Build in public

Build in public means the work is visible while it is unfinished. You show the product, the mistakes, and, often, the money. The phrase is a habit. For a few years the scene treated it as a growth system. The system got copied. What is left is a choice about what to show, not a promise that showing it will sell.

The timeline puts the era at roughly 2020–2023, overlapping the years when a following became part of the product. Twitter, now X, was the stage. A screenshot of a payments dashboard became a progress report. Followers were treated as the first audience. The numbers made a story concrete. They were also one person’s books, shown because showing them was part of the marketing.

What the habit was for

A private build has a small problem: nobody knows it exists, and the maker has no pressure to finish. A public log answers both, a little. Strangers can try the thing early. The maker has to ship something they are willing to attach their name to. That is the useful core. It does not require a revenue chart.

The chart arrived because it sorted the feed. A milestone is easy to understand. “$100k ARR” is a plot. In September 2021 Damon Chen posted that Testimonial.to had reached $100,000 in annual recurring revenue. Indie Hackers shared that post on 19 September. The share’s headline mentions three projects. Read the milestone as his public claim. It traveled because the number was legible, not because the thread was an audit.

Tony Dinh’s public record is the same habit across two products. Black Magic, a Twitter tool, reached a sale: in a May 2023 newsletter he wrote that he sold it for $128,000 after API pricing made it unprofitable, and that it was around $14,000 a month at the sale. TypingMind is the later number. In October 2025 he wrote that it was making about $130,000–$160,000 a month. Both figures are his. The habit is the point of this page. The businesses are not interchangeable.

Marc Lou made the habit a streak. Product Hunt’s 2023 Maker of the Year note says he launched 14 products that year. ShipFast went up on 31 August 2023. The launches were the content. The longer story is that most of 36 startups did not become the thing people remember. A public streak hides the misses unless the founder puts them back in. He sometimes does. The feed usually does not.

Why people say it died

buildinpublic.so argues the performance version died: daily MRR, gratitude posts, the milestone as the content. A plain build log is a different thing, and they do not bury that. LaunchKit is harsher. For the average maker the window closed around 2024, because the audience is other builders. They will like the post and not buy. Screenshots raised the bar until only huge numbers moved anyone. Transparency now feeds weekend clones.

Those are arguments about a channel, not a headcount. People still post. Indie Hackers still runs a build board. The posts are less likely to be a customer factory for a product whose buyer is outside that crowd. You do not need this audience to have a product. If this audience is your only plan, the “dead” posts are describing your plan.

The clone risk got sharper after vibe coding. A thread that explains the feature list is a spec. An agent can draft the spec into a repo. Levels’ 2026 warning about execution as a moat applies directly: the more complete the public tutorial, the cheaper the copy. Vibe coding did not invent the leak. It shortened it.

The credibility risk is older. Some screenshots were fake. Around 28 October 2025, in Marc Lou’s telling, Levels called that out, and Lou started TrustMRR the next morning: read-only Stripe, revenue the founder cannot edit by hand. MRR is still the number people want to see. A self-reported MRR and a Stripe-connected MRR are not the same evidence. Build in public that depends on unverifiable numbers is a genre. It is a weak proof.

What is left that is still true

Early users can still come from a small, specific post. A person who has the problem will try a rough product if they can see that a human is answering. That does not require a daily cadence or a follower goal. It requires the right people to see it once. One honest thread in a niche community can beat a month of milestone posts to other founders.

Accountability still works on the maker, which is easy to mock and hard to replace. A public “this week I shipped” is a commitment device. It does not need an audience larger than a few peers. Some communities exist mostly for that: a log, not a launch. WIP’s old idea — post what you shipped — is the non-performance version. Whether a given Discord is healthy in a given month changes. The shape is the point.

The record is still useful to other builders, which is why this site can cite it. A founder who writes the month down, including the bad one, leaves a primary source. Tony Dinh’s newsletter, Lou’s letters, Levels’ posts, Hallam’s pre-sale screenshot: the scene’s memory is build-in-public artifacts. That is a public good and a marketing channel at the same time. Readers should keep both in view.

What is no longer safe to assume is that the posting is the growth. The 2026 state is that attention is scarce enough to auction. A free post competes with people paying for placement. Organic build-in-public still happens. It is not the default unfair advantage it was when fewer people did it.

How to do it without turning it into a costume

Say what shipped. Say what failed if you can do it without handing a competitor the only hard part. Separate a founder-reported number from a verified one, and do not invent precision you do not have. If the buyers are not the people who follow builder accounts, spend the distribution effort where the buyers are, and let the public log be a side effect.

If the buyers are those people, the log is the market. Lou’s maker tools and Hallam’s X products live there. Pretending otherwise is a different kind of costume. The Marc Lou and Rob Hallam pages are those cases. Copy the match between audience and customer, not the posting frequency.

The phrase survived. The gold rush inside the phrase did not. A build log is a tool. A performance of revenue is a tactic that filled up. Use the tool. Do not expect the tactic to be scarce.

FAQ