Standish CHAOS Report: The Source Behind "Two-Thirds of IT Projects Fail"
The CHAOS Report is the single most-cited source on IT project failure. This page gives the methodology, sample, definitions, and the canonical citation so you can use the number with confidence.
What is the Standish CHAOS Report?
The Standish CHAOS Report is the most-cited study of IT project outcomes, published by the Standish Group since 1994. It sorts every project into three buckets: successful, challenged, or failed. In CHAOS 2020, about 31% of projects were successful, 50% were challenged (late, over budget, or under-scoped), and 19% failed outright, which is the source of the widely-quoted claim that roughly two-thirds of IT projects do not fully succeed. The full report is paywalled; the figures below are the versions reported in open secondary sources.
The classification: success, challenged, failed
CHAOS uses three buckets, defined consistently since 1994. A project is "successful" only if it delivered on time, on the original budget, and with the agreed feature set. Anything that slipped on cost, schedule, or scope is "challenged". Anything that was cancelled before completion or never used post-delivery is "failed".
| Outcome | Definition | 2020 share (all projects) |
|---|---|---|
| Successful | On time, on budget, agreed features delivered, user-satisfaction met | About 31% |
| Challenged | Completed but late, over budget, or with reduced scope | About 50% |
| Failed | Cancelled, never used, or written off | About 19% |
Shares vary slightly by report year and project-size band. Small projects fare considerably better than large ones.
Success by project size: the strongest signal in the data
Standish has argued since 1994 that project size is the single most important factor in project outcome. The CHAOS 2015 breakdown (Modern Resolution, FY2011–2015) makes the point starkly: only 6% of the very largest "grand" projects succeed, against 61% of small projects. Small projects succeed at roughly ten times the rate of the largest, which is why the report's central recommendation is to break big projects into small ones.
| Project size | Successful | Challenged | Failed |
|---|---|---|---|
| Grand | 6% | 51% | 43% |
| Large | 11% | 59% | 30% |
| Medium | 12% | 62% | 26% |
| Moderate | 24% | 64% | 12% |
| Small | 61% | 32% | 7% |
"Project size by CHAOS resolution", Standish Group CHAOS Report 2015, page 3 (Modern Resolution: OnTime, OnBudget, with a satisfactory result; FY2011–2015). Each row sums to 100%. Note that page 3 carries two tables: this one reads across a size band to its outcome split, while the second, "CHAOS resolution by project size", reads down each outcome column to show which size bands make it up, so its columns rather than its rows sum to 100%. The two are routinely confused. Figures read directly from the primary report.
Full breakdown: project success rate by size, and why size dominates the outcome ›
The trend: success has held near a third for a decade
On the Modern Resolution definition the all-projects success rate barely moves year to year. Across the five fiscal years in the CHAOS 2015 database it sat between 27% and 31%, and by CHAOS 2020 it was about 31%. The headline that "only about a third of IT projects succeed" has held for the past decade.
| Fiscal year | Successful | Challenged | Failed |
|---|---|---|---|
| 2011 | 29% | 49% | 22% |
| 2012 | 27% | 56% | 17% |
| 2013 | 31% | 50% | 19% |
| 2014 | 28% | 55% | 17% |
| 2015 | 29% | 52% | 19% |
| 2020 (Beyond Infinity) | 31% | 50% | 19% |
Modern Resolution of all software projects, Standish Group CHAOS Report 2015, page 2 (FY2011–2015 rows); the 2020 row is the CHAOS 2020 "Beyond Infinity" split as reported in open secondary sources. Each year sums to 100%. FY2011–2015 figures read directly from the primary 2015 report.
CHAOS 2020: Beyond Infinity, the latest (and last) edition
When people search for "the CHAOS Report 2020", they mean CHAOS Report: Beyond Infinity, written by the Standish Group's Jim Johnson. It is the most recent full CHAOS Report we know of, and it is the CHAOS Report the Standish Group's online store offers today, so 2020 remains the latest reference point. We have not read it: it is sold (US$450 for the digital edition), so its figures on this site are labelled as reported.
Beyond Infinity moves past the raw successful / challenged / failed split and argues that outcomes are driven by three Factors of Success: a Good Sponsor, a Good Team, and a Good Place. It names decision latency (how long an organisation takes to make project decisions) as the root cause of project failure, and traces poor decision latency back to low emotional maturity in the people making the calls.
The report also presents its outcome figures under several definitions side by side (traditional, modern, pure, and its "Bull's Eye" measure), which is why the same 2020 report is quoted with different success percentages depending on which definition a source uses. The 31% successful / 50% challenged / 19% failed figures on this page are the classic (modern-resolution) split, the version most widely cited.
Standish Group International, CHAOS Report: Beyond Infinity (2020). Edition status and the Good Sponsor / Good Team / Good Place framework confirmed against the Standish Group product listing at standishgroup.com and a contemporaneous review of the report (verified August 2026).
Every CHAOS Report edition by year, from the original 1994 (16%) to 2020 ›
Sample, methodology and limitations
Standish surveys IT executives directly. The CHAOS Report 2015 draws on more than 25,000 projects from fiscal years 2011 to 2015; larger database totals quoted for later editions come from Standish's own summaries. The exact sample frame is not published, and the raw data is proprietary, which has drawn academic criticism (Eveleens and Verhoef 2010 in IEEE Software argued the report's definitions inflate failure rates).
