---
title: "Startup Failure Statistics 2026: Sources and Limits"
slug: startup-failure-statistics-2026
description: "Startup failure statistics with source dates, populations and limits: CB Insights closure reasons, HBS definitions and shutdown figures reported by TechCrunch."
canonical: https://preuve.ai/blog/startup-failure-statistics-2026
author: Vincent
author_credentials: Founder of Preuve AI
date: 2026-07-04
last_updated: 2026-09-12
read_time: 8 min
---

# Startup Failure Statistics 2026: Sources and Limits

Startup failure statistics with source dates, populations and limits: CB Insights closure reasons, HBS definitions and shutdown figures reported by TechCrunch.

## Key takeaways

- **There is no single startup failure rate:** A closure count, an employer-establishment survival rate and a missed investor-return target have different denominators and endpoints. The sources here are dated, not a new 2026 cohort.
- **Failure definitions change the number:** In a March 2011 HBS interview, Shikhar Ghosh cited 30-40% for asset liquidation with investors losing most or all their money, versus 70-80% for missing projected returns. These are historical interview estimates.
- **CB Insights reports poor product-market fit in 43% of known-reason closures:** Its March 2026 report covers 431 public VC-backed closures since 2023, excluding prior exits. Reason shares use 385 companies with identifiable reasons and overlapping categories; running out of capital is most common at 70%.
- **Startup shutdowns rose 25.6% in 2024 to 966 recorded closures:** TechCrunch reported Carta customer closures, not all U.S. startups. Separately, SimpleClosure reported that 74% of observed shutdowns since 2023 were pre-seed or seed; this is a share of closures, not stage-specific failure probability.

## Scope and source dates

This September 12, 2026 editorial revision corrects source attributions and removes unsupported claims. It does not create a new 2026 cohort. A closure count, a survival rate and a missed investor-return target measure different events. A defensible probability needs a defined population, event and follow-up horizon.

Original research and provider-interview reporting are identified separately below. None of these sources estimates how much a Preuve scan changes business outcomes.

## What percentage of startups fail?

In an [HBS author interview dated March 7, 2011](https://www.library.hbs.edu/working-knowledge/why-companies-failand-how-their-founders-can-bounce-back), Shikhar Ghosh described historical estimates that depend on the definition:

| Definition in the 2011 interview | Estimate |
| --- | --- |
| Asset liquidation, investors lose most or all their money | 30-40% |
| Failure to see the projected return on investment | 70-80% |
| Falling short of a declared projection | 90-95% |

These are historical interview estimates, not a newly verified longitudinal dataset or current failure probabilities. Missing a return target does not mean returning no money. Substantial losses need not mean total loss.

## CB Insights closure reasons

The [CB Insights report published March 5, 2026](https://www.cbinsights.com/research/report/startup-failure-reasons-top/) covers 431 VC-backed companies that publicly shut down since 2023, excluding companies with a prior exit. Its evidence includes public post-mortems, founder interviews and shutdown announcements. The denominator for reason shares is 385 companies with identifiable reasons. Categories overlap.

| Reported reason | Share of 385 known-reason closures |
| --- | --- |
| Ran out of capital | 70% |
| Poor product-market fit | 43% |
| Bad timing or macro conditions | 29% |
| Unsustainable unit economics | 19% |

CB Insights interprets cash depletion as an endpoint and the other listed reasons as underlying problems. That is an interpretation of selected shutdown accounts, not a causal experiment.

The full 431-company cohort raised $17.5 billion in equity funding before closure, with a median of $11 million. Its 62 healthcare/biotech companies had raised $5.1 billion. These figures measure funding raised, not cash burned or investor losses.

The median interval from last fundraise to shutdown was 22 months; nearly a quarter closed more than three years after their last raise. This does not measure remaining cash or when founders recognized a problem.

## Industry survival and withdrawn figures

The [BLS Business Employment Dynamics survival tables](https://www.bls.gov/bdm/bdmage.htm) are the primary starting point for U.S. employer establishments by industry. An establishment is not interchangeable with a startup, a firm or every LLC registration. NAICS sectors are not the same as product categories such as SaaS.

I could not retrieve the original BLS tables during this revision. The earlier table relied on [Commerce Institute secondary reporting](https://www.commerceinstitute.com/business-failure-rate/), which was also unavailable for verification. The numerical table, industry ranking and claimed industry gap have been withdrawn. Arithmetic alone would not validate those figures.

Use a named birth cohort, consistent follow-up horizon and recorded table version. The retained sources do not establish when most startups fail. Cumulative closure by an anniversary differs from the share of closures occurring during a particular year.

