Most posts about launching a community platform are written by people who already have an audience. I don't. So this one is different.

This isn't a retrospective. It's a plan, written down so I'm accountable to it.

Here's my actual situation as ProductLog goes live:

0 users besides me
A small Turkish following on X, mostly not makers
Effectively zero reach in English
2-3 hours a day for this project, split across building, marketing, and bug-fixing
A day job, a family, and other projects in flight
No budget worth mentioning
No co-founder

Community platforms are notoriously hard to start. They're worthless when empty, but they only stop being empty if people come, and people only come if the platform isn't empty. The classic chicken-and-egg.

Most advice on how to escape this loop assumes you have leverage I don't have. Big audience. Big budget. Big personality. Big network of investor friends who'll tweet about your launch.

This post is about what I'm planning to do instead, in two phases, and why I think it's the only honest approach when you don't have any of that.

The lie of organic growth

The first thing I had to accept: "build it and they will come" has never worked for community platforms.

It didn't work for Reddit. It didn't work for Product Hunt. It didn't work for Indie Hackers. The founders of each of those platforms spent months manually seeding content before anything took off.

Reddit's founders famously created fake accounts and posted to their own site to make it look populated. Product Hunt started as Ryan Hoover's curated email list. Indie Hackers became a forum only after Courtland Allen had personally interviewed dozens of makers.

None of them got lucky. They all did the unglamorous work first.

If you're launching a community platform and you're hoping users will just show up, you're not launching a platform. You're launching a wish.

I had this wish. I had to put it down.

Phase one: be my own user

The first phase isn't about getting other makers in. It's about being a real user of my own product.

I have other projects I'm shipping in parallel with ProductLog. For the next stretch, every one of those projects gets its product page, its updates, its roadmap, and its feedback board on ProductLog. When I talk about a project on X, I link to the ProductLog update for it. When I ask for feedback, I send people to the ProductLog board. When I launch something, I launch it on ProductLog first.

This serves three purposes at the same time, and that's the only reason it fits into 2-3 hours a day.

First, it surfaces every bug, every rough edge, every empty state, every confusing flow. I'm the one running into them, so I'm the one motivated to fix them. The platform gets quietly polished while I use it.

Second, it puts real content on ProductLog. Visitors land on a platform that has actual products with actual updates and actual roadmaps — not a maker manifesto and three placeholder pages. The page looks alive because it is.

Third, it sends a small but steady trickle of curious people through ProductLog. A few of my X followers click the update link, see what the platform is, and stick around. A few of the people I asked for feedback start looking at the trending page. Nothing viral, but not nothing.

Phase one ends when ProductLog is something I'd be willing to send a stranger to without apologizing. When the empty states are real, the onboarding is guided, the bugs that matter are fixed, and the page doesn't feel like a construction site.

Phase two: invite the makers

Only then do I start the outreach.

The plan is to personally invite the first thirty to fifty makers. One by one. Not a mass email. Not a Twitter announcement. Specific people, with specific messages, after I've read their work.

Most makers will ignore the message. Some will say yes. The ones who say yes shape what the platform becomes.

What I'm not doing

A list, because half the advice for early-stage founders is about things you shouldn't do:

Not running ads. ProductLog has no audience to amplify yet — paid acquisition would just spend money to send strangers to a half-empty site.
Not chasing virality. Viral mechanics need a base of users to spread through. No base, no spread.
Not doing a big launch. Product Hunt and Hacker News reward platforms with existing momentum. I don't have any. A launch now would burn the one shot at attention I get with these channels.

The temptation is to do all of these because they feel like work. They're not. At this stage, they're avoidance.

What I'm uncertain about

A few things I haven't figured out and don't want to pretend I have:

When phase one ends. I have no objective marker for "ready enough to invite strangers." I'll know it when I see it, but I might be wrong. The risk is staying in phase one too long, polishing forever, never sending the messages. I have to fight that.

Whether to start phase two with Turkish makers first. I have a small Turkish following. Reaching ten Turkish makers is easier than reaching ten global indie hackers from a #coldstart. But the long-term audience for ProductLog is global. There's a real argument for starting where I have leverage and a real argument for not creating the wrong first impression. I haven't decided.

Whether my own projects will produce enough content to make phase one work. Phase one assumes I'm shipping things in parallel with ProductLog. If those projects stall, ProductLog stalls with them. I'm betting on my own consistency, which is not always a safe bet.

What happens after phase two. If I have a hundred active makers, that's a small but real community. What's the next layer of growth? I don't know yet. I'm going to find out by getting there first.

The only honest answer

When you have no audience and a few hours a day, the honest answer is: use the platform yourself until it's good enough, then send specific messages to specific people, build their thing for them so saying yes is one click, and do this until you have enough activity that the next phase is even an option.

I'm going to do that. I'll write the follow-up post in 90 days, either explaining what worked, or explaining what didn't and what I learned.

#coldstart #buildinpublic #productlog

Cold-starting a community platform with zero audience and a few hours a day