Helping people who sell worn underwear and fetish content trade safely since 2001. Figures, sources, dates.

Start Regular buyers How do I turn one buyer into ten purchases?

Court 19 May 2026, California: a judge threw out almost the whole case. The one part left alive is against the firms that answer messages for creators — not against the website. If someone else writes in your name, the blame lands on them.
PLACEHOLDER 1:1 · flash, low angle
454,000 visits/month:

The market hands you one buyer. The rest you build.

Between half a sale and four sales a year. That is what the market hands each seller on its own. So your problem is not making more things. Your problem is getting the same person to buy again. Turn one buyer into five and you have beaten the market without finding a single new person.

K10: 460,000+ sellers against 454,000 visits/month (SimilarWeb, measured). We use visits as a stand-in for buyers, because nobody counts buyers. One visit is not one buyer, so the real number can only be lower — that is why this is our maths, not a measurement. Our maths

Why this page exists

The other pages ask what to sell. This one asks the question that is left after your first sale: will they come back?

The sums

Five purchases beats five buyers

454,000 ÷ 460,000 = 0.99 0.99 × 0.5 buys = 0.49 0.99 × 4 buys = 3.95

Buyers per seller, then sales per seller per year at two and at five purchases each. The same number of people, 2.5 times the sales. Not one new buyer appears anywhere in these sums. Our maths

FY2022 audited:

Half the money is not the price tag

52.1% is not the monthly fee

More than half the money on OnlyFans is not the monthly fee — it is tips, paid messages and one-off purchases. If you only put things on a shelf and wait, you are working the smaller half of the business.

OnlyFans (Fenix International), audited annual accounts FY2022. Checked

Block 1

The conversation is the business

If more than half the money is not the monthly fee, then keeping a buyer is not a side job next to selling. It is the selling.

PLACEHOLDER 16:9 · sodium, gold
Charged per message:

A file sells once. A conversation has no last page.

A video is one sale per person. To sell it again you need a new person. A message is charged one at a time, and there is always another message. That is the difference between a shelf of goods and an income.

Our maths, from how the products work, backed by the 52.1 % that is not the monthly fee. Nobody publishes how often buyers come back for each type of thing. Our maths

What people are really paying for

A lot of the money is paying for company, not for sex

And those are the buyers who come back most often. Three things we can prove point that way: more than half the money is not the monthly fee, buyers are charged message by message, and the software built for this trade keeps track of your mood, what he likes and what you will not do — not of stock.

This is our reading of those three facts, not a measurement. No study puts a number on how much of the money is paying for company. We say what we found. We do not tell you what to think of it. Our maths

n = 123,674:

The regulars who pay your rent are also the risk

35.7% live off one or two big spenders

Some sellers make their money from a small circle of regulars. Of those sellers, 35.7 % get half their money or more from a couple of big spenders. Among sellers who are visible on the front page it is only 10.5 %. The study says it plainly: losing even one of those big spenders can badly hurt your earnings.

Every performer on the site was counted: 123,674 of them. Read this carefully: the 35.7 % is of that smaller group, not of all sellers. The short version gets passed around and it is wrong. Full source no. 2. Checked

So what do you do about it

The answer is not fewer regulars. It is more of them.

Every regular buyer is worth more to you than a stranger. That also means it hurts more when one walks away. The fix is not to stop building regulars — it is to have enough of them that no single one can sink you. Which is the same answer the half-a-sale-to-four-sales figure gives, coming from the other side.

Both halves of this come from the same study. Selling you regulars without telling you the risk would be telling you half the story. Checked

Block 2

How the big operations do it

We describe it, we do not judge it. This is the factory version of the same thing one seller does by hand.

PLACEHOLDER 1:1
3–4 accounts each:

Your girlfriend is a shift rota

Big accounts hire people to answer their messages. The job has a name in the trade: a chatter. One chatter works an eight-hour shift and answers messages for 3 or 4 different creators at once, for £250 to £2,500 a month on shifts of 8 to 12 hours. Most are hired in Pakistan, India and the Philippines.

