The Top Indie Hackers Are Quietly Going Dark — Here's Why That Changes Building in Public

I've been watching indie Twitter for about three years, and something shifted in the last two months. Scroll Indie Hackers today and the MRR screenshots are thinner. The "$X MRR → $Y MRR" tweets are rarer. Several founders who were loud about revenue 12 months ago have scrubbed their public numbers. Indie Hackers themselves ran a post last week titled something close to "top indie hackers are suddenly disappearing."

The first read is the lazy one: burnout. The 2023–2024 build-in-public cohort is tired, the shine wore off, they moved on to full-time jobs. There's probably some of that.

The real read, after talking to five people in this cohort and reading the new ghost-posts carefully, is different. AI vibe-coding has made good ideas trivially cloneable, and transparency has become a liability rather than a marketing channel. The specific mechanism is new, the founders affected are paying attention, and the tactical answer is not "stop building in public." It's "build in public about different things."

If "build in public" has been your whole growth strategy, the game just changed a level. Worth walking through why.

The Shift That Actually Happened

From about 2020 through 2023, building in public was a genuine unlock for indie founders. Transparency built audience, audience drove distribution, distribution beat incumbents. You tweeted your MRR, you tweeted your stack, you tweeted your lessons, you shipped the product in front of everyone, and a real percentage of your readers became customers. The tactic worked because the copycat cost was high. A reader who loved your approach could not easily ship a clone in a weekend. Building something that actually worked still required months of iterating.

2026 is different. AI vibe-coding has collapsed the copycat cost. A tweet-thread teardown of your product, plus a look at your public TypeScript types, plus an afternoon with Cursor or Codex, is now enough for a competent developer to stand up a clone. I have personally watched this happen. Three separate products I follow have had near-identical clones ship inside the window between the original founder posting an MRR milestone and the next week's milestone.

The copycats don't always win. Distribution, brand, and trust are still real moats, and a clone that ships in a weekend doesn't inherit any of those. But the clones do siphon a measurable percentage of the top-of-funnel traffic. They confuse the search results. They eat SEO. They sit on the HN front page with "I made this in a weekend" posts. The specific founder whose idea got cloned has a slightly worse month than they would have otherwise.

When that happens to you twice, you notice. When it happens to your friend and they tell you about it, you notice harder. After a cluster of that, the rational thing is to stop broadcasting the specific data that makes the wedge copyable.

The Specific Mechanism

The input that makes a clone cheap is not "I read a tweet thread." The input is: a specific problem statement, a visible product surface, a public stack, and a revenue signal that the idea is working.

Each piece of that costs the founder nothing to share and saves the cloner roughly a week of validation work.

A tweet that says "I built a tool for X audience to do Y, using Z stack, and I hit $5K MRR in three months" is a complete business plan for a cloner. The problem is validated. The audience is named. The stack is specified. The business model is confirmed. The revenue milestone tells the cloner "this is worth copying, not ignoring."

Without AI, that complete plan still required months of execution to validate. With AI, the execution gap closed. The months of execution become a weekend.

The specific revenue milestone is the trigger that moves a cloner from "maybe I'll look at this someday" to "I'm opening Cursor right now." Founders who hit $3K–$15K MRR are in the worst spot — big enough to attract copycats, small enough that the clone can meaningfully dent their trajectory. Founders below $1K MRR aren't worth cloning yet. Founders above $50K MRR have a brand and distribution moat the clones can't touch.

Who's Going Dark Vs. Who's Staying Loud

The pattern is legible if you watch for it.

Going dark: a specific cluster of $10K–$50K MRR founders have quietly stopped posting numbers. Not dramatic exits. Just a gradual fade. Their Twitter bios no longer say "$20K MRR indie maker." Their Indie Hackers profiles have stopped updating. Their product pages still exist. They still tweet about their craft, their lessons, their failures — but the specific revenue numbers and the specific stack details are gone.

Staying loud: two groups. Founders below $3K MRR who need the audience more than they need the wedge protection, and founders above $50K MRR who are past the copycat window. Pieter Levels still posts numbers. Marc Lou still posts numbers. At their scale, the clones don't matter — the distribution and brand are the moat, not the specific product. The new $2K MRR founder still posts numbers because without the audience they don't have a business at all.

