Selling digital products direct to an audience

The research was done by AI agents that open web pages. Some sites refuse them; where that happened we say so. The small numbers point to the source list at the end. Where a figure is our own sum or guess, it is marked “our estimate” and the basis is given.

The claim

These are examples of the claim, copied from our collection of 54 tweets about this way of making money¹. View counts are as collected on 6 October 2026. Dates are worked out from each tweet’s ID number. The page judges the method, not the people who tweeted.

“I rejected every fancy way to make money online: Youtube, Personal branding, Dropshipping, freelancing. And started selling simple $19 PDFs, made $50k last month.”

@Manu_Sisti, 1 October 2026, 93,364 views¹

“I’ve failed at AirBnB, YouTube, and Dropshipping. But in the last 11 months, I’ve made $700,000 selling a simple $19 PDF.”

@TommiPedruzzi, 20 September 2026, 223,375 views¹

“I made >$3.5M in 2024 selling digital products”

@X_FINALBOSS, 7 October 2025, 315,488 views¹

“Kids are making $25K/month with digital products. But most people don’t know how.”

@ChrisClickUp, 15 May 2023, 5,517,746 views¹

We quote only the part that states the income. We did not open the threads under the tweets.

The opening lines of the two “$19 PDF” tweets do not say what the PDF is or where it is sold, and we did not open the threads under them. Both authors’ profiles point to publishing books on Amazon: one reads “7-Figure by selling eBooks on Amazon KDP (without writing)”, the other “Want to learn how to sell books?”⁵². If the $19 PDF is a book on Amazon, which supplies the buyers, the claim belongs to our page on Amazon ebook publishing. This page tests what it would take if the PDF were sold direct.

What the scheme is

A “digital product” is a file sold as a download: an ebook, a PDF guide, a spreadsheet or planner template, a pack of prompts (ready-made instructions for an AI chatbot). The same file can be sold any number of times.

The seller puts the file on a checkout platform. The two named most often in the tweets are Gumroad and Whop¹. The platform takes the payment, delivers the file and keeps a fee. The seller has to bring the buyers: by posting on X, through the platform’s own marketplace, or with paid adverts¹.

The tweets describe two versions. One is a cheap PDF, about $19, sold in large numbers. The other is an AI-written ebook sold for hundreds of dollars: “create 5 ebooks , 200 pages each within 20 minutes with AI , sell each for $500” with “99% margins”¹.

People do sell files this way and some earn a lot. The questions are how many sales the claimed incomes need, where the buyers come from, and how many sellers get there.

The arithmetic

Step 1: what one sale leaves

Gumroad charges 10% plus $0.50 on each sale made through the seller’s own page or link. Card processing of 2.9% plus $0.30 comes on top². There is no monthly fee². Our sum for a $19 PDF: $1.90 + $0.50 + $0.55 + $0.30 = $3.25 in fees, about 17% of the price. The seller keeps $15.75.

Two other Gumroad rates matter. If the buyer finds the product through Gumroad’s own marketplace, called Discover, the fee is a flat 30% including processing²³, which leaves $13.30 (our sum). And for big sellers: “Once your paid sales in a calendar month reach $20,000, new direct sales that month are 5% + $0.50 instead of 10% + $0.50”². Those later sales leave $16.70 each (our sum).

Whop lists 2.7% plus $0.30 per sale on a domestic card, with 1.5% more for an international card and 1% if the currency is converted⁴. The same page lists $0.03 and $0.07 per transaction for two card-security checks, and optional extras of 0.8%, 0.5% and 2% that bring the total to “6% + $0.30”⁴. A domestic $19 sale leaves about $18.09 to $18.19 with no extras, depending on whether the security-check fees apply to every sale, and about $17.50 with all of them (our sums). Moving the money to a bank costs $2.50 per next-day transfer⁴.

So “99% margins” is not right for a low-priced item, but fees are not the obstacle: the seller keeps 70% to about 95% of the price (our sum).

Step 2: how many sales a monthly income needs

Sales needed each month, before adverts, tools, refunds and tax (our sums from step 1; the higher prices use Gumroad’s own-link rate):

The $50,000 row for Gumroad’s own link allows for the lower fee above $20,000 of sales in the month.

At $19, 635 sales a month is about 21 a day and 3,050 is about 102 a day. At $300 to $500, $10,000 a month is about one sale a day. A low price needs volume. A high price needs strangers to trust the seller with hundreds of dollars.

Step 3: what the tweets’ own figures imply

Dividing a claimed income by the stated price gives a number of sales, which the tweet text we read does not state. The figures below are our estimates and assume all the money came at that price. They show what each figure would need at that price. They are not a test of whether any claim is true, which we did not check.

• “$19 PDFs, made $50k last month”: $50,000 / $19 = 2,632 sales in a month, about 88 a day¹.
• “$700,000 selling a simple $19 PDF” in 11 months: 36,842 sales, about 110 a day¹.
• “$7000+ made while I slept”, where “Each sale was $300”: about 23 sales in one night¹. The full tweet, read through a copy of X, adds that buyers “also get few videos + discord community access”, and that the seller “built up audience and announced launch 5 days before launch”⁵¹.
• “I made >$3.5M in 2024”. The full tweet gives a recipe addressed to the reader: “~400 people view the checkout page” and “u get 20 sales for $500 each. $10K/month profit”⁵¹. At $500, $3.5 million is 7,000 sales in the year, about 19 a day (our sum). The recipe describes what a reader would get: 20 sales a month, which is $120,000 a year (our sum). That is about one-thirtieth of the income the author states as their own; the tweet does not say the two figures describe the same business, and it also says “$449K in past 5 weeks”⁵¹. The same account quotes “$300” and “$50+” in other tweets¹; at those prices the count is about 11,700 or 70,000.

For the cheap version, either the volume is very large or the sales happen somewhere that supplies buyers, such as Amazon. For the expensive version, the tweets themselves describe more than a PDF: videos, a community and an audience gathered before launch.

Step 4: how many visitors that takes

No platform publishes the usual share of page visitors who buy. We found only sellers’ own reports. In a 2021 Indie Hackers thread, sellers reported “roughly 7%”, “8.9%” (inflated by free products, the poster said) and two of about zero¹⁰. Another seller’s report for September 2025 shows 118 “sales” from 1,461 page views, about 8%, worth $160.77 in total because many were free downloads¹¹.

Visitors to the sales page needed each month at different buying rates (our estimates):

These are people who click through to the sales page, not people who see a post. No source gives a buying rate for a $300 or $500 PDF from a stranger. The recipe in one tweet assumes 5% (20 buyers from 400 visitors)⁵¹. We would expect fewer people to pay $500 than $19, but nobody has published a measurement.

Step 5: where the visitors come from

The platform. Gumroad’s marketplace, Discover, is closed to a new seller. Gumroad’s help page says an account must “Reach a balance of at least $100 USD from enough genuine sales”, and then be checked by its risk team, “which can take on average 3 weeks after you pass the $100 threshold”⁷⁵. Each product also needs at least one sale that was not refunded⁷⁵. Gumroad’s payout page gives the same line from the other side: the review typically needs “3-4 sales with a balance over US$100” and takes one to three weeks, and it warns that buying your own product can get the account banned⁷⁶.

Correction, 7 October 2026: this page first gave the rule from sellers’ reports as “a first sale, or about $10”²¹²³ and said we had not found Gumroad’s own statement. We have now read it. The account line is $100, not $10⁷⁵.

So a new seller gets no buyers from the marketplace until they have brought about $100 of sales themselves. One with 26 products and no sales found that every view came from links he had posted himself²³. For established sellers, Gumroad’s 2020 survey is the evidence: 96 of 601 creators (16%) named Discover a top-three sales channel, the same number as named Google Search. Their own websites and social media ranked above both⁵.

Whop looks the same. Sacra, a research firm, says that in April 2025 “direct sales were about 97% of GMV, while Discover was only about 3% of GMV”⁸¹. GMV means the total value of sales, and Discover is Whop’s marketplace. We read that in Sacra’s machine-written summary of its own research, not in Whop’s records⁸¹.

Free posts on X. A study of 70 million posts by brand accounts found the average X post got 2,979 views in 2025 (1,430 in 2024). X had the lowest rate of likes and replies per follower of the networks measured: 0.12% in 2025 and 0.10% in the first half of 2026¹⁵. Brand accounts are not new personal accounts, so this is a rough guide.