For citation purposes you should treat CHAOS as an industry survey, not an audited dataset. It is reliable for direction (large projects struggle more than small ones, agile delivery outperforms waterfall) but the absolute percentages should be paired with methodology context.
Agile versus waterfall: the CHAOS 2015 finding
The single most-quoted result from the CHAOS Report is the agile-versus-waterfall comparison in the CHAOS Report 2015. Across more than 10,000 software projects (FY2011–2015), scored on the Modern Resolution definition (OnTime, OnBudget, with a satisfactory result), agile projects succeeded at almost four times the rate of waterfall projects, and waterfall projects failed at roughly three times the agile rate. The gap widens sharply with project size: waterfall barely scales, while agile holds up.
| Project size | Method | Successful | Challenged | Failed |
|---|---|---|---|---|
| All sizes | Agile | 39% | 52% | 9% |
| Waterfall | 11% | 60% | 29% | |
| Large | Agile | 18% | 59% | 23% |
| Waterfall | 3% | 55% | 42% | |
| Medium | Agile | 27% | 62% | 11% |
| Waterfall | 7% | 68% | 25% | |
| Small | Agile | 58% | 38% | 4% |
| Waterfall | 44% | 45% | 11% |
CHAOS RESOLUTION BY AGILE VERSUS WATERFALL, Standish Group CHAOS Report 2015, page 7. "Successful" here uses the Modern Resolution definition (OnTime, OnBudget, with a satisfactory result), not the older OnTime/OnBudget/OnTarget definition, so the success percentages are lower than in reports that use the traditional metric. Figures verified against the primary report.
Full breakdown: agile vs waterfall success rate, by project size ›
How to cite
Document: CHAOS Report 2015 (PDF, copy hosted by Info-Tech Research Group), read 8 October 2026. CHAOS 2020: Standish Group product listing (paid; we have not read it, so its 31/50/19 split is as reported).
Direct CHAOS report PDFs are paywalled. The PMI library and IEEE Software contain peer reviews and secondary citations that are openly accessible.
Related references on this site
- CHAOS Report by year: every edition from 1994 (16%) to 2020
- Agile vs waterfall success rate: 39% vs 11%, full table by size
- Project success rate by size: 61% small vs 6% grand
- McKinsey: large IT projects 45% over budget
- PMI Pulse of the Profession: 43% of projects over budget
- IT project budget overruns: full sector page
- Statistics hub with cross-industry data
Frequently Asked Questions
What is the Standish CHAOS Report?
The CHAOS Report is a periodic study of IT project outcomes published by the Standish Group since 1994. It classifies projects as successful (on time, on budget, with the agreed features), challenged (partially failed: late, over budget, or reduced scope), or failed (cancelled or never used). The most-cited figure is that only about a third of projects succeed, with the rest split between challenged and failed.
How many projects does the CHAOS Report sample cover?
The CHAOS Report 2015 covers more than 25,000 software projects from fiscal years 2011 to 2015, about 5,000 a year, and its agile-versus-waterfall table more than 10,000. Sample sizes quoted for other editions, including the paid 2020 report, come from Standish's own summaries and secondary write-ups. Standish does not publish an open dataset, so external verification is limited; this is one of the report's recurring methodological critiques.
Is the Standish CHAOS Report publicly available?
Partly. The CHAOS Report 2015 is publicly available as a PDF hosted by Info-Tech Research Group, and it is the edition whose figures we have checked directly. CHAOS 2020: Beyond Infinity is sold by the Standish Group (US$450 for the digital edition when we checked on 8 October 2026); its figures circulate in secondary write-ups, so cite them as reported.
Is there a newer CHAOS Report than 2020?
No. CHAOS 2020, subtitled Beyond Infinity and written by the Standish Group's Jim Johnson, is the most recent full CHAOS Report we know of, and it is the CHAOS Report the Standish Group's online store offers today, so the 2020 figures remain the latest reference point. Beyond Infinity introduced three Factors of Success (Good Sponsor, Good Team, Good Place) and named decision latency as the root cause of project failure.
How does CHAOS project success vary by project size?
Size is the single strongest predictor of outcome in the CHAOS data. In CHAOS 2015 (Modern Resolution, FY2011-2015), only 6% of grand (very large) projects were successful, against 51% challenged and 43% failed. Large projects succeeded 11% of the time, medium 12%, moderate 24%, and small projects 61%. So small projects succeed at roughly ten times the rate of the very largest, which is the empirical basis for Standish's advice to break big projects into small ones.
How has the CHAOS project success rate changed over time?
On the Modern Resolution definition (OnTime, OnBudget, with a satisfactory result), the CHAOS 2015 report shows the all-projects success rate held roughly steady across FY2011-2015: 29% successful in 2011, 27% in 2012, 31% in 2013, 28% in 2014, and 29% in 2015, with failures between 17% and 22%. By CHAOS 2020 the figure was about 31% successful, 50% challenged, 19% failed, so success has stayed near a third of projects for the past decade.
What did the CHAOS Report find about agile versus waterfall?
The CHAOS Report 2015 (more than 10,000 projects, FY2011-2015, Modern Resolution: OnTime, OnBudget, with a satisfactory result) found agile projects succeed at almost four times the rate of waterfall projects: 39% of agile projects were successful versus 11% of waterfall, while 29% of waterfall projects failed versus 9% of agile. The gap widens with size. On large projects, agile succeeded 18% of the time against just 3% for waterfall; on small projects the gap narrows to 58% agile versus 44% waterfall. Standish's summary was that waterfall does not scale well and agile scales much better.