Vice, Eloise Hendy, 24.10.2023. We took these numbers out of the article itself, not from someone quoting it. Checked

PLACEHOLDER 1:1
A reporter took the job:

The character is written down before the first message

A journalist got himself hired as a chatter. On his first day he was handed a sheet describing how the creator talks — he was told to be “40 percent girly” — and a price list for every photo and video she had for sale.

Brendan I. Koerner, WIRED, 15.05.2024. One firm, seen from the inside by the reporter himself. Checked

Where we have to correct ourselves

The “files on buyers” are not in the Vice article

Careful — this part is our thinking, not a quote

What the Vice article really says: that chatters work “eight hour shifts” and cover “three or four creators at a time”. It says nothing at all about keeping files on buyers.

So this part is ours: if several people take turns on the same conversation, they must be writing things down somewhere. We think that is right, and the trick is worth copying — but it is us working it out from the shifts. It is not something Vice reported.

And one more number from Vice that we do not use at all: you will see “50 to 60 % of earnings come from messaging” passed around. The article actually says “Content creators, he says, can make 50 to 60 percent…” — that is one firm’s owner guessing, and he has money riding on the answer. Nobody measured it. The audited 52.1 % says much the same thing and can be checked, so we use that and leave his guess out.

We read the original article on 13.09.2026. This correction goes against our own earlier notes, which had the looser numbers in them. Not proven — so not used

Word for word · where it IS written down

“When another chatter opens the same fan later, they see what was said, what was offered, and whether a follow-up is needed.”

Manual for buyer-tracking software

That is the handover the shift system needs, spelled out by the people who sell the software for it. This one is not us guessing, and it is not from Vice.

Software maker’s own manual, read on 13.09.2026. That company has since stopped selling the product. The manual proves people work this way; it does not mean you can go and buy that tool. Checked

Word for word · what the software sorts people by

“Group fans based on subscription length and spending power for targeted retention and engagement strategies.”

Sales page for buyer-tracking software

How much a person spends is the thing everything else hangs on. The reporter who went undercover heard the same idea in plainer words: there are two sorts of customer, and you spend your time on one of them.

Software maker’s own pages, read 13.09.2026; WIRED 15.05.2024 for how it works in practice. Checked

Why outside software exists at all — the websites leave a gap

The websites tell you whether someone is subscribed, whether it renews by itself, and whether it has run out. What they do not tell you: how much each person has spent, what he is worth to you over time, or when he was last online. Those three things are exactly what the outside software adds.

So keeping notes on buyers is not some creepy extra. It fills a hole the websites left. That is simply what is going on, and you can say it without accusing anyone.

Here is what the websites do give you, straight from their own help pages: messages sent in bulk to everyone whose subscription ran out, automatic messages when someone subscribes, renews, tips or follows, a separate inbox for your biggest spenders, a list of who bought each item, a tip menu where a word or an amount gets you a particular thing, and one-off purchases priced anywhere from $1.00 to $499.99. When a website builds a button for something, that is stronger proof it happens than any guide telling you to do it.

Block 3

What one person can take from all this

The big firms do this with shift rotas and paid software. The part you can copy is much smaller, and it costs nothing.

PLACEHOLDER 16:9 · asphalt, 35 mm
Notes, not memory:

Write down what he told you. That is the whole trick.

The firms need their notes because several people share one conversation. You do not have that problem. You have the other one: you forget. The software manual lists what the notes are for, and it works exactly the same for one person: what he likes, what you will not do, how you talk to him, what he has bought before.

Our maths: we took the written-down workflow and applied it to one seller. Our maths

The one rule that keeps your notes safe

Write down the money. Not what he is into.

  • SaferWhat he spent, when he last bought, whether he is a big or small spender
  • RiskierAnything that ties his name to what he watched or what he is into

This is not us being fussy about style. It is the exact line the judge drew: the danger is the note that links a person’s name to what he watched. The amount he spent is not the problem.

