TThe First 100 Members Problem

A community is not built by its thousandth member. It is built by its hundredth. By the time a chapter has welcomed a hundred people through the door, the unwritten rules are already set. What gets celebrated. What gets ignored. Whether you can ask a question without being patronised. Whether seniors show up for juniors. Whether the WhatsApp group is a place where work gets done, or a place where forwards go to die. The first 100 decide all of this, and by the time the 101st walks in, they are not choosing the culture. They are choosing whether to fit into one that has already been cast.

Most chapter founders in India misunderstand this. They run after numbers in the first quarter because numbers feel like proof. A list of 500 sign-ups on a Google Form looks like momentum. It is not. It is a queue. A community of 30 people who actually meet every fortnight in a Koramangala cafe, share work in progress, and answer each other's messages within the hour, is a stronger asset than a 5,000-member Telegram channel where the founder posts alone.

¶Why the first 100 carry disproportionate weight

The first 100 set three things that are very expensive to change later. Vocabulary, standards, and the default response to a stranger. Vocabulary is what words mean inside the room. If your first cohort uses "shipped" to mean "deployed to a real user," that meaning sticks. If it uses "shipped" to mean "showed a Figma to my manager," that sticks too. Standards are what kind of work is treated as serious. If the first 100 include three people doing rigorous machine learning research and seven people running paid newsletters, the room learns which of those gets the longer questions and which gets a polite nod.

The default response to a stranger matters most. In Indian professional settings, the default is often hierarchy first, help second. You ask where someone studied, you ask whom they work for, and only then do you decide how warmly to engage. A chapter that lets this default carry over will reproduce every gatekeeping pattern its members already deal with at work. A chapter whose first 100 deliberately invert the default, who ask "what are you working on" before "where are you from," builds something materially different.

¶What founders get wrong

The most common mistake is recruiting for fame. Founders invite the loudest people in their network, assuming visibility equals contribution. It does not. The loudest accounts on X are often the least useful in a room of forty people on a Saturday morning, because their job is to perform, not to participate. Coimbatore and Pune have shown this clearly in the last two years. The chapters that took off were seeded with quiet operators, a CA in their second year of article-ship, a registrar at a medical college, a product manager from a Tier-2 SaaS firm, not with the local podcast circuit.

The second mistake is geographic laziness. A founder in Bengaluru recruits her last three roommates and four ex-colleagues and calls it a chapter. The first 100 must come from at least a dozen overlapping but distinct networks, or the chapter will feel like a private club to anyone walking in from outside that lattice. NLS alumni, AIIMS residents, IIM Bangalore PGP students, government employees on deputation, a couple of school teachers, founders from Indore who moved for funding, an artist who teaches at Srishti. Width at the start is cheap. Width retrofitted at member 800 is impossible.

¶How to actually pick them

Treat the first 100 as a deliberate composition, not an inbox. Maintain a written charter, even one page, that names what the chapter is for and what it is not. Send personal invitations. Real ones, with a sentence about why this specific person belongs in this specific room. Aim for a ratio that holds: roughly a third in their first five working years, a third in years six to fifteen, a third senior. Roughly a third women. Roughly a quarter from outside the host city. At least ten members who are not in tech, because a chapter that is only engineers will produce only engineering conversations, and the human layer for AI is broader than that.

Set a participation bar before you set a membership bar. Show up to two of the first four meets. Share one piece of in-progress work in the first month. Reply to at least three other members' shares. The bar should be low enough to be human and high enough that being a member means something. People rise to clear expectations. They drift past vague ones.

¶What to do this month

If you are launching or seeding a chapter, write down the names of the first 100 you want, before you send any invitation. Do it in a spreadsheet, by hand. Mark each one with the network they come from, what they are working on, and one sentence on what they bring to the room. If your list has more than 40 people from a single employer or institute, rebuild it. If it has fewer than 15 people from outside your home city, rebuild it. Send the first ten invitations this week, by voice note or in person, not by broadcast. Then host one small event, no larger than 25 people, and watch what gets said in the corners. That is your culture. Tune it now, while it is still soft.