Software projects frequently miss budget, timeline, or outcome expectations. Oxford research discussed by PMI in 2026 found that only 0.5% of projects in a database of more than 22,000 major initiatives across industries delivered on budget, on time, and with their promised benefits. BCG's 2024 research found that only 30% of surveyed companies running large-scale technology programs met their timeline, budget, and scope expectations.
AI-powered no-code platforms like Hercules let business professionals build working software by describing what they need instead of coordinating traditional development teams. For suitable projects, this approach can reduce upfront development costs, shorten timelines, and let teams make changes without relying on external developers.
Key takeaways
- Budget overruns remain substantial: IT projects ran 73% over budget on average in Flyvbjerg and Gardner's 2023 research.
- Large-scale initiatives face substantial risk: Projects in the extreme-overrun group averaged a 447% cost overrun. BCG found that only 30% of surveyed companies running large technology programs met timeline, budget, and scope expectations.
- Build software around existing workflows: Helixon Law built more than 40 applications on Hercules and reduced proposal preparation time from approximately five hours to 40 minutes.
- AI tools are widely used by developers: Stack Overflow's 2026 developer survey found that 65.9% of surveyed technologists used AI coding assistants or agents at work.
- Launch business applications faster: Petony Transportes launched its logistics ERP in two months rather than the eight months quoted by a development agency. The company also avoided a $230,000 development quote.
1. Only 0.5% of more than 22,000 major projects met budget, timeline, and benefit targets
Oxford researcher Alexander Budzier discussed a database of more than 22,000 major projects across 126 countries in July 2026. Only about 0.5% delivered on budget, on time, and with their promised benefits. This finding shows that meeting all three measures of success remains rare. The dataset covers capital investment projects across industries, not software development alone.
2. Only 29% of software projects succeeded in the 2015 CHAOS Report
The Standish Group's 2015 CHAOS Report classified 29% of software projects as successful, 52% as challenged, and 19% as failed. Challenged projects finished late, exceeded budgets, or failed to meet satisfactory outcome criteria. The report used a definition of success based on on-time delivery, on-budget delivery, and satisfactory results. These findings are historical and should not be interpreted as current 2026 failure rates.
3. Only 30% of companies running large tech programs met timeline, budget, and scope expectations
BCG's 2024 global research found that 30% of surveyed companies running large-scale technology programs met timeline, budget, and scope expectations. Another 35% missed all three targets, while 35% met budget expectations but missed timeline and scope expectations. The research shows that large-scale technology initiatives carry substantial risk, even at well-resourced organizations.
4. 31% of complex projects failed to achieve their full intended benefits in 2026 research
PMI's 2026 Pulse of the Profession found that 31% of complex projects failed to deliver the full scope of their intended benefits. This was more than twice the rate reported for projects overall. The finding highlights how competing priorities, organizational dependencies, and changing requirements can affect project outcomes. The research covers projects across industries rather than software projects exclusively.
5. 28% of very complex software projects failed in the 2015 CHAOS Report
The Standish Group's 2015 CHAOS Report found a 28% failure rate among very complex software projects, while 57% were challenged and only 15% succeeded. The research classified projects according to their complexity and used data from fiscal years 2011 through 2015. These historical results show an association between higher complexity and poorer project outcomes.
6. IT projects ran 73% over budget on average
Flyvbjerg and Gardner's 2023 research, summarized in a project cost database reference, found that IT projects ran 73% over budget on average. The figures come from the cost-risk base rates in How Big Things Get Done, covering projects from a larger database of more than 16,000 projects across industries. This historical finding shows that cost overruns can be substantial in large IT initiatives. The statistic should not be treated as the expected overrun for every custom software project.
7. IT projects with severe budget overruns averaged a 447% cost increase
The same Flyvbjerg and Gardner research reported that 18% of IT projects exceeded their original budgets by at least 50%. Within this extreme-overrun group, the average overrun was 447%. These outlier projects can create serious financial exposure. A 447% overrun would increase a $100,000 original budget to approximately $547,000.
8. Large IT projects ran 45% over budget and delivered 56% less value than predicted
A McKinsey and Oxford study covering 5,400 IT projects found that large projects ran 45% over budget on average and delivered 56% less value than predicted. Published in 2012, the research defined large IT projects as those with initial budgets above $15 million. The combination of higher costs and lower delivered value demonstrates why software investment decisions need more than an initial development estimate.
9. 17% of IT projects experienced problems severe enough to threaten the company
The same McKinsey research reported that 17% of IT projects experienced problems severe enough to threaten the company's existence. These high-impact outcomes represent the extreme end of project risk rather than typical performance. The finding reinforces the importance of financial controls, realistic scope, and early testing, particularly for business-critical systems.
Imjomat, a data labeling company in the Philippines, previously spent more than $36,000 on outsourced developers who had not delivered the system it needed after more than a year. Its founder then built the system using Hercules. Imjomat also reports 2.4x revenue growth after launching a Hercules-built website that helped attract new enterprise clients.
10. Organizations wasted almost 12% of project investment due to poor performance
PMI's 2019 Pulse of the Profession found that organizations wasted almost 12% of project investment due to poor performance. The research measured estimated wasted project spending across industries, not software development alone. This highlights the financial value of stronger planning, performance monitoring, and managing changes before costs increase.
11. 37% of surveyed executives cited unclear objectives as the leading cause of failed strategic initiatives
PMI's 2017 Pulse of the Profession found that 37% of surveyed executives identified unclear or unachievable objectives and milestones as the most important factor behind failed strategic initiatives. Unclear objectives can make it difficult for teams to measure progress and agree on intended outcomes. The finding covers strategic initiatives across industries rather than software development alone.