The best-documented success on X gives a second yardstick. Daniel Vassallo, a former Amazon engineer, reported $104,754 of sales from an ebook and a video course in four months, with 95% “directly attributable to my Twitter account”, after his tweets had been shown 35 million times in 14 months¹³. That is about $3 of sales per 1,000 times a tweet was shown, and about 2.5 million views a month (our estimates; the views cover a longer period than the sales). At that rate $10,000 a month needs about 3.3 million views a month, a little more than he averaged, and $50,000 needs about 17 million (our estimate). A new account whose posts get a few hundred views each is roughly a hundred times short of that (our estimate).

Paid adverts. A September 2026 round-up of advert benchmarks on Meta (Facebook and Instagram) relays two datasets¹². WordStream puts the average cost of a click at $0.60 for campaigns that aim for clicks. Triple Whale’s figures for online shops, mostly selling physical goods, are middle values: 1.53% of clicks end in a purchase and $38.99 of advert spend buys one sale. Its cheapest category is e-learning, at $26.80 per sale with an average order of $34.33¹². We read the round-up, not the two reports.

At those costs a $19 PDF that leaves $15.75 loses about $11 to $23 on each sale (our sum). To break even at $0.60 a click, 3.8% of clicks must buy, about two and a half times the shop figure (our estimate; the click cost and the buying rate come from different datasets).

The one seller we found who published advert results shows where adverts did work. Vassallo advertised his Amazon-cloud book in the Reddit forum for that exact subject. At about $0.50 a click “it was barely breaking even so I decided to stop”. At $0.10 a click during the 2020 pandemic “the initial results were excellent”. When prices rose again he was “struggling to break even” and stopped, adding “there’s likely more opportunity there”. That advert brought $13,734 of sales; he does not give the spend¹⁴.

Step 6: what it costs to try

Gumroad has no monthly fee². Tweet Hunter, the post-scheduling tool named in the tweets, starts at $29 a month¹⁶. Claude Pro, one of the AI chatbots named, costs $20 a month¹⁷. So the tool bill is about $50 a month (our sum). An email list, the tool most tied to higher earners in Gumroad’s survey⁵, is free to start: MailerLite is free up to 250 subscribers and 2,500 emails a month and then starts at $12 a month, and beehiiv is free up to 2,500 subscribers⁸³. Stan, a checkout platform aimed at Instagram and TikTok sellers, charges $29 a month according to Sacra; its own pricing page did not open⁸². The real costs are unpaid hours and, in the adverts version, advert spend. Low cost cuts both ways: little money is at risk, and anyone else can produce the same AI-written PDF this afternoon.

Step 7: how many sellers reach these numbers

Gumroad published a full count for 2020 from its own records, on 12 January 2021. It is the only full-year earnings count from a platform that we found, and it is nearly six years old⁵. Two newer measures, weekly counts from 2021 and Stan’s averages by audience size, follow the table.

In Gumroad’s words: “the median creator made only $70 in 2020. In fact, the top 1% of creators earned about 60% of the money in 2020”⁵. It adds: “The top 10% of creators earned nearly 92% of the total GMV”⁵. GMV means the total value of sales. Gumroad describes the top group as “A few hundred people, with large audiences and polished products”⁵. “Median” means the middle creator: half earned more and half less. People who listed a product and sold nothing are not in the 45,917, so the true shares are lower⁵. $10,000 a month is $120,000 a year, and $50,000 a month is $600,000. Either would put a seller among the top 0.4% or fewer of 2020’s earners⁵.

Weekly counts from 2021. Gumroad’s newsletters in mid-2021 printed counts for the week from its own records. The issue of 18 July 2021 reads⁶⁶:

Five other issues from June to September 2021 show the same shape: 12,300 to 13,100 creators earning, 590 to 680 over $1,000 and 38 to 47 over $10,000⁶⁶. The people with a good week are not necessarily the same each week. The counts cover everything sold on Gumroad, not only PDFs, and are five years old⁶⁶.

Stan’s averages by audience size (report covering 2025). Stan published figures from its own records in a report that also markets its product⁶⁷. More than 59,000 creators earned money on Stan in 2025, and the average digital download sold for $51⁶⁷. It is the only platform data we found that ties sales to audience size: “The larger your audience is, the higher your average sales typically are”⁶⁷.

These are averages, which a few large sellers pull up, so the typical seller in each row earns less. The report does not say whether sellers with no sales are counted, or over what period⁶⁷. Even the top row is about $16,540 a year, far below the incomes in the tweets (our sum)⁶⁷. A second-hand figure puts Stan at 85,000 active creators in August 2026⁶⁹. Set against the 59,000 who earned in 2025, that suggests roughly 3 in 10 active creators earned nothing (our estimate; the two figures are from different dates and sources)⁶⁷⁶⁹.

Is the market growing? It depends on the platform. Gumroad’s annual report to the US Securities and Exchange Commission shows revenue of $17.79 million in 2025, down from $18.95 million in 2024⁶. Gumroad’s revenue is mainly its percentage of sales, so its sellers’ total sales were probably flat or falling (our inference). Its About page says sellers earned $2,274,585 “last week”⁷, which would be about $118 million a year (our estimate) against $142 million of sales in 2020⁵. The page does not say whether that is before or after Gumroad’s fees; if after, the gap is small or nil.

A second source supports the inference. Sacra reports that sales through Gumroad “fell from its COVID peak of $185M in 2021 to $171M in 2023”, after Gumroad moved to a flat 10% fee in early 2023⁸⁰. Sacra also says Gumroad’s marketplace “doesn’t drive a meaningful amount of demand” and that its smallest sellers have been measured leaving at up to 20% a month; it does not show the basis for that figure⁸⁰. Over its whole life the platform has paid out a large sum: “Gumroad has now paid over $1 billion to creators since launch”, its chief executive wrote in March 2026⁷⁰.

Stan has stopped growing fast. Sacra estimates its growth slowed from 765% in 2023 to 93% in 2024 and 24% in 2025, with sellers leaving about as fast as new ones join⁸².

Whop is growing fast. Its homepage says “$4.6B+ paid to businesses on Whop”⁸. A research firm, Sacra, puts Whop’s lifetime sales at $2.67 billion in February 2026⁹. If the two figures measure the same thing, nearly $2 billion was added in about eight months (our estimate). But much of that is not PDFs. Sacra describes Whop’s sellers as offering courses, memberships and betting tips, and says four of the five top performers in May 2024 sold sports picks or trading signals⁹. Neither platform publishes how much of its sales are ebooks or PDFs.

Whop publishes no middle figure. Sacra reports 183,628 sellers and an “average of $8,413 per month”, an average that a few very large sellers pull up⁹. It also reports that 258 sellers had earned over $1 million on Whop in total by June 2025, about 0.14% of sellers (our sum)⁹.

A warning. Lines such as “89% of Gumroad products never sell” trace back to InsightRaider, a firm that collects public Gumroad pages. Its own report says most of its 146,271 products show no sales count, and “that is missing data, not proof of $0 revenue”²⁵. A second tracker, Profitable, also estimates from public pages²⁴. We treat neither as a measurement. The same goes for InsightRaider’s line that “Only about 2% of products clear $1,000 a month” on Whop, which we saw only as a search extract⁸⁶, and for its list of what winning products share, which rests on the same partial data²⁵. We do not use either.

What people who tried it report

Most of the following are the sellers’ own reports and are not independently audited. One figure was checked by a news outlet: Easlo’s, below.

Correction, 7 October 2026: this page first said none was independently audited. That was wrong for Easlo⁶¹.

Large results from people who had an audience first.

• Daniel Vassallo reported $237,207 of sales and $210,822 of profit in 261 days from a 173-page PDF about Amazon’s cloud service and a 100-minute video. He wrote: “I had a bit over 12,000 Twitter followers when my first product went on the market. Directly or indirectly, every dollar I made in this business can be attributed to this initial audience. I could have created the same two products before I had an audience, and I’m almost certain I wouldn’t have sold any”¹⁴. His second product was a course on building a Twitter audience¹⁴.
• Adam Wathan made $61,392 in three days and $138,835 in 11 months from a programming ebook priced from $29 to $135. He launched “to a list of 1500 people and about 4000 Twitter followers”, after years of free teaching¹⁸.
• Thomas Frank wrote: “My team and I made $1 million selling @NotionHQ templates in a single year (Jan-Dec 2022)”. The team is eight people, and his second YouTube channel grew from 43,000 to 110,000 subscribers that year. He added: “I’ve been creating content since 2010” and “This is a long-game business”¹⁹.

Sales confirmed by platform counters. Gumroad shows a sales or review count on some pages. These are platform records, not screenshots. We did not research these sellers’ audiences, and past prices are unknown, so revenue cannot be worked out.

