About BudgetOverrun.com
An independent reference for project budget overruns. Cited statistics by industry, a working EVM calculator, famous case studies, causes, prevention strategies, and the legal and insurance context. No project-management software to sell, no consultancy to pitch.
Why this site exists
Project-overrun statistics are cited everywhere but rarely with named sources. The often-quoted "45% over budget" figure for large IT projects traces back to a specific McKinsey-Oxford study from 2012, with conditions that buyers should know about before quoting the headline number. The 9-in-10 transport infrastructure figure comes from Flyvbjerg, Holm and Buhl's 2002 study of 258 projects, with a defined selection criterion. The Standish Group's challenged-or-failed rate for software projects (71% in its 2015 report) has a published methodology critique that should travel alongside the figure.
Most pages ranking for budget-overrun queries are written by project-management consultancies with governance services to sell. They quote the headline number without methodology, attach their own service pitch, and move on. This site does the opposite: every figure traces to a named publisher with the year, the sample, and the methodology context.
The other gap is a working tool. Earned Value Management is the industry-standard method for forecasting overruns on active projects, but most EVM material online is either PMP-prep theory with no calculator or vendor-software demos behind a sales gate. This site publishes a free EVM calculator that runs entirely in your browser, with the formulas shown and presets for the three sectors most readers come from.
Who builds this
Oliver runs Digital Signet, an independent reference-content studio that builds data-led cost and decision tools using public datasets. After two decades as a solutions architect and tech lead across media, utilities, satellite, and data, he founded Digital Signet to apply autonomous AI development methodology to real software at scale.
Reach Oliver at [email protected] or LinkedIn.
About Digital Signet
BudgetOverrun.com is part of a portfolio of cost-reference and calculator sites operated by Digital Signet. The portfolio sits inside the project-management and engineering-economics surface, so the sister sites cross-reference each other where the topics meet:
Rework cost as a share of engineering capacity. NIST $59.5B macro plus per-team math.
Technical debt economics: developer-velocity drag, refactor ROI math.
Independent CI/CD pricing comparison across GitHub Actions, CircleCI, GitLab CI, and self-hosted.
Cloud financial operations: cost-allocation, anomaly-detection, and FinOps tooling pricing.
Estimated cost of a production incident: lost revenue, engineering response time, customer churn.
Internal developer platform team economics: build vs buy, team-size sizing, productivity math.
Editorial position
This is a reference site, not a project-rescue consultancy, not a programme-governance practice, and not a software reseller. Digital Signet does not sell project-management software and does not run a project-recovery service line. The monday.com partner link is labelled where it sits and never changes a figure.
Where sources measure different things (the McKinsey-Oxford 45% average overrun on large IT projects versus the Standish share of software projects challenged or failed; the Flyvbjerg 9-in-10 transport figure versus PMI's cross-sector survey), both figures are presented with the methodology context that explains why they differ. The site does not pick a single headline number when the underlying research does not justify one.
What this site covers
Editorial principles
Every overrun statistic on this site traces to a named document: PMI Pulse of the Profession reports, Standish Group CHAOS reports, McKinsey articles, Bent Flyvbjerg's peer-reviewed papers and book, US GAO and UK NAO audit reports, and HM Treasury guidance. The sources ledger on the methodology page lists each document, the link, what we cite it for, and the date we last read it.
No figure, ranking or position on this site is paid for. One sponsorship slot is offered on the sponsorship page and is currently unsold; a sponsor would be labelled as one. The site does not sell project-management software, does not act as a project-rescue consultancy, and does not run a programme-governance practice.
Outbound links to source publishers, vendor pricing pages, and reference documents are plain URLs, with one exception: the monday.com link in the PM software boxes is a labelled partner link that may pay a commission. It sits beside plain links to Smartsheet and Wrike and has no bearing on any figure. This site is a reference and collects no leads.
Each headline percentage (43% PMI, 9-in-10 transport infrastructure, 71% Standish, 98% megaprojects) links to the document that states it, with the date we read it. Where a figure comes from a paid report or a book we could not check page by page, the page says so. Where the underlying methodology is contested (the Standish CHAOS critique, the Flyvbjerg selection question), the page flags the limitation rather than quoting the headline number alone.
The verification date is held in one constant (LAST_VERIFIED_DATE) imported by every page. Footer text, schema dateModified, and visible "data verified" labels all read from that single source, and the date moves only when a substantive source review has actually happened. Currently reads October 2026.
Where sources measure different things (for example McKinsey's 45% average overrun on large IT projects against Standish's share of software projects that are challenged or failed) the page presents both with the methodology context, not a single point estimate. Calculator outputs are working bands, not predictions for a specific project.
Methodology in brief
Statistics come from named documents: PMI Pulse of the Profession reports, Standish Group CHAOS reports, McKinsey articles, Bent Flyvbjerg's peer-reviewed papers and book, US GAO and UK NAO audit reports, and HM Treasury guidance. Each row in the sources ledger gives the document, the link, what we use it for, and the date we last read it. This is a compilation of published research; the site holds no dataset of its own.
EVM formulas follow PMI/AACE standard notation. Calculator presets are illustrative inputs for trying the formulas, not published figures. Historic case-study figures (Sydney Opera House, James Webb, Big Dig, HS2) come from audit reports, official statements and peer-reviewed papers where those exist; where only press reporting exists, the page says so. For the full framework, see the methodology page.
Contact and corrections
Spotted a stale figure, a missing publisher, or a project case study with new audit data? Email [email protected] with the page URL and the source you would like cited. Substantive corrections are typically actioned within five business days.
Disclosures
- ●One labelled partner link (monday.com, on /prevention, /templates and the three study pages) may earn a commission; every other link is a plain URL.
- ●No email-gated downloads, quote forms, or sales redirects.
- ●Not affiliated with PMI, the Standish Group, McKinsey, Oxford Said Business School, KPMG, Atlassian, Microsoft, Smartsheet, or any other named publisher or vendor, apart from the monday.com partner link noted above.
- ●Calculator outputs are working bands; production outcomes depend on project-specific risks not modelled here.