Industry
Part of Creator memberships: the parts worth your attention
Creator membership tiers that stop being worth running: common patterns and fixes
Creator membership tiers that stop being worth running, from Patreon and Substack to Discord rooms, with common patterns and fixes that remove something.
The examples below are public. Each is a real creator, a real membership, and a real change that was announced or reported. What each one broke on is the part worth copying into your own plan.
What to take away
- Most failed tiers broke on hours, not on demandthe perk that sold the tier was the perk that consumed the creator.
- Several fixes were subtractions, including caps, closed seasons and pulling a product out of the tier.
- A tier that promises a reply from one person has a hard ceiling, and the ceiling arrives before the revenue does.
- A member list you cannot export is a list you do not own.
- Check your own numbers before you change anythinghours per member per month, and the month members leave.
What "failed" means here
A tier fails when it stops being worth running. Sometimes it is cancelled. Sometimes it is capped, repriced or folded into another tier. Sometimes it quietly stops being mentioned, which is the slowest version.
The cases below cover all four endings. None of them is a warning about memberships in general. Each is a warning about one specific promise.
Example: the archive tier that ran out of archive
Tom Scott's channel memberships and his later subscription work sat alongside a very large free back catalogue. The paid draw was never the old material, because the old material was already free. It was the new, regular thing.
That is the pattern to copy. An archive is a finite object, and a finite object is a product. It sells once. A membership sells the next one.
The fix shape: sell the archive as a one-time product and keep the membership for what arrives monthly. Fewer people join, and the ones who do have a reason to still be there in month four.
Example: the feedback tier with no cap
This is the most common failure in the category, and the clearest public version is the coaching and critique tier sold by writing and craft creators, where the top level promises a personal reply to each member's work. N.K. Jemisin's Patreon critique tier is one named example of that promise.
Is your feedback tier broken?
- Count hours per reply
- Multiply by expected top-tier members
- Compare to hours left after main work
- If step 3 < step 2, tier is broken
It sells fastest of any tier. A reply from a named person is worth money. Then the queue becomes the job.
The arithmetic is simple and worth doing before you launch:
- Count the hours one reply takes you, including reading, thinking and writing.
- Multiply by the number of members you expect in the top tier.
- Compare that total to the hours you have left after making the thing members joined for.
If step three is smaller than step two, the tier is already broken. The fix is a published cap and a waitlist, or a monthly group session with a queue instead of an open inbox. Both lose members. Both keep the creator working.
Example: the forever membership
Substack and Patreon both host large numbers of open-ended monthly memberships with no stated schedule. Substack's Founding Member tier and Patreon's Annual Membership option are two named versions. The ones that stall tend to stall the same way: a prospective member cannot tell what they get, and an existing member cannot tell whether they missed anything.
Seasons fix this. A season has a theme, a start, an end and a stated set of deliveries. Renewal becomes a decision rather than a default. Churn between seasons disappears because there is nothing between seasons to churn from.
The trade is real: a season gives you a natural moment to talk to your whole audience again, and it also gives members a natural moment to leave.
Example: the room that went quiet
Discord-based tiers on Patreon and Ko-fi follow a predictable curve. Patreon's Discord role sync and Ko-fi's Discord integration are the named tools behind those tiers. Week one is busy. Month two is a handful of names. Month three is silence, then cancellations.
The cause is almost never the members. It is the absence of a rhythm. A room with no scheduled prompt waits for someone to start something, and nobody wants to be first.
What tends to work: written rules, a second moderator, and one weekly prompt from the creator. Removing a member and refunding them is a decision to make in advance, alongside the cancellation terms the Federal Trade Commission's negative option rule page sets expectations for.
Example: the support tier that got "improved"
Low tiers with no perks, sold as support, work. CGP Grey's Patreon is a public example. Members pay because they want the work to exist. It is a stable motive and it asks nothing of the creator.
It breaks when the creator gets nervous about the low price and starts adding perks to justify it. Each added perk creates work, and the work was never what those members bought. The fix is to leave it alone and send one short monthly note about what the money made possible.
Example: the tier that lived inside one platform
Patreon's 2017 fee change and its 2018 tier restructuring both pushed creators to explain new terms to their members. Some members followed. Some did not. Kickstarter's Drip, a membership platform that shut down in 2018, is another named case.
The lesson is not that platforms change. It is that the member list and the billing records should exist somewhere you control. Export the list monthly. Keep the payment records in your own files, which is what the Internal Revenue Service's recordkeeping guidance asks of anyone self-employed.
The same rule applies to every platform-dependent income line: what you cannot take with you was never entirely yours.
The pattern
Every fix above removed something. A perk, an unlimited tier, an archive, a sense of forever. The membership guide covers why the tier that costs you per member decides whether the whole thing survives.
The second pattern is competition for hours. In each case the membership was competing with the creator's other work, and when the membership won, the audience feeding it started to shrink. Which shape of career a membership belongs in is worth deciding before it exists.
Common questions
Which of these fails most often?
The uncapped feedback tier. It sells best and breaks most reliably, for the same reason each time: people will pay for a reply, and a reply from one person does not scale.
Can any of these be fixed without losing members?
Rarely. Each fix above lost some members and improved retention among the rest. A smaller, steadier membership is the one still running in year three.
How do I tell which case I am in?
Look at hours and at exits. High hours per member points at the feedback tier. Members leaving in month two points at the archive tier. Cancellations clustered in quiet months points at the missing season.
Do I need a CPA for any of this?
Membership income, platform fees and refunds all land on your return, and treatment varies by jurisdiction. Ask a CPA or licensed tax professional where you file, and confirm platform terms with the platform's own published documentation.