• Easlo’s $79 template shows 7,071 sales²⁶. Business Insider reported that he made $239,000 from templates in 2022 and wrote that it “verified his revenue with documentation”⁶¹. His path is set out under “The upside”.
• Steph Smith’s ebook, now $99, shows 5,301 sales²⁷. By her own account, in an interview with a section headed “On crossing $130K in 8 months with an ebook”, it made over $130,000 in eight months, launched to about 7,000 followers after years of writing online⁷⁸.
• Daniel Vassallo’s two products show 425 and 132 ratings on his Gumroad store today⁷⁴. Ratings are a fraction of sales and cannot be turned into revenue.
• Modest Mitkus tweeted on 27 April 2023: “I make $16,000/month selling digital products”¹. His Gumroad profile, read today, shows “4.9 from 1701 verified reviews across 7 products” and the description “Build your digital product business”²⁸. The reviews support real sales. The monthly income is his own statement.

None of these is an AI-written PDF sold by a new account.

Ordinary results are small.

• A designer selling five template products on Gumroad, promoted on LinkedIn and Medium, wrote: “I made $427.11 in a year!” His monthly payouts ran from $10 to $50, and a free template had 2,400 downloads and $80 in tips²⁰.
• Another template seller reported: “For the first two months, I had zero revenue”, then $50, then a little over $250 after three months²¹.
• In May 2026 a seller launched four AI-assisted products priced $9 to $47 in three weeks and had no sales: “Decent views. Zero conversions” (no buyers)²². Three weeks is too short to be final.
• The seller with 26 products wrote in April 2026: “26 products on gumroad. zero sales”²³.
• In a September 2020 roll-call of Gumroad sellers on Indie Hackers, one wrote: “Launched it in August, didn’t promote it much, and so far 26 sales ($1,309 total)”⁸⁵. Another had sold 60 books and about 40 advance orders at $26, and a third found sales hard with an email list of just over 300⁸⁵. These are from before AI-written products.

What Gumroad’s survey says. Gumroad surveyed 601 creators in December 2020 and says the sample is not representative. Among those earning over $1,000 in the year, 60% kept an email list, against 31% of lower earners. And: “Only 13.4% of HECs said that 2020 was their first year making money as a creator” (HEC is its label for those higher earners)⁵. The same survey compared 156 higher earners with 213 who earned $1 to $1,000⁵. Higher earners “averaged 6.6 products for sale as opposed to 5”, and were nearly three times as likely to pay other people a share for sales they brought in, though Gumroad cautions that this may be a result of success and not a cause⁵. Of all who replied, 20% said they had replaced their day job⁵.

The recent failures and the older successes point the same way. AI has made the product quick to produce. It has not changed where buyers come from.

We found no source for the share of newcomers who reach $1,000 a month. For the usual time to a first sale we found one survey, of creators of every kind who were still at it: nearly five months to the first dollar⁷⁹. It is described under “What it takes to compete”. Our search for successes by new accounts with AI-written products was limited; we found none with evidence, which is not proof there are none.

What the rules allow now

Selling a file you made, on an ordinary subject, is allowed on every platform we read. The version in many of the tweets meets rules at several points. We found the rules, not how often they are enforced.

A product about getting rich. Gumroad’s prohibited list includes “services (including, but not limited to, “get rich quick” schemes, business opportunities, investment opportunities, mortgage consulting, real estate purchases, mortgage reduction services and credit repair and protection)”, and says the list “may change abruptly and without notice”³⁶. That item is about services. Gumroad does not say whether it covers a guide about earning money, and such guides are on sale there today¹⁴²⁸. Payment companies are plainer. Stripe, used for many online checkouts, will not serve “schemes that claim to offer high rewards for very little effort or up-front work; and sites that promise fast and easy money”⁴². PayPal’s policy also bars “get rich quick” schemes⁴³.

Income claims in the marketing. Whop requires sellers to “Ensure earnings claims aren’t false, deceptive, or misleading in any respect, including by implying that atypical results are typical”³⁸. It lists “untrue income guarantees” as a reason for suspension³⁹.

Resold and copied material. Gumroad bans “reselling private label rights products”, meaning ebooks bought with a licence to resell them as your own³⁶. One tweet suggests copying other accounts’ tweets into a document and selling it on Whop¹. Whop lists selling products that infringe someone else’s copyright as a reason for suspension³⁹.

AI-written text has no copyright in the US. The US Copyright Office wrote in January 2025: “Copyright does not extend to purely AI-generated material”⁴⁴. Selling such an ebook is allowed on the pages we read, but the seller cannot stop others copying and reselling it. Work a person wrote or substantially edited is protected⁴⁴.

Money already earned can be held.

• Gumroad’s terms, updated 14 September 2026, let it keep back 25% of a seller’s funds for 90 days once refunds pass 15%, and suspend or close the account above 25%³⁷. They also allow a hold with “no fixed or maximum hold period” if Gumroad believes the seller’s marketing is “misleading, deceptive or otherwise create risk to Buyers”³⁷.
• Whop says that when it suspends an account “it holds the balance for 90 days by default” and “doesn’t guarantee their release afterward”³⁹.
• A dispute is when a buyer asks their bank to reverse a payment. Whop charges $15 for each⁴, so one dispute on a $19 sale costs $34 (our sum). Whop says card networks expect disputes to stay under 1.5% of sales; above that, or on a new account with little history, it can keep back part of the money⁴⁰. Stripe’s guide gives the same lines: “any chargeback rate above 1% means a business might be deemed high-risk”, and Mastercard’s monitoring programme starts at 1.5%⁸⁴. A chargeback is another word for a dispute.

Tax and refunds. Gumroad has collected and paid sales tax and VAT (the European sales tax) itself since 1 January 2025³. On Whop that is the seller’s job unless they pay 2% per sale for a managed option⁴¹. In the US, self-employment tax of 15.3% starts once net earnings reach $400 in a year⁴⁹. EU buyers can cancel most online purchases within 14 days; for a download the right is lost only if the buyer “expressly agreed to lose your right of withdrawal” first⁴⁸.

Automated accounts on X. Some tweets advise posting hundreds of AI-written posts and letting a bot answer every direct message¹. We could not open X’s own policy pages⁴⁵. News sites reported that X removed 42,000 chatbot-reply accounts on 24 and 25 July 2026⁴⁶.

US regulator action. We found no action against selling PDFs direct to an audience as such. The Federal Trade Commission (FTC), the US consumer regulator, has acted against sellers of courses on related methods. All three cases below were settled; they are the FTC’s allegations, not court findings. These cases concern the companies named in them only. We found no regulator action against any account quoted or named on this page, and we do not suggest any connection.

• Publishing.com, 2026. The company sold courses on publishing AI-assisted ebooks and audiobooks on Amazon, typically for $1,995³². The FTC’s complaint alleged that most buyers “never achieved the income the company promised in its advertising”, which was “$1k to $3k a month in passive income”³⁰. The company agreed to pay $1.5 million and settled without a court ruling on the allegations. The order became final on 2 July 2026³¹.
• Lurn, 2023. The offers included a $947 course on making money from ebooks sold on Amazon. The FTC alleged that “very few, if any consumers actually made money with these programs”. The case was settled without a court ruling on the allegations: Lurn and its chief executive agreed to turn over $2.5 million, and the FTC sent more than $2.4 million in refunds in June 2024³³.
• Growth Cave, 2025 and 2026. Buyers paid $3,500 to $9,800 to create an online course and sell it through YouTube adverts. The complaint alleges that in most instances buyers of the programme “do not make a single sale and therefore do not earn any income”³⁴. The defendants settled in January 2026, without a court ruling on the allegations³⁴.

Who makes money from it

What the tweets lead to. Of the 54 tweets, 15 show nothing for sale in the text we stored. The other 39 steer the reader somewhere: a course, a training video, a community, a tool or something sent by direct message¹. The counts rest on labels added when the tweets were collected and are approximate.

Thirteen tweets ask the reader to follow, repost or reply to get something free. They hold about 6.62 million of the 10.18 million views, roughly 65%¹. The immediate gain from such a post goes to the poster: followers, and a list of people to message.

Of the 30 most-viewed tweets, eight state a price of some kind: six for a product, one for access to a class and one for a community. In the text we stored, none gives the number of sales, the number of visitors, advert spend, fees, refunds, hours worked, or whether the figure is sales or profit, and none links to a record checked by a payment company¹. The stored text is partly cut short, and we did not view attached images, videos or threads, so any of these may be given there.

The accounts behind the example tweets. These are facts from the accounts’ own pages, read through a copy of X, not x.com, and one pointer to press coverage. They show what each account offers, not whether its income claim is true.

