Rules
Telling whether a niche social network is still alive or just dormant
Niche social networks: why small ones are worth building on, how they actually die, and how to check whether one is alive before investing a year in it.
Small networks do not obey the rules that large ones do. A large network is valuable because everyone is there. A small one has to be valuable for some other reason, and the reason is almost always the same: the people in it share a thing specific enough that a general audience would dilute it.
That makes researching them a different job. The questions that work on a big platform, starting with how many people use it, tell you least about a small one. Even on the large services, size is the weakest column available; here it is close to useless.
What to take away
- Check activity, exports, and who runs it before you commit.
- Keep the same short sheet for every specialist network you evaluate, so you are comparing like with like.
- Building an audience on a specialist network is a bet that depth beats reach for what you do.
Why anyone builds on a small network
The honest advantages are narrow and real.
Ravelry for fiber crafts, Letterboxd for film, and Untappd for beer are examples. Each is built around a specific interest, not a general audience.
Context is free. In a specialist network you do not have to explain the basics before you can say anything. On a general platform every post has to work for someone who arrived by accident, which flattens everything toward the introductory.
The audience is pre-qualified. A thousand people who all do the specific thing are worth more, for most purposes, than a hundred thousand who might. This is why professional networks and hobbyist ones keep existing despite every large platform trying to absorb them.
Norms can hold. Small enough groups can enforce standards through social pressure rather than policy. That is a genuine capability large platforms cannot buy.
You are visible. Being one of two hundred active posters is a position that no amount of effort buys you on a service with millions.
The corresponding risks are just as narrow: the whole thing can end quickly, and usually for reasons that have nothing to do with you.
How small networks actually die
Almost never from competition. The common causes:
How small networks die
- Founder gets tired
- Bill comes due
- Core drifts away
- Single moderation crisis
- Infrastructure dependency breaks
- The founder gets tired.A large share of specialist networks are run by one or two people, often unpaid, alongside jobs. When they stop, there is frequently no succession plan and no second administrator with access.
- The bill comes due.Hosting, moderation, and legal exposure scale faster than a hobby project's income. The moment a small network needs real money is the moment it either changes character or closes.
- The core drifts away.Twenty people were carrying it. Five leave, the rest notice it is quieter, and the decline compounds over a few months.
- A single moderation crisis.A small network has no institutional capacity to absorb a serious dispute. One unresolvable fight can end it.
- An infrastructure dependency breaks.A payment provider, a host, an authentication service, an app store. Small operators have no bargaining power when one of these changes terms.
Cohost, a small text-focused network, announced in 2024 that it would close after running out of money. That is the bill-comes-due cause in practice.
Notice that none of these are visible in a member count, and most of them are visible if you know what to look at.
Researching whether a small network is alive
Alive or dormant?
- Read the middle, not the front page
- Count distinct names in last 50 posts
- Find newcomers and see if answered
- Identify who is actually running it
- Check failure history via archives
- Test exits before entrance
Read the middle, not the front page. Find a section that is not the busiest one and look at whether anything has happened in it recently. Active networks are broadly active; dying ones concentrate into a single room.
Look at the participation spread. Count the distinct names in the last fifty posts. A high count means a community; a low one means a few people talking and an audience watching, which is much more fragile than it looks.
Find the newcomers. Search for people who joined recently and see whether anyone answered them. A network that has stopped absorbing new members is on a clock, however busy it feels today.
Identify who is actually running it. One person, a small team, or a company? Is there any public statement about funding? Is anyone paid? You are not looking for reassurance (most of these are one person, and that is fine), you are looking for whether you know the risk you are taking.
Check the failure history. Search for past outages, data loss, or governance disputes. Use web archiving services such as the Wayback Machine, archive.today, and Perma.cc to read the announcement pages as they were written rather than as they were later edited. How the operators handled the last bad week is the best available predictor of the next one.
Test the exits before the entrance. Data portability is the property to check first, because it is the only one that stops mattering after the network closes. Can you export your own posts? Your contacts? Is there any way to reach people you meet there if the network goes away? Run the export on day one rather than the day you need it. Use the platform's own tool, such as Mastodon's CSV export, Discourse's user archive, or Reddit's data request.
Watch for the second home. Small networks nearly always have an unofficial overflow somewhere: a chat, a list, a group elsewhere, usually running on one of the messaging services everyone already had. Being in it is the difference between surviving the network's closure and disappearing with it.
What to record
Keep the same short sheet for every specialist network you evaluate, so you are comparing like with like.
Evaluation sheet fields
Field
- Defining specificity
- Shared thing general platforms dilute
- Operator
- Individual, volunteer group, company; funded how
- Bus factor
- How many could keep it running
- Participation spread
- Distinct posters in recent sample
- Newcomer response
- Whether recent arrivals got replies
- Archive
- Durable, searchable, exportable old material
- Contact portability
- Can reach members if network stops
- Overflow
- Where members go when down
What to record
- Defining specificity
- Operator
- Bus factor
- Participation spread
- Newcomer response
- Archive
- Contact portability
- Overflow
| Field | What to write |
|---|---|
| Activity dates | The newest post in at least three sections, not just the busiest one. |
| Distinct posters | The number of different names in the last fifty posts. |
| Export availability | Whether you can download your posts and contacts, and which tool does it. |
| Operator | Who runs it, whether anyone is paid, and any public funding note. |
| Second home | Where members gather if the network closes, such as a chat or mailing list. |
| Failure history | Past outages, data loss, or governance disputes, with dates. |
| Newcomer response | Whether recent joiners got replies within a few days. |
| Infrastructure dependencies | The host, payment provider, authentication service, or app store it relies on. |
The trade you are making
Building an audience on a specialist network is a bet that depth beats reach for what you do. Sometimes it plainly does: if your work has forty possible buyers in the world, being known to thirty of them is the whole job, and a general platform cannot deliver that at any price.
The bet goes wrong in one specific way: people mistake a network's intimacy for durability. It feels permanent because you know everyone. Feeling known is not the same as being safe.
Behave as though it could close next year. Keep the export current and keep a route to the people you care about that does not depend on the operator, and the bet becomes cheap. Skip those and a good decision turns into a total loss for reasons that were entirely predictable.
Some of this overlaps with running a group of your own; the shape questions are set out in the notes on community platforms.
Common questions
Is a small network worth it if I already have an audience elsewhere?
Often yes, but for a different reason than you expect. The value is usually input rather than output: being where the specific conversation happens, rather than broadcasting to it. Judge it by what you learn, not by how many followers accrue.
How do I tell a quiet network from a dead one?
By response, not volume. Post a real question in a normal part of the network and see whether anyone answers within a few days. A quiet network answers. A dead one does not, and the difference is invisible from the traffic alone.
Should I cross-post everything from a large platform?
No, and it is the fastest way to be resented. Specialist networks notice syndicated content immediately, because it is written for someone who does not know the basics. If it is worth posting there, it is worth writing for the people there.
What if the network has no export at all?
Then treat everything you put there as unrecoverable, and adjust what you put there accordingly. That is not automatically a reason to avoid it, plenty of worthwhile conversations happen in places you cannot back up, but it should be a decision rather than a surprise.
In this guide
- Niche social networks facts: how to verify four claimsNiche social networks facts: a method for checking claims about independence, privacy, community control and open-source code against available evidence.
- Measuring niche social networks dataNiche social networks data: use counts and consistent windows, account for snowball sampling, set thresholds, and record each observation.
