Percentage of Businesses That Fail, and What Killed Ours
We can tell you exactly what our own numbers say: 38 of the 406 businesses in our index are documented failures — launched, tracked, and did not work.…
We can tell you exactly what our own numbers say: 38 of the 406 businesses in our index are documented failures — launched, tracked, and did not work. Broad national failure percentages belong to government statisticians, not to us, and the U.S. Bureau of Labor Statistics Business Employment Dynamics survival data is where that percentage should come from.
Table of contents
- ·What percentage of businesses fail, and why?
- ·What the official failure rates measure — and what they miss
- ·What actually killed the ones we documented
- ·How much is each of these numbers actually worth?
- ·Why a list without failures is worse than useless
- ·How to use this without fooling yourself
- ·Frequently asked questions
What percentage of businesses fail, and why?
The percentage is not ours to publish. What we hold is 38 documented failures inside a 406-business index spanning 22 countries — each one with a named product, a stated outcome, and a cause. That is the layer a percentage cannot give you, and the layer nobody else bothers to keep.
Six of those 38 disclose a clean monthly figure before or during the failure. Their median is $8K/mo, and the range runs from $29/mo to $400K/mo.
Read that range again. A business can fail at $400K/mo. Most failure content quietly assumes failure means zero.
What the official failure rates measure — and what they miss
Official survival rates measure whether an establishment still has employees on a payroll at a given anniversary. That is a clean, auditable definition — and it is not the question you are asking. Start with the SBA Office of Advocacy small business FAQ and the BLS series above for the headline odds.
Here is what those series cannot tell you: the mechanism. Two businesses both counted as "failed" can fail for opposite reasons — one because nobody wanted the product, one because everybody wanted it and the founder walked away.
A percentage gives you the odds. It does not give you a diagnosis. Our whole reason for keeping failures is that the diagnosis is the transferable part.
What actually killed the ones we documented
Four causes, four named entries. Each one is in all indexed ideas with its evidence class attached, not buried in a footnote.
Built something nobody needed. LeadLedger — $35,000 spent across two products ($15,000 for an in-app email-capture SDK, $20,000 for a Facebook Lead Ads connector). Product one sold exactly one licence, at $300/year, to a developer in England. Product two was priced at $29/month, $99/year, or a $199 one-time lifetime licence and sold off a single webinar. Money ran out, partners lost interest, the site is dead. Founder-reported.
Grew, then churned. Yorbie — Viral Content Database + Remixer (and the churn that followed the peak) peaked at ~$8,000/month gross volume in March, dropped to ~$6,700/month over the following four weeks, and ran $0 revenue for the first two months after the founder quit his job. Churn was acknowledged on camera. Founder-reported.
Claimed without proof. Cleo (AI Content Assistant) claimed $60K MRR in 53 days with zero proof attached. The video making the claim is itself one link in the launch funnel. Unproven — and that grade is the finding.
Won, then quit the winner. Rook · Turnkey AI-Avatar Finance Channel did $47K over two months with a peak of $6,000+ in a single day, on an audience with a $24.5 RPM. The model depended on cloning a real person's voice and background. It worked until it structurally could not.
The broadest cause in the group comes from The 97% Audit: 97% of 1,000+ tracked vibe-coded apps were dead, abandoned, breached, or under $500/mo after six months. Creator-relayed, so hold it loosely — but the direction matches everything else here.
How much is each of these numbers actually worth?
The 38 failures break down as 5 third-party verified, 21 founder-reported, 6 creator-relayed, 6 unproven — 68% verified or founder-reported. Our full method is at how we grade the evidence.
| Evidence class | Count | What it means |
|---|---|---|
| ✅ Third-party verified | 5 | Confirmed by an outside source, not the operator |
| 🗣 Founder-reported | 21 | The operator stated it, often on camera |
| 📎 Creator-relayed | 6 | Someone retold another person's figure |
| ❓ Unproven | 6 | Claimed, with nothing attached |
This is the argument the site rests on: a founder-reported figure and a verified figure are not the same claim, and quoting them identically is the standard failure of business content. Most of the internet publishes the number and drops the evidence. That is worse in failure stories than in success stories, because a failure with no evidence class is just a rumour with a moral attached.