• @ChrisClickUp (5.5 million views, May 2023). The tweet gives away a prompt pack “Normally $39” for a follow, a repost and a reply¹. Its text states no income of the author’s own and offers the pack free, not for sale⁵⁰.
• @X_FINALBOSS wrote 9 of the 54 tweets, with claims from “$7000+ made while I slept” to “$500K+ PER MONTH”¹. The account’s display name is Alex Suzuki, and its profile reads “$5M+/yr digital products on X” and points to a “2 hr training video”⁵¹.
• @TommiPedruzzi and @Manu_Sisti, the two “$19 PDF” tweets. Their profiles name the same company, Nespola, as founder and co-founder, and both link to the same free workshop, PublishingOS, which is about publishing books on Amazon⁵². Its page says the seat is free and points to a sales call. Prices of the paid programmes are not published⁵². The workshop’s earnings disclosure, updated 12 September 2026, says of its showcased results: “They are not typical and are not what most participants achieve”, and “You should assume you may earn £0 / $0”⁵³.
• Press coverage. The news site Pirate Wires published an article in 2025 about Tommi Pedruzzi’s Amazon ebook business and its course²⁹. We read only the part outside the paywall, so we do not summarise its findings. It concerns a different business from the “$19 PDF” tweet and does not test that claim.

The platforms are paid on every sale whether or not the seller comes out ahead: about 17% of a $19 sale on Gumroad, or 30% through its marketplace², and 2.7% plus $0.30 and payout fees on Whop⁴. They also pay for bringing in new sellers. Whop: “Earn 30% of Whop’s profit from businesses you refer”⁵⁵. Skool, a platform for paid communities: “Refer people to Skool and earn 40% of monthly recurring revenue for life”⁵⁶. We did not check whether any tweet carries a referral link.

The course about selling the course. In one variant the product is a course that the buyer is licensed to resell. Truth in Advertising, a non-profit watchdog, examined one that “costs $487” and another at “either $497 or $597” where resellers “keep 85% of the profit”⁵⁷⁵⁸.

The upside

People do earn real money selling files this way. This section sets out the best outcomes we could document. Each is marked with who checked the figure, and with whether the person also sells teaching on earning money or growing an audience. None of them is the tweets’ version: we found no evidenced case of a new account earning five figures a month from an AI-written PDF.

The best documented outcomes

Checked by a third party

• Easlo (Jason Chin), templates for the note-taking app Notion: $239,000 in 2022. Business Insider published his account and wrote: “Insider verified his revenue with documentation”⁶¹. It does not say which documents, or whether the figure is before or after fees⁶¹. The rest is his own account. He gave templates away for about two months in summer 2021 with tips switched on, charged from November 2021, and emailed everyone who had downloaded a free one⁶¹. The first paid month brought over $3,000⁶¹. One $130 product made over $100,000 of the 2022 total⁶¹. He never paid for adverts and calls the work “definitely a full-time commitment”⁶¹. Gumroad’s own blog dates the start. In December 2021 it wrote that eight months earlier he “didn’t have an audience online or had never generated a dollar online”, and that he now had “nearly 18k followers on Twitter” and had “made over $25k online”⁶². So the path was over $25,000 in the first eight months and $239,000 in the second year⁶¹⁶². By December 2022 he had over 211,000 Twitter followers⁶¹. Does he sell teaching on the subject? Partly. His products are templates, but he said in 2022 that his tweets “focus on helping others grow their audiences and ventures”⁶¹. In 2024 interviews he put his income at $20,000 a month with 356,000 followers on X, which is his own statement⁷⁷.
• Marc Lou, a code template called ShipFast: $1,272,045 since July 2023, as read on 7 October 2026. The figure is on TrustMRR, a site that reads revenue from the payment company Stripe: “Revenue is verified with Stripe API key”⁷¹. It is a live page, so the figures move⁷¹. The product is starter code for programmers, not a PDF. It costs $199 or $249, or $299 in a bundle with his coding course⁷¹. The last 30 days showed $1,793 on the same date, so most of the income came earlier⁷¹. He has 406,800 followers on X and also sells a coding course, so he does sell teaching to the same audience⁷¹. Correction, 7 October 2026: this page first gave $1,271,746 and $1,494, from a summary made by our fetch tool. The figures here are from the raw page, read later the same day⁷¹.

The seller’s account of a total the platform told him

• Brian Hahn, study outlines for the US bar exam: $1 million over about ten years. He is a patent attorney who started in 2014 and ran this beside a law-firm job⁶³. He wrote: “Gumroad informed me I’d surpassed $1 million in revenue. They were right. I checked”⁶³. That is roughly $100,000 a year on average (our estimate). He wrote that he has four digital products as his main source of revenue⁶³. His Gumroad store today lists six, priced from $35 to $197⁷⁴. His buyers come from “word-of-mouth referrals along with Google SEO and engaging on platforms like Reddit” (SEO means being found through search), not from a large social following⁶³. He does not sell a course on selling digital products⁶³. He also wrote: “digital products are NOT a source of passive income”⁶³. His $197 product shows 155 ratings on Gumroad today⁷⁴.

The seller’s own figures

• Philip Kiely, a book called “Writing for Software Developers”: over $25,000 in three months. He was a computer-science student who had written paid tutorials for developer magazines⁶⁴. He emailed 100 well-known developers and put 11 interviews in the book⁶⁴. It made $20,000 in the first 24 hours after he posted it to Hacker News, a discussion site for programmers⁶⁴. Gumroad’s blog reports: “he’s sold 700 copies, earning a Net Income of more than $25,000”⁶⁴. That is about $36 a copy (our sum: $25,000 / 700). The figures are his, relayed by the platform, and the post does not say Gumroad checked them⁶⁴. The book is about writing, not about selling digital products⁶⁴.
• Steph Smith, an ebook about writing online: over $130,000 in eight months. The period is from a section heading in the interview⁷⁸. She wrote: “I already had maybe 7,000 followers before launching and also spent years writing online before creating this”⁷⁸. She asked in a tweet whether people would pay, took orders before the book existed, raised the price in steps as copies sold, and paid other people a share for sales they brought in⁷⁸. Gumroad’s counter of 5,301 sales is a platform record that fits her account²⁷. She later sold a productivity course to the same buyers⁷⁸.
• LifeMathMoney: $1 million of Gumroad sales in total. This is his own statement in a guest post on Gumroad’s blog⁶⁵. He wrote: “I have ~400k followers on X, a blog that gets hundreds of thousands of readers, and a fairly big email list”⁶⁵. He wrote that one self-improvement guide, Live Intentionally, has over 12,500 sales⁶⁵. His store lists it today from $19.99⁷⁴. His two main products show 474 and 500 ratings on Gumroad today⁷⁴. He does sell a product on the subject: “The Art of X: Build a Business That Makes You $100/Day”, listed today from $89.99⁷⁴. He wrote in January 2025 that its top variant costs $297⁶⁵.
• A healthcare professional named in Stan’s report. Stan’s 2025 report says a seller with under 10,000 followers made over $250K in eight months; its own profile of her says $180K in seven months, which may be an earlier date⁶⁷⁶⁸. The profile says she sells to other healthcare providers, so by Stan’s account the edge was professional knowledge and a narrow group of buyers, not reach⁶⁸. The figures are Stan’s and were not independently checked.
• Daniel Vassallo, Adam Wathan and Thomas Frank are set out under “What people who tried it report”: $237,207 in 261 days, $138,835 in 11 months and $1 million in a year with a team of eight¹⁴¹⁸¹⁹. All three are the sellers’ own reports. Vassallo’s second product was a course on building a Twitter audience¹⁴.

The two checked cases and Hahn’s sold a subject: templates, code, exam outlines. Of the sellers above, LifeMathMoney sells a guide to earning money, and Vassallo, Easlo and Marc Lou sell or post some teaching on audience growth or their craft^14616571.

What the top tenth and the top hundredth earn

Gumroad’s 2020 report says: “the top 1% of creators earned about 60% of the money in 2020. The top 10% of creators earned nearly 92% of the total GMV”⁵. GMV means the total value of sales. From those shares, the $142 million of sales and the 45,917 earning creators, we worked out four figures⁵. All four are our estimates, not Gumroad’s.

The averages are pulled up by the largest sellers. The entry figures are our estimate, read off between the bands Gumroad published, so they are rough⁵. Read plainly: about one earning seller in ten made a few thousand dollars or more in the year, and about one in a hundred made a wage or more⁵.

Step 7 has two newer measures: Gumroad’s weekly counts from 2021, in which about 5% of the creators who earned in a week made over $1,000 that week (our sum)⁶⁶, and Stan’s averages by follower count, from its report covering 2025⁶⁷.