N.Z. v. Fenix, C.D. Cal., Dkt. 251, 19.05.2026. In the EU and Germany there is a second reason: what someone is sexually into counts as extra-protected data under Art. 9 GDPR, so a “what he likes” note is the risky one there too. Checked

Which things get bought again

Some things sell again by their nature

  • YesMonthly subscription — it renews by itself unless he stops it
  • YesPaid messages — he pays per message, and messages never run out
  • YesWorn underwear — it gets used up, so he needs another pair
  • OnceA video — one sale per person; you need new people to sell it again
  • MaybeMade-to-order requests — only as long as he sticks around

Our maths, worked out from how each thing works. Nobody measured this. How often buyers actually come back for each type of thing: no study and no website publishes it. This is the biggest hole on the page and we are not hiding it. Our maths

How a posted parcel turns into a second order

“a handwritten note, a polaroid photo, an audio message”

Extras you can order — the stated reason: “build loyalty”

These little extras are the documented link between one parcel and the next order: they turn a plain sale into something that feels personal. But they cost you on the other side — handwriting can be compared and matched, and every instant photo has a 10 or 11 digit code printed on it that shows which batch of film it came from and ties your pictures together. The thing that brings him back is the same thing that helps someone identify you.

The websites’ own pages for the extras; the film maker’s own documents for the code. Read “Who can find you” before you put either in a parcel. Checked

Block 4

Where it stops being worth it

This is not about right and wrong. It is about a court case with a file number.

PLACEHOLDER 16:9 · dark base
Dkt. 251, 19.05.2026:

The website walked away. The message firms did not.

California, May 2026. People sued a website and the firms that answer messages for creators. The judge threw out nearly everything. Just one complaint was left standing, and only against the firms: that they broke a US privacy law about who is allowed to see what you watch. The same complaint against the website was thrown out for good, along with the claims about organised crime, broken contracts, fraud and unfair competition. If you hand this work to someone else, the trouble lands on them.

N.Z. et al. v. Fenix International Limited et al., No. 8:24-cv-01655-FWS-SSC (C.D. Cal.), Judge Fred W. Slaughter, Dkt. 251, 19.05.2026. The surviving claim is under the VPPA, not CIPA — the wiretapping claims in this same case already failed in December 2025. And note what “left standing” means: the judge only said the complaint is worth hearing. Nothing has been proven and nobody has been found guilty. Checked

Word for word · the judge

“…allegedly disclosed PII from Fans to Chatters by sharing login information or via CRM software.”

The ruling of 19.05.2026

In plain words: handing buyers’ personal details to the hired message-writers, either by sharing the login or through the buyer-tracking software. The judge counts that software as one of the ways the details got out. The person who keeps and fills in those notes is the one who passed them on — not the website.

At this stage a judge only decides whether a complaint is worth hearing. Nothing is proven, no other court has to follow it, and nothing on this page is legal advice. Checked

Word for word · the excuse that did not work

“…not persuaded that the alleged disclosure of PII to Chatters falls within … ordinary course of business…”

The same ruling

The firms argued this was just normal business. The judge did not buy it — not for notes kept to sell people more. And the small print people tick when they sign up protected only the website. It did not protect the firms.

Four of the five complaints were thrown out. This one was not. The rule you can actually use is in the box above: write down the money, not what he is into. Checked

Refutations

What people say / what is actually true

Three of these are us correcting numbers we had wrong ourselves, before we went and read the original.

What people say

50–60 % of what creators earn comes from messaging — and Vice said so

What is true

That number is in Vice, but it is one firm owner’s guess. The audited figure is 52.1 %.

“Content creators, he says, can make 50 to 60 percent…” He sells this service, so he is not a neutral witness, and nobody measured it. The real figure is close, and you can go and check it. Checked

What people say

These firms keep files on buyers — and Vice proved it

Vice wrote about shifts and juggling several accounts. Not about files on buyers.

The files are real — the proof is the software makers’ own manuals, not that article. Naming the right source matters more to us than keeping a good line. We fixed our source

What people say

Message-writers cover 3–7 accounts on 8–10 hour shifts

The article itself says 3–4 accounts, eight-hour shifts, £250–2,500 a month on 8 to 12 hour shifts

Vice, 24.10.2023, read the original ourselves. The bigger numbers come from people repeating it second-hand — and one of those people was us. A number is only as good as the document it came from. Checked

What people say

The top 1 % take 33 % of the money — from a 2026 company report

It is from 2020, and an outsider collected it. There is no such company report.

The website never published this. The figure you can actually check comes from the big study: the top 10 % get more than 69 % of all tips, and the bottom half get under 1 %. Checked

What people say

Everyone gives loyalty discounts and a free gift with repeat orders

We could not find this anywhere. It is a rumour until somebody shows us a source.