Compare the classes directly. NoFap / Men's Self-Improvement App (Jaxon) reports $6K/mo in its first month live with ~1,100 paying users and is third-party verified. Leftclick (AI automation agency) — and its founder's case for why the skill is expiring reports ~$400K/month across businesses and is founder-reported. Both are useful. They are not equally certain, and we will not print them as if they were.
Why a list without failures is worse than useless
A directory that deletes its failures is selling survivorship bias with a straight face. We keep 38 of them inside 406 entries on purpose, because the deleted cases are where the causes live.
What you lose when a list shows only winners:
- 1.The denominator. Without failures you cannot tell whether a tactic works or whether you are reading the one time it did.
- 2.The mechanism. AI Faceless Facebook Content Monetization Page is third-party verified at $32.5K/mo with 314M views in 28 days on a single page. Impressive — and meaningless as guidance unless you also know what happened to the pages that did the same thing and got nothing.
- 3.The failure modes that hit at scale. Six of our 38 failures had disclosed revenue, median $8K/mo, top of range $400K/mo. Revenue does not immunise anyone.
One caveat we state every time: this index over-represents businesses that worked and then chose to talk about it. Use the medians to size up a claim someone shows you, not as a forecast of your own result.
How to use this without fooling yourself
Four checkable moves this week, each tied to something above.
- 1.Get the base rate from the government, not from a blog. Pull survival figures from BLS or the SBA Advocacy FAQ before you accept any number in a YouTube title.
- 2.Label the claim in front of you. Verified, founder-reported, creator-relayed, or unproven. Of our 38 failures, only 5 are third-party verified and 6 are unproven — the same spread you will meet in the wild.
- 3.Write down your churn assumption before launch. Yorbie went from ~$8,000/month to ~$6,700/month in four weeks. If your plan has no churn line, it is not a plan.
- 4.Test demand before you spend build money. LeadLedger spent $35,000 and sold one $300/year licence. Structure the check with the SBA's guide to planning a business.
Further reading with the same discipline: How to Build an n8n AI Agent Business That Keeps Its Margin, App Ideas That Make Money: Revenue Evidence, Not Brainstorms, and the Why They Fail hub.
Frequently asked questions
Is it true that 90% of businesses fail?
We have no data supporting that figure and will not repeat it. Our index holds 38 documented failures out of 406 businesses — a curated sample, not a national rate. For an actual failure percentage, use BLS Business Employment Dynamics survival tables or the SBA Advocacy FAQ, both linked above.
What business has a 90% success rate?
None that we can evidence. The closest thing to a success-rate figure in our data runs the other way: 97% of 1,000+ tracked vibe-coded apps were dead, abandoned, breached, or under $500/mo after six months, and that figure is creator-relayed rather than verified. Treat any advertised success rate as an unproven claim until graded.
What percentage of businesses make $500,000 a year?
Our index cannot answer that, and neither can most sites that try. What we can say: of 406 indexed businesses, 106 disclose a clean monthly figure, with a median of $27K/mo. Among the 38 failures, six disclosed, median $8K/mo, topping out at $400K/mo.
How reliable are these percentage of business that fail figures?
Every figure here carries its class. For the failure group: 5 third-party verified, 21 founder-reported, 6 creator-relayed, 6 unproven — 68% verified or founder-reported. Across the whole index: 57 verified, 184 founder-reported, 121 creator-relayed, 44 unproven. The class is part of the number, not a disclaimer beneath it.
How many of the 38 indexed businesses actually disclose a monthly number?
Six. The rest describe outcomes too loosely to compare, and we leave them uncounted rather than estimate. Those six have a median of $8K/mo and span $29/mo to $400K/mo. 23 of the 38 were one-person businesses, which is worth knowing before you assume failure requires a team to blame.