A neighbouring field hints at what time does. An anonymous poster’s analysis of 5,079 small software projects with Stripe-read revenue on TrustMRR found a middle figure of $169 a month, with the top tenth above $10,000 a month⁷². The middle project earned $148 a month in its first year and $2,399 in its fifth⁷². That is software, not PDFs. The projects chose to list themselves, the later years hold only the survivors, and we did not reproduce the analysis⁷². It is a hint that those who last several years earn far more, nothing stronger.

A realistic good result

This is our judgement from the sources above, not a measured group.

• Year one, for a competent person with a real subject: a few hundred to a few thousand dollars. That is about where the top tenth began on Gumroad in 2020 (our estimate)⁵. It fits Stan’s averages of about $1,070 a year under 1,000 followers and about $3,280 a year at 1,000 to 10,000 (our sums; averages run above the typical seller)⁶⁷. It also fits ordinary sellers’ reports, such as 26 sales and $1,309 in the weeks after a launch⁸⁵.
• Five figures in year one or two is reachable for people who build an audience or hold a credential buyers trust. Easlo passed $25,000 in his first eight months⁶². Kiely passed $25,000 in three months⁶⁴. Both are well above the usual result: about 4% of Gumroad’s earning creators made $10,000 or more in 2020 (our sum)⁵.
• Six figures a year took an audience built before the product, a credential buyers trust, or many years in every case we found: Easlo in his second year with over 211,000 followers⁶¹, Vassallo with about 12,000 followers built first¹⁴, Hahn at about $100,000 a year averaged over ten years (our estimate)⁶³, Frank after creating content since 2010¹⁹. Wathan and Steph Smith passed $100,000 with about 4,000 and about 7,000 followers, after years of free teaching or writing¹⁸⁷⁸.

What it takes to compete

A low price needs volume and a high price needs trust. Both need buyers the seller brings, because the platforms do not supply them to a newcomer⁷⁵⁸¹.

What the people who succeed have

• They knew something specific that people pay for, often for work: a cloud service¹⁴, a programming method¹⁸, a productivity tool¹⁹, a professional exam⁶³, writing for programmers⁶⁴.
• They had an audience before the product. About 12,000 followers¹⁴. An email list of 1,500 after years of free teaching¹⁸. Twelve years of making content and a team¹⁹. About 7,000 followers and years of writing⁷⁸. Easlo began posting before he had anything to sell, on the advice to “create an audience before having something to sell”⁷⁷.
• Or they had a credential and a narrow group of buyers in place of reach. A patent attorney selling to bar-exam candidates through search and word of mouth⁶³. A student who had already been paid to write for developers, with interviews of known developers in the book⁶⁴. A healthcare professional selling to other healthcare providers, by the platform’s account⁶⁸.
• They gave work away first. Easlo’s free templates came before his paid ones, and his first buyers were the people who had downloaded them⁶¹. Wathan taught for free for years¹⁸.
• They kept an email list and several products. In Gumroad’s unrepresentative survey, 60% of higher earners kept an email list against 31% of lower earners, and higher earners had 6.6 products for sale against 5⁵.
• The audience took time. Most higher earners in that survey were not in their first year of earning⁵. The fastest documented cases took about eight months: Vassallo built his audience in about eight months first, with a senior Amazon background¹³¹⁴, and Easlo took eight months to reach about 18,000 followers and $25,000⁶². Hahn took ten years⁶³. No source measures the usual time for PDF sellers.
• The product was about a subject, not about getting rich^1418196364. That version stays clear of the payment companies’ and Whop’s rules on easy-money promises and income claims³⁸⁴².

What it costs

Cash: very little. Gumroad has no monthly fee². An email list is free to start: MailerLite is free up to 250 subscribers and then starts at $12 a month, and beehiiv is free up to 2,500 subscribers⁸³. With the tools in step 6 the bill is about $50 a month or less (our sum)¹⁶¹⁷.

Time: months before the first dollar, and more than a year before a living, among those who stayed. The Tilt, a company that sells education for creators, surveyed 1,043 creators in 2023⁷⁹. It wrote: “it takes nearly five months to earn the first dollar and just over a year to begin working full time as a creator”, and supporting yourself “takes over 18 months”⁷⁹. Those creators said they spent $10,700 on average to start⁷⁹. The survey covers content businesses of every kind, not PDF sellers. It was recruited through creator newsletters, and people who gave up are not in it⁷⁹. Easlo calls his template business a full-time commitment and in 2024 said he spent two full days a week on content⁶¹⁷⁷.

More competition, by one seller’s account. Easlo said in 2022: “More competition has emerged in the template market, and I’ve had to adapt Easlo around that”⁶¹. That was before AI-written products. Sacra reports that sales through Gumroad fell from a peak of $185 million in 2021 to $171 million in 2023⁸⁰, still above the $142 million of 2020⁵. We found no newer count of sellers, so we cannot say how many people share that total.

What a newcomer can reasonably expect

This is our estimate, not a measured figure. With no particular expertise, no audience and an AI-written general PDF: nothing to under $100 in a year. The $70 middle figure counts only sellers who sold something, 55% of those earned under $100 (our sum), and the two recent no-audience reports we found had no sales⁵²²²³. Stan’s averages put sellers with under 1,000 followers at $89 a month, but that is an average from a company that sells the tool, and it does not say whether sellers with no sales are counted⁶⁷. With real expertise and months of useful free posting first, a few hundred to a few thousand dollars in the first year is plausible. The basis is the $427 year²⁰, the 17% of earning creators above $1,000⁵, the top-tenth estimate above⁵ and the launches above. That is a judgement; nobody has measured it.

How to tell early which side of the line you are on

1. Before making anything, count who is listening. Can you name the subject you would be trusted on? Do strangers already ask you about it? If not, the first job is months of free, useful posts.
2. Ask before you build. Steph Smith’s book began as a tweet asking whether people would pay, followed by orders taken in advance⁷⁸. If nobody says yes, that is the answer for now.
3. Check your reach. At about $3 of sales per 1,000 views, an expert’s result, $1,000 a month needs roughly 330,000 views a month (our estimate from¹³). If your posts get a few hundred views each, the sales arithmetic does not start yet.
4. Look at where page visits come from. If every visit comes from a link you posted yourself, the platform is sending no one. That is normal on Gumroad until the account has $100 of genuine sales and has passed a review⁷⁵.
5. Give something away first. If a free version gets downloads but no replies, questions or sign-ups, a paid one is unlikely to sell²⁰. A good sign is the one Easlo describes: “I began receiving tips of up to $100 on Gumroad” for free templates⁶¹.
6. Watch for the first real line: three or four genuine sales adding up to $100. That is what Gumroad’s review typically needs before it pays out or lists a product in its marketplace⁷⁵⁷⁶. As a rough yardstick for time, creators who stayed took about five months to their first dollar⁷⁹.
7. Treat adverts as a test with a fixed budget. The one documented case where they paid was aimed at a narrow community that already wanted the subject, when clicks were cheap¹⁴. At around $0.60 a click, fewer than 4 buyers in 100 clicks means each $19 sale loses money (our estimate from¹²).
8. Keep disputes rare. Stripe says a dispute rate above 1% can get a business treated as high-risk⁸⁴.
9. Be wary when the product is about making money. Payment companies restrict easy-money promises⁴²⁴³, it may fall under Gumroad’s list³⁶, and regulators have acted against sellers of courses on related methods³⁰³³³⁴.

None of these is a measured threshold for PDF sellers.

Trying without adverts or a course costs almost no cash. The stake is time. The costly mistakes are advert spend, and paying four figures for training before a first sale.

