Quick disclaimer before we get into it. The creator in this story is a composite, built from patterns that repeat across a lot of memberships. The figures are illustrative rather than reported, and she is not a real person you could go and verify. What she ran into, though, is one of the most common growth problems there is, and the way through it is worth borrowing.
Here is the scene. Around three hundred paying members. A subject where the gap between somebody in their first month and somebody in their eighth year is enormous. And a founder who kept getting two completely different pieces of feedback in the same week: this is too advanced for me, and this has gotten too basic.
For about a year she treated that as a communications problem. It was not. She had two communities living in one room, and they were slowly making each other miserable.
The symptom she noticed first
Her experienced members had stopped posting questions. They still showed up, still renewed, still answered other people generously. But they had quietly stopped bringing their own hard problems into the room.
When she finally asked one of them about it, the answer was almost embarrassing in how obvious it was. He said that when he posted something genuinely difficult, half the replies were from people who did not understand the question, and answering their follow ups took more time than solving his own problem. So he stopped, and took his real questions to a group chat with three friends instead.
That is what a community looks like right before its most valuable people drift away. Not anger. Not complaints. Just a gentle relocation of the interesting conversations somewhere else.
Meanwhile her newest members were having the opposite experience. They would open the community, see an exchange between two very advanced people, and conclude that they had joined something above their level. Several of them told her later that they lurked for two months because they were afraid of asking a stupid question in front of professionals.
So both groups were being politely driven out by the presence of the other. Nobody was behaving badly. The room was just doing something no single room can do.
The wrong fixes, in order
She tried tags first. Every post got a level marker so people could filter. Adoption was about forty percent for two weeks and then effectively zero, because nobody wants to classify their own confusion before they can ask about it.
She tried a beginners channel. Sensible, except that it read as the kiddie table. New members posted there once, saw that the founder and the interesting people mostly hung out in the main room, and either went quiet or migrated up and felt out of place again.
She tried a rule asking experienced members to be patient with basic questions. This one is worth dwelling on, because it looks like leadership and it made things worse. It framed beginners as a burden to be tolerated, and it added an unpaid obligation to the people who were already giving the most. Two of her strongest contributors cooled off noticeably in the following weeks.
The lesson she took from that stretch is one I would tattoo on every founder: when the same problem survives three different fixes, the problem is structural. Stop tuning and start deciding.
The decision, and how she framed it
She split the membership into two communities with separate rooms, separate calls, separate pricing, and separate onboarding. One built for people learning the craft. One built for people who make their living from it.
The framing she landed on took her three weeks to get right, and it is the single most important part of this whole story. She did not present it as basic and advanced. She presented it as two different jobs. One community is for building the skill. The other is for the business and craft problems that only show up once you are working professionally.
That distinction is not cosmetic. Levels imply a hierarchy and make everyone below the top feel provisional. Different jobs implies a choice, and a member choosing which room fits their current life feels like an adult making a decision rather than a student being sorted.
She also refused to gatekeep. No application, no proof of experience, no approval. You pick the one that fits, you can switch at any time, and switching costs nothing but a message. About a dozen people picked the professional room when she privately thought they should not have, and most of them either grew into it or moved themselves after a month. Nobody needed policing.
How she actually ran it
She told people six weeks out, which was longer than felt comfortable and turned out to be right. The first announcement explained the problem honestly, including the part about experienced members going quiet, because she wanted people to understand that this was a response to something real rather than a pricing exercise.
Then she did the unscalable thing again. She wrote to every member individually, a short personalized note with a recommendation of which room she thought suited them, phrased as a suggestion with an explicit invitation to disagree. Three hundred notes over eight days. She has since said it was the highest return work she has ever done, because almost nobody felt sorted by an algorithm.
Pricing was the part she nearly fluffed. The professional community cost more, which is normal, and her first instinct was to raise the price for existing members immediately. She did not. Everybody kept their current rate for a full year, and new members paid the new rate. That decision cost her real revenue in year one and bought her something better, which was a complete absence of the story where longtime supporters got a bill for being loyal.
She kept the old archive available to both groups, permanently. People had paid for that material and she was not going to make it a bargaining chip.
And she was careful with the mechanics. Nobody's payment details were moved around without consent, nobody was auto enrolled into a more expensive plan, and every change to what somebody paid was confirmed by that person in writing first. That sounds like paperwork. It is actually the difference between a reorganization and a scandal.
What broke in the first month
Two things, both predictable in hindsight.
The professional room felt empty for about three weeks. Half the size of what everyone was used to, and the members who had asked for it most loudly were suddenly the ones with nothing to say. She fixed it the way you fix any quiet room: she went and asked six people, privately, to bring one real problem each. Within a month it had the density it needed.
The bigger problem was her own time. Two communities meant two sets of calls, two calendars, two moderation loads, and roughly double the emotional surface area. She burned through about six weeks of energy before admitting it was not sustainable and cutting the live calls in the learning community from weekly to fortnightly, replacing the gap with a written thread. Nobody complained. She had been running on the assumption that reducing anything would read as decline, which is a very common founder delusion.
The part she did not see coming
Referrals went up, noticeably, in both communities. Not because she incentivized anything. Because people could finally describe what they were part of in one clear sentence.
Before the split, a member trying to recommend her had to say something vague about a community for people interested in the subject, at all levels, sort of. After, they could say this is where you learn it, or this is where working professionals talk shop. Clarity travels. Vagueness does not, no matter how good the thing is.
The other surprise was that the learning community turned out to be more fun to run. With the professionals in their own room, the beginners started answering each other. A member six months in became the person who explained the basics beautifully, precisely because she remembered being confused. That role had been unavailable to her while there were experts in the room.
Would she do it again, and would you
She would, and she would do it a year earlier.
But this is not a universal prescription, and she is careful about that when people ask. A split makes sense when you can point at two distinct jobs your members are hiring you for, when both groups are big enough to feel populated, and when you have the capacity to run two rooms without collapsing. Below roughly a hundred and fifty members, splitting mostly produces two rooms that both feel dead, and the honest answer at that size is to pick one audience and let the other go elsewhere.
If you cannot describe the two groups in a sentence each without using the words beginner and advanced, you probably are not ready. That constraint forces you to find the real difference rather than the obvious one.
Do this this week
Go and look at whether your most experienced members are still asking their own questions, or only answering other people's. That single check tells you more about the health of your community than any dashboard.
Then write down the two jobs you think your members are hiring you for, in one sentence each, with no reference to skill level. If you can do it easily and the two sentences look genuinely different, you have something worth thinking about. If you struggle, your community is probably still one thing, and you should go make it better rather than bigger.
If you do decide to move, give people six weeks, write to them one at a time, protect the price of the people who got you here, and keep the archive open. Then expect the new room to feel quiet for a month and go seed it with real problems rather than panicking.
Growing into two things is a good problem. It means enough people trusted you that they arrived at different stages of their lives. That is worth being a little bit thrilled about, even while you do the unglamorous work of sorting out the rooms.
