it feel more concrete. Someone asked a specific technical question and got back three paragraphs about keyword best practices. Someone opened a complaint about a competitor copying her listings and received tips on how to use tags. The questions get matched to a topic category and the category gets a canned response. Nobody's reading it.
The workaround that got mentioned a few times: frame it as an account security or payment issue rather than a general question. Apparently those categories have different routing.
One thing that became kind of a running theme: cheap products attract bad buyers. Multiple sellers with low-ticket physical items said the dispute and refund rate was noticeably worse than on their higher-priced stuff. Not just margin math - the actual customer behavior changes. Raising prices to filter the customer pool, not to make more money per unit, was something a few people had tried and said worked.
The Star Seller thing: the most experienced guy in the group basically said don't bother chasing it. His logic was that it all comes down to whether buyers leave reviews, and whether they accidentally give you 3 stars instead of 5. Too much random noise to build a strategy around. He said it, got pushed back on, held his position.
The SEO thing that changed how I think about tags
Before using a keyword, search for it on Etsy and actually look at the results. If page one is full of stuff that isn't what you sell, that tag is sending the wrong traffic, running up your click count without converting, and dragging your ratio down. Someone in the group was using "leopard wall decal" on their leopard wall decals. Totally reasonable. Except the top results for that search were leopard-print fabric. Etsy had learned the query meant something adjacent. Five minutes of checking would have caught it.
One person launched a shop with a $1,000/day ad budget
Actual spend was $200-250/day. He ran it like that early on, then tapered down to $50/day once organic was working. The reasoning: you're going to go through an algorithm learning phase no matter what, you can either limp through it at $10/day for six months or pay to compress it. Not something most people can do, but the underlying idea - front-load spend early, taper as organic builds - got general agreement.
There was also a whole thread about card security I wasn't expecting. A bunch of sellers figured out that Thai VPS providers they'd used for hosting had leaked card data. Cards linked to Etsy, Erank, Corjl, various tools started showing fraudulent charges. The fix they landed on was setting a $1/month spend limit on any card used for online tools and only raising it when making a payment. Paranoid sounding until you're the one with a $200 charge from "FACEBK" at 3am.
Anyway. Three years of people talking Etsy shops. The honest summary from the most active voice in the group, translated:
"Almost every shop needs its own vector. Some shops I've been experimenting with for two or three months. Etsy loves changing its strategies. The main thing is: wait, do, wait again."
Sounds about right.