What we did not verify

• The income claims. None of the claims in the tweets was verified or refuted. We found no evidenced case of a new account succeeding with an AI-written PDF. We found no payment-verified record for any. We read the stored tweets, and the full text of four through a copy of X, not the threads, links, attached images or videos.
• The tweet counts rest on labels added at collection. Nobody re-opened each profile.
• Current earnings on Gumroad. The only full-year count is for 2020 and the weekly counts are from 2021, both before AI-written products. Whether their shares still hold is unknown. We did not watch Gumroad’s public board meetings, which may hold newer figures. The entry figures for the top tenth and top hundredth are our estimate, read off between the published bands, not published numbers.
• The success stories. Only Easlo’s $239,000 was checked by a third party, and Business Insider does not say which documents it saw or whether the figure is before fees. We read its article in the Yahoo Finance copy because businessinsider.com refused us⁶¹. His own accounts differ on how long he gave templates away before charging; we used the two months in the article. Gumroad’s blog posts on Kiely, LifeMathMoney and Easlo’s first $25,000 do not say Gumroad checked the figures. Ratings counts on store pages are not sales counts, and today’s store prices may differ from past ones. The ShipFast figures are from a live page and will have moved. The analysis of 5,079 software projects is one poster’s work.
• Stan. The follower table gives averages with no middle figures, and does not say who is counted or how followers were measured. The two Stan pages on the healthcare seller give different figures. The 85,000 active creators and a figure of over $600 million earned by creators in total come second-hand from Contrary Research⁶⁹. Stan’s price comes from Sacra.
• The Tilt survey is unweighted, covers content businesses broadly and leaves out people who quit⁷⁹.
• Whop. No middle figure exists. The seller counts and the average come from Sacra, not Whop. Whop’s fees page lists no percentage platform fee or marketplace fee for new sellers; Sacra describes earlier ones and says the marketplace fee was cut to 0%⁴⁹. A Sacra summary dates the cut to 12 May 2025; Whop’s own changelog was not opened⁸¹. The 97% direct share is also Sacra’s. Whether the two security-check fees apply to every sale is not stated.
• Buying rates and adverts. No measured buying rate for PDFs at any price. The advert benchmarks are second-hand. We found nothing on the Nigerian version in one tweet (a ₦9,800 PDF sold through Facebook adverts).
• Refunds and disputes. No data on rates for digital products, or on how often platforms hold money. Figures seen in search results came from vendor blogs we did not open, and we do not use them.
• Pages we could not open. X’s policy pages, refused by three routes⁴⁵. Meta’s rule on adverts that promise unrealistic outcomes⁴⁷. Etsy’s rules on AI-made items, relevant because two tweets name Etsy⁵⁹. The pricing pages of X Premium, Kit and Stan, so the cost of paid reach on X is missing. Gumroad’s filing⁶ refused a second request today. The Kit creator-economy report, which we read second-hand only: a summary says 18% of the “over 1,000 creators” surveyed earn over $100,000 a year, from a sample of users of a paid email tool that is far from typical and covers all creator income⁷³. The Business Insider article on Easlo, listed here before as unopened, has now been read in a syndicated copy⁶¹.
• Summaries, not wording. Notion’s marketplace terms (8% plus $0.40 a sale, and an approval that may take a few months) reached us through our fetch tool’s summary and are not used in the sums⁸³.
• Pirate Wires. Read above the paywall only, through the r.jina.ai reader; a direct request was refused today, and again when the wording on this page was reviewed. We do not summarise its findings, and any reply from the person it is about is unseen. Its date is August or September 2025: the text calls a tweet of 11 August 2025 “earlier this month”, and one of our checks read 9 September 2025 in the page’s code.
• The FTC cases. We read press releases and complaints, not the final orders. The FTC’s 2021 warning on money-making claims and its Business Opportunity Rule do not say whether a cheap ebook counts as a “business opportunity”³⁵. One 2026 FTC settlement was read second-hand only, in a law-news report, and is not used on this page⁶⁰.
• A checkout page found by search carries the same display name as @X_FINALBOSS but shows no X handle⁵¹. We could not confirm whose it is, so we do not treat it as that account’s offer and do not describe it.
• Other accounts in the collection. We did not research what the other accounts sell, beyond what is stated above.
• Not researched. Payhip, Lemon Squeezy, SendOwl and TikTok. Stan was read for this update, from its own report only. Tax and consumer rules outside the US and EU. Regulators outside the US. Whether AI-written PDFs are refunded more often. A fuller search for new-account successes.