The middle tier is where the posture changed. That middle tier is also where most of the indie hacker cohort actually sits.

The Nuance Most Takes Miss

"Is building in public dead?" posts will dominate indie Twitter for the next six months. Most of the takes will either cheer ("transparency forever!") or doom ("AI killed indie"). Neither is useful.

The tactic isn't dead. The scope changed.

Post about your craft, your lessons, your failures, your weekly shipping rhythm. These are not clonable. A weekly post about "I shipped X small feature and learned Y from it" builds audience and trust and is basically impossible to weaponize against you. Nobody clones your reflections. They clone your business plan.

Stop posting the specific revenue numbers, the specific stack details, and the specific problem-plus-solution pairing. These are the clonable pieces. A post that says "I had a great month for my indie project" is fine. A post that says "my project for X audience doing Y with Z stack hit $8K MRR this month" is a business plan you're giving away.

The distinction is crisp: share the process, protect the wedge. Share the lessons, protect the numbers. Share the failures, protect the stack details that would let someone clone the successes.

My Own Shift

In the last 60 days I've stopped sharing three things.

Exact revenue numbers for specific projects. I used to tweet monthly MRR screenshots. I don't anymore. I'll share rough qualitative framings ("this was a good month," "this project crossed a milestone") and I'll be honest about direction, but the specific dollar figures are no longer public. The tradeoff: I lose some audience growth I would have gotten from number-worship. The upside: my actual wedge is less targeted.

Stack specifics for the products making real money. For Solo Operator, I'm completely open about the stack — Astro, Keystatic, Vercel, the blog is the product. For the other things I run, I've stopped publishing the full stack details. I'll write about individual tools when the lesson is generalizable, but I won't publish a "here's my exact stack for product X" post anymore.

The specific problem-plus-solution pairing for validated ideas. If I've found something that works, I'll talk about the general category — "this is a kind of problem worth solving" — without specifying the exact audience and exact feature set that makes it work for me. Again, the tradeoff is real: I'm less visibly "helpful" than I used to be. The upside is that my specific validated ideas take longer to get cloned.

What I still post, and more of now than before: craft posts, lesson posts, failure posts, thinking-out-loud posts. The long-form stuff on this blog. The weekly shipping logs. The specific AI-workflow posts where the lesson is general enough that the "clone" is just "have a better workflow." None of that is clonable. All of it is useful audience material.

The Contrarian Steelman

It would be wrong to skip this. Pieter Levels and the Nomad List cohort argue, loudly and often, that copycats never actually matter. Their claim is that distribution and brand eat "better product" every time, and that worrying about clones is a waste of founder attention that could be spent on the real moat: shipping, trust, and audience depth. They have real evidence for this — Nomad List has been cloned dozens of times and is still Nomad List. Marc Lou's ShipFast was cloned within weeks and still prints money.

That steelman is worth sitting with, even if you disagree. Two honest reads on it.

First: at Levels-scale and Lou-scale, the steelman is right. Above a certain brand gravity, clones are noise. The distribution advantage is so big that a clone cannot pierce it.

Second: below that scale, the steelman is wrong. A founder at $5K MRR does not have Levels-level brand gravity. A clone can cost that founder 20% of their top-of-funnel. The advice "clones don't matter" is correct for people who have already won. It is optimistic advice for people who have not.

If you are already past the event horizon — real brand, real distribution, real audience depth — post whatever you want. If you are still building that moat, be selective about what you give away.

Practical Takeaway

Three things, in order of importance.

Audit what you've shared about your products in the last six months. Not your lessons. The specific numbers, stack, and problem-solution pairings. Ask yourself: if a competent AI-assisted developer read all of this, could they ship a clone in a weekend? If the answer is yes, the next six months of posts should not add new data points to that clone kit.

Shift your public surface toward craft and process. Write about how you work, what you're learning, what's breaking. That content builds audience and does not help cloners. It's also, honestly, more useful to readers than MRR screenshots.

Keep a private log for the specific stuff. Your MRR, your stack, your specific product decisions — all of this should still be written down. Just not in public. I keep mine in a personal Notion. I still look at the numbers every week. I just don't tweet them anymore.

Building in public is not dead. Building in public about the specific things that make your wedge copyable — that's dead. If the mental model shift takes a Saturday afternoon, it's a cheap one.