## Funding stage and shutdown counts

[Mary Ann Azevedo's January 26, 2025 TechCrunch reporting](https://techcrunch.com/2025/01/26/2025-will-likely-be-another-brutal-year-of-failed-startups-data-suggests/) is based on provider interviews, not provider-native raw datasets independently checked here.

SimpleClosure reported that 74% of its observed shutdowns since 2023 were at pre-seed or seed, including 41% at seed. These are shares of closures, not stage-specific failure probabilities. The number of all companies at each stage is missing from that comparison. Carta is a separate source in the same article.

The earlier stage-probability table lacked a primary longitudinal source and has been removed. Not raising the next round is not the same event as closure. A stage-specific probability needs all companies entering that stage, a closure definition and a follow-up period.

| Provider, via TechCrunch | 2023 | 2024 | Change |
| --- | --- | --- | --- |
| Carta | 769 | 966 | +25.6% |
| AngelList | 233 | 364 | +56.2% |

Carta counted U.S.-based Carta customers leaving because of bankruptcy or dissolution. Its head of insights, Peter Walker, acknowledged missing shutdowns. AngelList reported winddowns in its own sample. Different coverage and possible overlap mean these totals should not be added together or treated as population-wide rates.

Carta sector shares reported by TechCrunch were enterprise SaaS 32%, consumer 11%, health tech 9%, fintech 8% and biotech 7%. These sector shares are not individual failure probabilities. Walker said they broadly aligned with initial funding shares.

Walker offered a working hypothesis about more funding producing more closures without an improved selection hit rate. The article's forecast for 2025 is not an observed 2025 or 2026 result. Neither establishes whether founders skipped customer research.

## First-time and repeat founders

An [HBS author interview dated February 2, 2009](https://library.hbs.edu/working-knowledge/the-success-of-persistent-entrepreneurs) reports research by Paul Gompers, Josh Lerner, David Scharfstein and Anna Kovner. Success means an IPO of a venture-backed company, not generic startup survival or profitability.

| Founder group | IPO success figure in the historical study |
| --- | --- |
| First-time founders | 22% |
| Repeat founders with prior success | 34% |
| Repeat founders with prior failure | 23% |

These figures belong to that historical interview and its endpoint. The association does not isolate preparation or show that a validation product reproduces the difference.

## Historical pivot association and practical limits

The [Startup Genome premature-scaling report, version 1.2 edited March 2012](https://s3.amazonaws.com/startupcompass-public/StartupGenomeReport2_Why_Startups_Fail_v2.pdf) studied high-growth technology startups, including unfunded firms, and explicitly lacked longitudinal data. It reported 3.6x user growth among startups that pivoted once or twice compared with those that pivoted zero or more than two times. This is an association, not a causal estimate that pivoting prevents failure. It does not establish a current universal failure rate or a venture-return definition of failure.

For your own idea, test specific assumptions with potential buyers. A [free Preuve scan](https://preuve.ai/app) can help organize desk research, but customer commitments and business outcomes need separate evidence. The [Preuve assessment benchmark](https://preuve.ai/blog/startup-validation-benchmarks-2026) describes model scores, not observed startup survival.

## FAQ

### What percentage of startups fail?
A defensible rate needs a defined population, outcome and observation horizon. BLS tracks employer-establishment survival; shutdown trackers count closures in their own samples. In a March 2011 HBS interview, Shikhar Ghosh gave different historical estimates for liquidation, missed projected returns and missed projections. None is a universal 2026 startup failure probability.

### What is the number one reason startups fail?
Running out of capital is the most cited reason at 70% in CB Insights’ March 2026 report; poor product-market fit appears in 43%. The denominator is 385 companies with identifiable reasons, within 431 VC-backed companies that publicly closed since 2023, excluding prior exits. Categories overlap. CB Insights interprets cash depletion as an endpoint and PMF as an underlying cause; this is not a causal experiment.

### What industry has the highest startup failure rate?
This revision does not rank industries. The earlier table and its ranking could not be verified against the original BLS cohort data and have been withdrawn. Use BLS Business Employment Dynamics survival tables for a specified birth cohort, horizon and NAICS industry. They measure employer establishments, not all startups or every business registration.

### Do most startups fail in the first year?
The sources retained here do not establish when most startups fail. That requires following the same birth cohort over time. Cumulative closure by an anniversary is different from the share of all closures that occurs during a particular year.

### How many startups shut down in 2024?
TechCrunch reported on January 26, 2025 that Carta counted 966 closures in 2024 versus 769 in 2023, up 25.6%, among U.S.-based Carta customers leaving due to bankruptcy or dissolution. AngelList reported 364 winddowns versus 233, up 56.2%, in its own sample. These provider-interview figures are not a census or independently obtained raw provider datasets.

Canonical HTML version of this post: https://preuve.ai/blog/startup-failure-statistics-2026