Sources

1. Does It Pay collection of 54 tweets on this scheme, collected 2026-10-06. View counts as collected; text as stored, partly cut short. Tallies are our own counts (estimate).
2. Gumroad Help, “Gumroad’s fees”. https://gumroad.com/help/article/66-gumroads-fees. Undated. Read at source on 2026-10-07.
3. Gumroad, Pricing. https://gumroad.com/pricing. Undated. Read at source on 2026-10-07.
4. Whop Docs, “Fees”. https://docs.whop.com/fees. Undated; “applies to new merchants that sign up today”. Read at source on 2026-10-07.
5. Gumroad, “Last Year in the Creator Economy”. https://gumroad.com/gumroad/p/last-year-in-the-creator-economy. Early 2021, covering 2020 (date inferred). Read at source on 2026-10-07. Re-read on 2026-10-07 at https://gumroad.gumroad.com/p/last-year-in-the-creator-economy, where the date shows as 2021-01-12.
6. Gumroad, Inc., Form C-AR annual report to the US Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/1532978/000153297826000003/primary_doc.xml. Filed 2026-04-08, for the year to 2025-12-31. Read at source on 2026-10-07.
7. Gumroad, About. https://gumroad.com/about. Live page. Read at source on 2026-10-07.
8. Whop, homepage. https://whop.com/. Live page. Read at source on 2026-10-07.
9. Sacra, research page on Whop. https://sacra.com/c/whop/. Undated; mentions events to March 2026. Secondary: relays Whop’s reported numbers and Sacra’s own estimates.
10. Indie Hackers, “What’s your conversion rate on Gumroad?”. https://www.indiehackers.com/post/what-s-your-conversion-rate-on-gumroad-9dc97be25f. 2021-06-15. Read at source on 2026-10-07; sellers’ own reports.
11. Low Content Profits, “How I Got 118 Sales in 30 Days on My Gumroad Store”. https://lowcontentprofits.com/gumroad-passive-income/. 2025-10-06, updated 2026-01-25, covering September 2025. Read at source on 2026-10-07; seller’s own report.
12. Coupler.io, “Facebook Ads Statistics”. https://blog.coupler.io/facebook-ads-statistics/. 2026-09-25. Secondary: summarises WordStream by LocaliQ and Triple Whale benchmarks, which we did not open.
13. Daniel Vassallo, “I sold $104,754 worth of videos and PDFs in 4 months”. https://www.indiehackers.com/post/i-sold-104-754-worth-of-videos-and-pdfs-in-4-months-1f8a41339d. 2020-04-27. Read at source on 2026-10-07; seller’s own report.
14. Daniel Vassallo, “How I made $210,822 selling a PDF and a video on the internet”. https://www.indiehackers.com/post/how-i-made-210-822-selling-a-pdf-and-a-video-on-the-internet-028d5b0a2c. 2020-09-14. Read at source on 2026-10-07; seller’s own report with dashboard charts.
15. Socialinsider, social media benchmarks including X. https://www.socialinsider.io/blog/twitter-benchmarks/. 2026-01-16, updated 2026-07-15 with data to the second quarter of 2026. Read at source on 2026-10-07; a vendor study of brand accounts.
16. Tweet Hunter, Pricing. https://tweethunter.io/pricing. Live page. Read at source on 2026-10-07.
17. Claude, Pricing. https://claude.com/pricing. Live page. Read at source on 2026-10-07.
18. Adam Wathan, “The Book Launch That Let Me Quit My Job”. https://adamwathan.me/the-book-launch-that-let-me-quit-my-job/. April 2017. Read at source on 2026-10-07 through the r.jina.ai reader; seller’s own report.
19. Thomas Frank, “$1 million in Notion template sales”. https://typefully.com/TomFrankly/dollar1-million-in-notion-template-sales-kuFT0iD. About January 2023 (date inferred). Read at source on 2026-10-07; seller’s own report.
20. UX Planet, “How much I made by selling digital products for 1 year”. https://uxplanet.org/how-much-i-made-by-selling-digital-products-for-1-year-4429d9a2cce4. 2023-08-27. Read at source on 2026-10-07 through the r.jina.ai reader (the direct request was refused); seller’s own report.
21. The Mindful Planner, “How I made $250 selling Notion templates”. https://themindfulplanner.substack.com/p/how-i-made-250-selling-notion-templates. 2024-10-22. Read at source on 2026-10-07; seller’s own report, no proof shown.
22. Indie Hackers, “I launched 4 digital products in 3 weeks. 0 sales”. https://www.indiehackers.com/post/i-launched-4-digital-products-in-3-weeks-0-sales-heres-what-i-m-actually-learning-772e569ddb. 2026-05-22. Read at source on 2026-10-07 through the r.jina.ai reader; seller’s own report.
23. Indie Hackers, “Gumroad has zero organic discovery and that’s actually fine”. https://www.indiehackers.com/post/gumroad-has-zero-organic-discovery-and-thats-actually-fine-e68bd40f4e. 2026-04-01. Read at source on 2026-10-07; seller’s own report and replies.
24. Profitable, Gumroad statistics. https://profitable.app/gumroad/stats. Updated 2026-09-03. Read at source on 2026-10-07 through the r.jina.ai reader; the tracker’s estimates from public pages.
25. InsightRaider, “State of Gumroad 2026”. https://insightraider.com/en/state-of-gumroad-2026. Published 2026-05-21. Read at source on 2026-10-07; a third-party collection of public pages, not Gumroad’s records.
26. Easlo, Gumroad store. https://easlo.gumroad.com/. Live page. Read at source on 2026-10-07.
27. Steph Smith, “Doing Content Right”, Gumroad page. https://stephsmithio.gumroad.com/l/doing-content-right. Live page. Read at source on 2026-10-07.
28. Modest Mitkus, Gumroad profile. https://modestmitkus.gumroad.com/. Live page. Read at source on 2026-10-07.
29. Pirate Wires, “Inside the Amazon Slop King’s $3M Hustle” (the headline is Pirate Wires’ own wording). https://www.piratewires.com/p/inside-the-amazon-slop-kings-3m-hustle. August or September 2025. Read at source on 2026-10-07 through the r.jina.ai reader, above the paywall only; journalism, with revenue figures as stated to the reporter.
30. US Federal Trade Commission, press release on the Publishing.com settlement. https://www.ftc.gov/news-events/news/press-releases/2026/04/publishingcom-pay-15-million-misleading-consumers-about-how-much-income-they-could-earn-using. 2026-04-13. Read at source on 2026-10-07.
31. US Federal Trade Commission, press release on the final Publishing.com order. https://www.ftc.gov/news-events/news/press-releases/2026/07/ftc-approves-final-order-against-publishingcom-settling-allegations-it-misled-consumers. 2026-07-02. Read at source on 2026-10-07.
32. US Federal Trade Commission, complaint against Publishing.com. https://www.ftc.gov/system/files/ftc_gov/pdf/2423055publishingcomcomplaint.pdf. Released 2026-04-13. Read at source on 2026-10-07.
33. US Federal Trade Commission, press releases on the Lurn case. Settlement, 2023-09-28: https://www.ftc.gov/news-events/news/press-releases/2023/09/ftc-acts-stop-online-business-coaching-scheme-lurn-deceiving-consumers-about-money-making-potential. Refunds, 2024-06-06: https://www.ftc.gov/news-events/news/press-releases/2024/06/ftc-sends-more-24-million-consumers-harmed-deceptive-business-coaching-scheme-lurn. Complaint: https://www.ftc.gov/system/files/ftc_gov/pdf/FiledLurnComplaint.pdf. Read at source on 2026-10-07.
34. US Federal Trade Commission, complaint against Growth Cave. https://www.ftc.gov/system/files/ftc_gov/pdf/GrowthCave-Complaint.pdf. 2025-02-10. Case page with press release titles: https://www.ftc.gov/legal-library/browse/cases-proceedings/growth-cave-llc. Settlement release, 2026-01-27: https://www.ftc.gov/news-events/news/press-releases/2026/01/ftc-secures-settlement-banning-growth-cave-defendants-marketing-selling-business-opportunities. Read at source on 2026-10-07.
35. US Federal Trade Commission, notice on money-making claims. https://www.ftc.gov/news-events/news/press-releases/2021/10/ftc-puts-businesses-notice-false-money-making-claims-could-lead-big-penalties. 2021-10-26. Business Opportunity Rule guidance: https://www.ftc.gov/business-guidance/resources/selling-work-home-or-other-business-opportunity-revised-rule-may-apply-you-1. Read at source on 2026-10-07.
36. Gumroad, Prohibited products. https://gumroad.com/prohibited. Undated. Read at source on 2026-10-07.
37. Gumroad, Terms of Service. https://gumroad.com/terms. Last updated 2026-09-14. Read at source on 2026-10-07.
38. Whop Docs, Community guidelines. https://docs.whop.com/trust-and-safety/trust-safety-overview/community-guidelines. Undated. Read at source on 2026-10-07.
39. Whop Docs, Account suspensions. https://docs.whop.com/trust-and-safety/suspensions/account-suspensions. What is not allowed: https://docs.whop.com/trust-and-safety/trust-safety-overview/what-is-not-allowed-on-whop. Undated. Read at source on 2026-10-07.
40. Whop Docs, Payment health. https://docs.whop.com/trust-and-safety/account-health/payment-health. Reserves: https://docs.whop.com/trust-and-safety/account-health/reserves. Undated. Read at source on 2026-10-07.
41. Whop Docs, Taxes. https://docs.whop.com/payments-and-billing/fees/taxes. Undated. Read at source on 2026-10-07.
42. Stripe, Prohibited and restricted businesses. https://stripe.com/legal/restricted-businesses. Last updated 2026-09-22. Read at source on 2026-10-07.
43. PayPal, Acceptable Use Policy. https://www.paypal.com/us/legalhub/paypal/acceptableuse-full. Last updated 2022-10-29. Read at source on 2026-10-07.
44. US Copyright Office, “Copyright and Artificial Intelligence, Part 2: Copyrightability”. https://www.copyright.gov/ai/Copyright-and-Artificial-Intelligence-Part-2-Copyrightability-Report.pdf. January 2025. Read at source on 2026-10-07.
45. X, Authenticity policy. https://help.x.com/en/rules-and-policies/authenticity. Reported as updated April 2025. Secondary: the page refused us by three routes; wording is from a search-engine extract.
46. Roboin, article on X account suspensions. https://roboin.io/article/en/2026/04/05/mass-account-suspensions-on-x-in-april-2026-timeline-and-root-causes-explained/. 2026-04-05. Let’s Data Science, on 42,000 removed accounts: https://letsdatascience.com/news/x-says-it-removed-42000-chatbot-reply-accounts-0e994445. 2026-07-27. Secondary: the original X posts were not opened.
47. Meta, advertising standard on unrealistic outcomes. https://transparency.meta.com/policies/ad-standards/deceptive-content/unrealistic-outcomes. Undated. Secondary: the page returned “not found”; wording is from a search-engine extract.
