The Apps I Quietly Killed
Not every project survives contact with users. An honest accounting of what got shut down, why, and what it actually cost — instead of the version where every product story has a happy ending.
Most of what I write here is about the products that made it — what's live, what's growing, what I learned keeping six things running at once. That's a skewed sample. Not everything I've shipped survived, and I've never actually written about the ones that didn't. Here's the honest version: what got shut down, why, and what it actually cost to let go of it.
Rent Buddy: The One That Died for the Right Reason
Rent Buddy was a peer-to-peer rental marketplace — tools, gear, skills, the kind of thing sitting unused in a garage that someone nearby would happily pay to borrow for a weekend. I built it during a private beta in 2020, and the product itself was fine. Nothing about the app was the problem.
The problem was the actual business model underneath it: a two-sided marketplace needs both supply and demand at the same time, in the same place, or neither side has a reason to show up. Convincing someone to list their tools requires believing renters exist. Convincing a renter to check the app requires believing there's already inventory worth looking at. Bootstrapping both sides simultaneously, alone, in my spare time, without capital to subsidize either side artificially, turned out to be a different kind of problem than building the product — and one I couldn't actually solve at that scale.
I shut it down once it became clear the chicken-and-egg problem wasn't going to resolve itself through incremental effort. That was the right call. It's also the shutdown I think about the least, because the reasoning holds up completely in hindsight. Some ideas need a level of capital or team that a solo indie studio genuinely isn't positioned to provide, and recognizing that isn't a failure — it's just an accurate read on what one person can actually bootstrap.
A Layman's BAC Calculator: Killed by Liability, Not Demand
This one's exactly what it sounds like — a simple blood alcohol content estimator, built in 2014 back when I was shipping small, fast, single-purpose apps to learn the ropes. It had real usage. That was actually part of the problem.
An app that estimates something people might use to decide whether they're okay to drive carries a liability profile that's completely disproportionate to how casually it was built. Any estimate can be wrong. Any wrong estimate, in that specific context, has a worse-case outcome I didn't want my name anywhere near. I killed it not because it failed, but because the downside risk of it succeeding — more people trusting a number I couldn't actually stand behind — was the wrong trade. Some ideas are more interesting to maintain the liability exposure on than they're worth building in the first place, and this was one of them, in hindsight, from day one.
Redinfinite: The One That Didn't Stay Dead
This is the complicated case, and it's worth being honest about the difference. Redinfinite — the infinite Reddit scroller — genuinely died in 2023, when Reddit's API pricing changed and the math stopped working for a solo dev running it for a few thousand users. I shut it down for real, wrote it off, moved on.
Then I rebuilt it from scratch in 2026, and it's live again, actively maintained, with its own ongoing blog. So it doesn't really belong on a list of things I "quietly killed" — it belongs on a shorter, stranger list of things I killed and then reconsidered. What made the difference wasn't that the original shutdown was wrong. It's that enough time passed, and enough changed about how I'd build it the second time, that the same idea became viable again under a rebuilt architecture. Not every killed product gets a second chance at that. Most don't. This one did, and I don't think I could have predicted that at the time I shut it down.
What Actually Killing Something Costs
The honest cost isn't really the sunk time, even though that's the part people assume. It's the specific discomfort of admitting, in public or even just to yourself, that something you built and believed in didn't work. Every one of these had a version of me who thought it was going to be the next thing. Writing the actual reason it wasn't — a business model problem, a liability problem, a market-timing problem — is a different kind of honesty than writing about what worked, and it's the kind that's easy to just never get around to.
The Bottom Line
I don't think a portfolio of only success stories is an accurate one, and I've been guilty of only telling the survivor version for a while now. Rent Buddy and the BAC calculator didn't fail because I built them badly. They failed for structural reasons that were true from the start and that I only fully understood after shutting them down. That's a more useful thing to have written down than another retrospective about the products that worked.