The empty-room problem
Community platforms, marketplaces, and niche networks share a failure pattern. The team specifies profiles, listings, messaging, reviews, groups, events, notifications, and a feed. It takes a year. It launches. And it is empty, because every one of those features only becomes valuable when other people are already there., marketplaces, and niche networks share a failure pattern. The team specifies profiles, listings, messaging, reviews, groups, events, notifications, and a feed. It takes a year. It launches. And it is empty, because every one of those features only becomes valuable when other people are already there.
Meanwhile the audience the platform was built for is still in a group chat, a forum, and a marketplace they complain about weekly. They did not leave because your feature list was shorter. They stayed because your room was empty.Meanwhile the audience the platform was built for is still in a group chat, a forum, and a marketplace they complain about weekly. They did not leave because your feature list was shorter. They stayed because your room was empty.
So the sequencing problem is the real problem: what can you ship that is useful to a single person on day one, before there is a community at all?So the sequencing problem is the real problem: what can you ship that is useful to a single person on day one, before there is a community at all?
Find the single-player utility
Most successful niche platforms started with something a person would use even if nobody else showed up. A tool, a directory, a record of their own stuff, a calculator, a place to organize what they already track badly. started with something a person would use even if nobody else showed up. A tool, a directory, a record of their own stuff, a calculator, a place to organize what they already track badly.
In an automotive community, that might be a person's own vehicle history and build record. For a professional niche, it might be a portfolio or a compliance tracker. The test is simple: would one motivated person still use this on a Tuesday if they were the only account?In an automotive community, that might be a person's own vehicle history and build record. For a professional niche, it might be a portfolio or a compliance tracker. The test is simple: would one motivated person still use this on a Tuesday if they were the only account?
Single-player utility solves acquisition and retention at the same time. It gives people a reason to arrive before the network exists, and a reason to return during the months when the network is still thin.Single-player utility solves acquisition and retention at the same time. It gives people a reason to arrive before the network exists, and a reason to return during the months when the network is still thin.
Choose a niche narrow enough to fill
A platform for everyone in a category is empty everywhere. A platform for one region, one segment, or one vehicle type can feel busy with a fraction of the users, because density is what makes a community feel alive, not headcount.A platform for everyone in a category is empty everywhere. A platform for one region, one segment, or one vehicle type can feel busy with a fraction of the users, because density is what makes a community feel alive, not headcount.
Pick the narrowest slice where you can plausibly reach most of the participants, and be willing to sound too specific in your positioning. You can widen later from a position of density; you cannot concentrate a thin general audience after the fact.Pick the narrowest slice where you can plausibly reach most of the participants, and be willing to sound too specific in your positioning. You can widen later from a position of density; you cannot concentrate a thin general audience after the fact.
Density also improves the content problem. Ten active people in one niche produce a room worth visiting. Ten active people spread across a whole industry produce a ghost town with a good logo.Density also improves the content problem. Ten active people in one niche produce a room worth visiting. Ten active people spread across a whole industry produce a ghost town with a good logo.
Sequence the multiplayer features
Once single-player utility works, add features in the order that requires the fewest people to be valuable. Public profiles and listings work with a handful of users. Search and discovery become useful with dozens. Feeds, reputation systems, and recommendations need real volume, and they look sad without it. become useful with dozens. Feeds, reputation systems, and recommendations need real volume, and they look sad without it.
Messaging deserves particular care. It is easy to build and hard to run: it introduces moderation, abuse handling, notification design, and support obligations. Add it when there is something worth messaging about.Messaging deserves particular care. It is easy to build and hard to run: it introduces moderation, abuse handling, notification design, and support obligations. Add it when there is something worth messaging about.
Every multiplayer feature also has a moderation cost. Ask, before building each one, who handles the bad version of it, because the answer determines whether you can afford the feature at all.Every multiplayer feature also has a moderation cost. Ask, before building each one, who handles the bad version of it, because the answer determines whether you can afford the feature at all.
Seed the room deliberately
Nobody joins an empty platform on principle. Early density is manufactured, and that is legitimate as long as it is honest. Import public data you have rights to. Recruit a founding group individually. Run the first events yourself. Create real content rather than fake accounts.Nobody joins an empty platform on principle. Early density is manufactured, and that is legitimate as long as it is honest. Import public data you have rights to. Recruit a founding group individually. Run the first events yourself. Create real content rather than fake accounts.
Do not simulate activity you do not have. Fake profiles and invented engagement destroy trust exactly once, permanently, and in a niche community word travels faster than in a general one.Do not simulate activity you do not have. Fake profiles and invented engagement destroy trust exactly once, permanently, and in a niche community word travels faster than in a general one.
Concentrate early activity rather than spreading it. One busy category beats twelve categories with one post each, so keep the structure narrow until there is enough activity to justify splitting it.Concentrate early activity rather than spreading it. One busy category beats twelve categories with one post each, so keep the structure narrow until there is enough activity to justify splitting it.
Instrument what tells you to keep going
Signups are the least informative number available. Watch instead whether people come back without being prompted, whether they complete the core action a second time, and whether any interaction happens that you did not personally cause.Signups are the least informative number available. Watch instead whether people come back without being prompted, whether they complete the core action a second time, and whether any interaction happens that you did not personally cause.
Set a decision point in advance: at this date, if these behaviors are not present, we change the wedge rather than adding features. Platforms rarely die from picking the wrong first feature; they die from spending three years adding features to a wedge that never worked.Set a decision point in advance: at this date, if these behaviors are not present, we change the wedge rather than adding features. Platforms rarely die from picking the wrong first feature; they die from spending three years adding features to a wedge that never worked.
The discipline is the same one we apply to our own ventures in development, including SWIIRV, which is an automotive social, community, marketplace, and marketing platform being built by this sequencing rather than by shipping a full feature list at once., which is an automotive social, community, marketplace, and marketing platform being built by this sequencing rather than by shipping a full feature list at once.
Questions we get
- How small should a first version be?
- Small enough that one person gets value with nobody else present, and specific enough that the intended audience recognizes it as built for them. If you cannot describe it in a sentence, it is still too large.
- When should we add messaging and feeds?
- After there is something worth talking about and someone responsible for moderation. Both features carry ongoing operational cost, and both look empty at low volume, which makes an early platform feel deader than it is.
- marketplace
- community
- launch
- product strategy