48. European Union, Your Europe, “Guarantees and returns”. https://europa.eu/youreurope/citizens/consumers/shopping/returns/index_en.htm. Last checked by the EU 2026-04-28. Read at source on 2026-10-07.
49. US Internal Revenue Service, Self-employment tax. https://www.irs.gov/businesses/small-businesses-self-employed/self-employment-tax-social-security-and-medicare-taxes. Undated. Read at source on 2026-10-07.
50. X profile of @ChrisClickUp, as shown with the tweet on the fxtwitter copy of X. https://api.fxtwitter.com/ChrisClickUp/status/1658200590550278144. As shown on 2026-10-07. Read at source on 2026-10-07 through that copy, not x.com.
51. X profile and full tweets of @X_FINALBOSS on the fxtwitter copy of X. https://api.fxtwitter.com/X_FINALBOSS/status/1975610300217860344 and https://api.fxtwitter.com/X_FINALBOSS/status/2063317534430920876. As shown on 2026-10-07. Checkout page: https://ebook-vault.com/checkout-page (footer dated 2024; found by search, not from the profile). Read at source on 2026-10-07.
52. PublishingOS workshop page. https://publishingos.io/begin?htrafficsource=X&el=X-Bio-Manu. Undated. Profiles of @TommiPedruzzi and @Manu_Sisti on the fxtwitter copy of X: https://api.fxtwitter.com/TommiPedruzzi/status/2101648552351191192 and https://api.fxtwitter.com/Manu_Sisti/status/2105696120450216102. Read at source on 2026-10-07.
53. PublishingOS, Earnings disclosure. https://publishingos.io/begin/disclaimer. Last updated 2026-09-12. Read at source on 2026-10-07.
54. Pascio, newsletter sign-up page. https://sidequests.pascio.com/subscribe. Live page. Read at source on 2026-10-07.
55. Whop Docs, “Refer businesses to Whop”. https://docs.whop.com/refer-businesses-to-whop.md. Undated. Read at source on 2026-10-07.
56. Skool, Affiliate program. https://www.skool.com/affiliate-program. Live page. Read at source on 2026-10-07.
57. Truth in Advertising, on a $487 resell-rights course. https://truthinadvertising.org/articles/make-money-with-fran-passive-digital-mastery/. 2025-04-29. Read at source on 2026-10-07.
58. Truth in Advertising, on Digital Wealth Academy. https://truthinadvertising.org/articles/rachell-medero-digital-wealth-academy/. 2025-08-18. Read at source on 2026-10-07.
59. Etsy, Creativity Standards. https://www.etsy.com/legal/creativity. Not opened: the page refused us directly and through the reader. Listed only as a page we could not read.
60. National Law Review, on the FTC’s settlement with Air AI. https://natlawreview.com/article/company-banned-marketing-business-opportunities-settle-ftc-charges-it-misled. 2026-03-26. Secondary; the FTC’s own release was not found.
61. Business Insider, as-told-to essay by Easlo on making $239,000 from Notion templates, in the Yahoo Finance copy. https://finance.yahoo.com/news/made-239-000-selling-budgeting-120626735.html. Archived copy: https://web.archive.org/web/2023/https://finance.yahoo.com/news/made-239-000-selling-budgeting-120626735.html. 2022-12-20. Read at source on 2026-10-07 in the syndicated copy; businessinsider.com refused us. The revenue figure is stated as checked by the outlet; the rest is his own account.
62. Gumroad blog, “From 0 to 27k in 8 months” (Easlo). https://gumroad.gumroad.com/p/from-0-to-27k-in-8-months. 2021-12-20. Read at source on 2026-10-07; the platform’s post, which does not say it checked the figures.
63. Gumroad blog, “Creator Spotlight: 10 years to $1 million in the most boring niche in the world”, by Brian Hahn. https://gumroad.gumroad.com/p/creator-spotlight-10-years-to-1-million-in-the-most-boring-niche-in-the-world-by-brian-hahn. 2025-03-20. Read at source on 2026-10-07; seller’s own account of a total the platform told him.
64. Gumroad blog, “Creator Spotlight: Philip Kiely’s success on Gumroad”. https://gumroad.gumroad.com/p/creator-spotlight-philip-kiely-s-success-on-gumroad. 2020-08-19. Read at source on 2026-10-07; seller’s figures relayed by the platform.
65. Gumroad blog, “How LifeMathMoney made $1m on Gumroad”. https://gumroad.gumroad.com/p/how-lifemathmoney-made-1m-on-gumroad. 2025-01-14. Read at source on 2026-10-07; seller’s own guest post.
66. Gumroad newsletter, “From ‘How-to’ Articles To $25k In One Month”, with weekly creator counts; five other issues from June to September 2021 read the same day. https://gumroad.gumroad.com/p/how-to-have-a-25k-product-launch. 2021-07-18. Read at source on 2026-10-07.
67. Stan, “State of the Creator Economy”. https://stan.store/blog/state-of-the-creator-economy/. 2026-02-14, updated 2026-07-21, covering 2025. Read at source on 2026-10-07 through the r.jina.ai reader; the platform’s own figures in a marketing report.
68. Stan blog, profile of Lindsay Hill. https://stan.store/blog/lindsay-hill/. 2025-10-29. Read at source on 2026-10-07; the platform’s marketing case study.
69. Contrary Research, page on Stan. https://research.contrary.com/company/stan. Undated; cites August 2026. Secondary: relays a co-founder’s podcast statements, which we did not open.
70. Gumroad blog, “What we shipped, what’s next, and our 2026 annual meeting”. https://gumroad.gumroad.com/p/what-we-shipped-what-s-next-and-our-2026-annual-meeting. 2026-03-02. Read at source on 2026-10-07.
71. TrustMRR, ShipFast page. https://trustmrr.com/startup/shipfast. Live page. Read at source on 2026-10-07, first through our fetch tool’s summary and then as the raw page through the r.jina.ai reader; the figures used are from the raw page.
72. Indie Hackers, “I analyzed 5,079 Stripe-verified startups”. https://www.indiehackers.com/post/i-analyzed-5-079-stripe-verified-startups-f0f6bd053f. 2026-03-12. Secondary: one poster’s analysis of TrustMRR listings, not reproduced.
73. Net Influencer, on Kit’s (ConvertKit’s) State of the Creator Economy 2024. https://www.netinfluencer.com/convertkit-state-of-the-creator-economy-2024/. 2024. Secondary: Kit’s own pages refused us.
74. Gumroad store pages of Daniel Vassallo (https://dvassallo.gumroad.com/), Brian Hahn (https://makethisyourlasttime.gumroad.com/) and LifeMathMoney (https://lifemathmoney.gumroad.com/), for ratings counts and listed prices. Live pages. Read at source on 2026-10-07.
75. Gumroad Help, “Gumroad Discover”. https://gumroad.com/help/article/79-gumroad-discover. Page timestamp 2025-09-10. Read at source on 2026-10-07 through the r.jina.ai reader.
76. Gumroad Help, “Getting paid”. https://gumroad.com/help/article/13-getting-paid. Undated. Read at source on 2026-10-07 through the r.jina.ai reader.
77. Indie Hackers, interview with Easlo. https://www.indiehackers.com/post/YQNh2atLObz6Jj4SYeBl. 2024-09-19. Ben Lang’s Notes, interview with Easlo: https://benlangsnotes.substack.com/p/tens-thousands-customers-selling-notion-templates-meet-easlo-22-yo. 2024-03-05. Read at source on 2026-10-07 through the r.jina.ai reader, the first above its sign-up wall only; his own statements.
78. Indie Hackers, “Steph Smith on making $130K w/ an ebook”. https://www.indiehackers.com/post/steph-smith-on-making-130k-w-an-ebook-creating-a-course-in-20-days-and-the-latest-trends-b1eed16421. 2021-10-21. Read at source on 2026-10-07 through the r.jina.ai reader; seller’s own report.
79. The Tilt, “Creator Economy Research Preview 2023”. https://www.thetilt.com/industry-news/creator-economy-research-preview-2023. 2023-05-01. Read at source on 2026-10-07 through the r.jina.ai reader; an unweighted survey by a company that sells creator education.
80. Sacra, “Gumroad at $21M”. https://sacra.com/research/gumroad-at-21m/. 2024-09-18. Sacra, research page on Gumroad: https://sacra.com/c/gumroad. Undated. Secondary: relays Gumroad’s reported numbers and Sacra’s own assessments.
81. Sacra, question-and-answer summary on Whop’s marketplace. https://sacra.com/chat/h/c44645cd-0fda-4300-888d-6d65174e49b7. Page timestamp 2026-10-07; describes April and May 2025. Secondary: a machine-written summary of Sacra’s research; Whop’s changelog was not opened.
82. Sacra, “Stan’s Shopify vs Whop’s Amazon”. https://sacra.com/research/stans-shopify-vs-whops-amazon/. 2026-04-08. Secondary: Sacra’s estimates.
83. MailerLite, Pricing. https://www.mailerlite.com/pricing. beehiiv, Pricing: https://www.beehiiv.com/pricing. Notion Help, “Selling on Marketplace”: https://www.notion.com/help/selling-on-marketplace. Live pages. Read at source on 2026-10-07; the Notion page through our fetch tool’s summary only.
84. Stripe, “What is an average chargeback rate?”. https://stripe.com/resources/more/what-is-an-average-chargeback-rate-a-guide-for-ecommerce-businesses. Undated. Read at source on 2026-10-07.
85. Indie Hackers, “Do you have a product on Gumroad?”. https://www.indiehackers.com/post/do-you-have-a-product-on-gumroad-c93ea4c0a5. 2020-09-21. Read at source on 2026-10-07 through the r.jina.ai reader; sellers’ own reports.
86. InsightRaider, “How much money can you make on Whop”. https://insightraider.com/en/answers/how-much-money-can-you-make-on-whop. Undated, 2026. Secondary: seen as a search-engine extract only; a tracker’s estimate.

Corrections

If you are quoted or named on this page and think something is wrong, want your reply shown beside it, or want to be removed, write to [email protected]. We aim to reply within 14 days, and always within one month. Factual errors are corrected with a dated note here. A sentence that is seriously disputed comes down while we check it.

Change log:

• 2026-10-07: Added “The upside” and “What it takes to compete” in place of “When it can work”. Corrected Gumroad’s marketplace entry rule from “a first sale, or about $10” to a $100 balance plus a review⁷⁵. Corrected “none is independently audited”: Easlo’s 2022 revenue was checked by Business Insider⁶¹. Added Gumroad’s 2021 weekly counts⁶⁶, Stan’s averages by audience size⁶⁷ and sources 61 to 86. After review: corrected the ShipFast figures to the raw page⁷¹, marked the price per copy of Kiely’s book as our sum⁶⁴, moved store prices to the store pages⁷⁴, and reworded the six-figure line, which had said “a large audience or many years”, to include audiences of about 4,000 to 12,000 built before the product¹⁴¹⁸⁷⁸.
• 2026-10-07: wording about named accounts and companies reviewed; a partly read press report and a page matched by name only were cut, and the FTC cases are worded as allegations.