We went through the websites’ own help pages on 13.09.2026. What is really there: bulk messages to people whose subscription ran out, automatic messages, tip menus, and separate inboxes for big spenders. Could not check

What people say

Pricing tricks — a high price first, or a bad option next to the one you want — are proven to work here

They are proven in shop and lab studies. Nobody has tested them in this trade.

What is documented is offering pricier things to people who already spend more (WIRED, one firm). We found no pricing experiment in the adult trade at all. A finding from another trade is a guess here, so we will not sell it to you as proof. Not proven here

Block 5

So what is the catch?

Tucked away because it is not the nice part. Open it anyway.

The more a buyer is worth to you, the more it hurts when he goes

The same study that shows regulars are the way to a real income also measures the danger. Among the sellers who live off a small circle of regulars, 35.7 % get half their money or more from one or two big spenders. For that group, big spenders bring in 38.42 % of all the money, against 15.69 % for sellers who are visible on the front page. And the less well-known sellers had more of these big spenders, not fewer.

Read the number carefully, because the short version gets passed around: the 35.7 % is of that smaller group, not of all sellers.

The answer is not to have fewer regulars. It is to have more of them.

The dangerous moment is ending it, not having it

This belongs on this page because a regular buyer is a relationship, and one day it has to end. Two completely separate fields of research found the same shape. Studies of sex workers’ clients found men turn obsessive at the moment they are cut off — blackmail, showing up where she works. And in the much better researched field of women killed by partners, leaving roughly triples the risk, and it goes up about nine times if the man was very controlling.

So do it in this order, and leave the hard block for last, not first:
go quiet → put your prices up → take longer to reply → warn him → limit him → block him

There is a money reason too. Blocking someone does not refund his subscription by itself. If a refund does get approved, it comes out of your earnings. And he keeps whatever he already bought, forever. So a sudden block leaves you with a paying customer who now has a genuine complaint. Take his access away before you end it. The full version is on “Who can find you”.

The risk figures are aOR 3.64 and aOR 8.98 — full source no. 10.

The number this whole page stands on came from the company itself

An earlier version of this page opened with “1.28 buyers per seller”. We got that by dividing two numbers a website published about itself, and nobody checked either of them. Our own calculations refuse to use those two numbers — they appear in none of our working documents. We used measured traffic instead. That route gives between half a sale and four sales per seller per year, or up to about ten if only one in ten listed sellers is actually active.

Both routes point the same way: a handful of sales per seller per year, at most. That direction we trust. The exact decimals we do not. Treat 1.28 as a rough picture, not as a head count.

The honest one-line version: no website publishes checked buyer numbers, and however you do the sums, a seller makes a few sales a year, not dozens.

Nobody searches for this, which is why this page came last

Almost nobody types “how do I get a buyer to come back” into a search engine. Every other page here was built around things people actually search for. This one was built because it is the most useful page on the site and brings in the least traffic.

That fits the rest of the plan. The popular search terms everyone fights over are the ones where an AI answer box replies and nobody clicks through. We would rather be the page that gets quoted than the page that comes first.

Block 6

The other ways in

Every kind of selling shows up on every one of these pages — as a row in a table, never as a page of its own.

Start All 19 things people sell All four questions, 19 kinds of selling, and the full table Page 1 What pays best What pays best for my time? — regulars matter most once your body caps how much you can make Page 2 Who can find you Who can recognise me? — the longer someone stays, the more he learns about you Page 3 Getting paid Where does my money get stuck? — money frozen on a website costs you more than a lost buyer Page 4 Sending it What arrives, what gets stopped? — the little extras that bring him back can also identify you

Free tool · nothing to upload

Buyer notes, kept the safer way

The one thing worth copying from an industry that runs on shift rotas is the habit of writing things down: what he spent, when he last bought, whether he is a big spender, how you talk to him. The judge’s rule is built into the template — the money, not what he is into — because the note that links a person to what he watched is the one that got these firms sued.

Our maths, from the written-down workflow and the ruling Dkt. 251. Tools live on the Gear page. These pages link to it, never the other way round. Our maths