12. 86% of surveyed software companies identified changing client requirements as a recurring challenge
A 2026 GoodFirms survey of more than 100 software development companies found that 86% of respondents identified changing client requirements as a recurring challenge. Requirements often change as businesses learn more about their needs or discover limitations during development. These changes can create additional work, affect delivery schedules, and increase costs.
13. 65.6% of surveyed companies said scope creep increased project costs by 10% to 25%
In the same 2026 GoodFirms survey, 65.6% of respondents reported that scope creep typically increased project costs by 10% to 25%. Scope creep occurs when a project expands beyond its originally agreed requirements. Teams can limit unexpected spending by defining priorities, testing features early, and reviewing the cost of new requirements before approving changes.
14. 57% of companies cited insufficient business and technology leadership
BCG's 2024 technology program research found that 57% of surveyed respondents identified a lack of joint leadership from business and technology teams. When business experts are unavailable to define and validate requirements, development teams may struggle to prioritize the right work. Shared responsibility helps teams connect technical decisions to business needs.
Helixon Law, a military law firm, built more than 40 applications on Hercules. Founder Will Helixon reports a 30% increase in client close rates and a reduction in proposal preparation time from approximately five hours to 40 minutes. The firm's applications let staff use workflows built around the founder's legal expertise.
15. 81% of project professionals said projects have become more complex
PMI's 2026 research found that 81% of project professionals believed projects had become more complex in recent years. Another 37% described the increase as significant. The report identified organizational dependencies, changing technology, regulation, and competing stakeholder needs as important sources of complexity. These findings show why managing requirements and interconnected work matters during software development.
16. 30.9% of daily AI users spent at least four hours per workday using AI tools
Stack Overflow's 2026 AI usage survey found that 30.9% of respondents who used AI tools daily reported using them for at least four hours of their workday. This suggests substantial AI integration into daily technical workflows. However, time spent using AI is not a direct measure of productivity, software quality, or successful project delivery.
17. Large agile projects succeeded at 18% versus 3% for waterfall projects
The Standish Group's 2015 CHAOS Report found an 18% success rate for large agile projects, compared with 3% for large waterfall projects. While both rates remained low, the success rate for large agile projects was six times that of large waterfall projects in this historical sample. Project complexity, team experience, and organizational conditions may also affect these outcomes.
18. 24.6% of surveyed project professionals frequently used agile approaches in 2023
PMI's 2024 Pulse of the Profession found that 24.6% of respondents reported using agile approaches always or often in 2023. Agile usage was slightly higher than the 23% recorded in 2020 but lower than in 2022. The findings show that agile remains an established project management approach, although adoption has not increased consistently every year.
19. 31.5% of project professionals frequently used hybrid approaches in 2023
The same PMI 2024 report found that 31.5% of respondents used hybrid project management approaches always or often in 2023, compared with 20% in 2020. Hybrid methods combine elements of predictive planning and agile delivery. Organizations can use structured plans for major milestones while allowing teams to adjust implementation as requirements change.
20. Effective complexity management was associated with an 88% project success rate
PMI's 2026 Pulse of the Profession found that projects managed by teams highly effective at navigating complexity achieved an 88% success rate. Teams rated slightly effective or ineffective achieved a 14% success rate. This difference highlights the relationship between complexity management and project outcomes. It does not prove that any single methodology or tool guarantees success.
21. 65.9% of surveyed technologists used AI coding assistants or agents at work
Stack Overflow's 2026 Developer Survey found that 65.9% of 17,464 respondents used AI coding assistants or coding agents in their work. These tools help with tasks such as generating code, reviewing changes, and finding technical information. The figure reflects AI use among surveyed technologists, not the percentage of businesses using no-code app builders.
22. 26.2% of surveyed technologists used AI agents or automated workflows
The same 2026 Stack Overflow survey found that 26.2% of respondents used AI agents or automated workflows at work. These tools can perform sequences of tasks rather than responding to individual requests. Their adoption suggests growing interest in automating development workflows, although the survey does not measure whether projects using agents are more successful.
23. 73% of AI coding assistant users relied on them daily
Stack Overflow's 2026 AI usage research found that 73% of respondents who used AI coding assistants or coding agents used them daily. This indicates that AI tools have become part of regular development work for many users. Teams still need to test generated code, review functionality, and manage changes before using software in production.
24. 48% of surveyed technologists trusted AI outputs when they could verify them
The 2026 Stack Overflow survey found that 48% of respondents trusted AI-generated outputs when they could easily verify the results. Trust was lower for important decisions. This highlights the continued importance of human review, testing, and clear accountability when using AI-generated software.
How Hercules addresses software project failure risk
The Hercules platform is an AI-powered no-code app builder that helps businesses create custom software through conversation. Instead of writing detailed requirements documents that developers interpret months later, business professionals can describe what they need and see working software within minutes.
- Commercial-use requirements: Hercules requires a Business or Enterprise plan for commercial use, including internal business operations and revenue-generating applications. Hercules's terms distinguish AI Credits from Cloud Credits, with separate usage requirements. Businesses should review current plan entitlements and billing settings before launching production applications.
- Customer results: Petony Transportes, a Brazilian logistics company, avoided a $230,000 development agency quote by building its ERP on Hercules. The company launched in two months rather than the eight months an agency quoted.
This customer example is an individual outcome, not a guaranteed result. Avoided development quotes are not the same as verified net savings. Project complexity, ongoing platform usage, testing, and integration requirements affect the total cost and timeline of any software initiative.
Frequently asked questions
What percentage of custom software projects fail?
Failure rates vary by definition and study. The Standish Group's 2015 CHAOS Report classified 29% of software projects as successful, 52% as challenged, and 19% as failed outright. More recent Oxford research discussed by PMI in 2026 found that only 0.5% of major projects across industries delivered on budget, on time, and with promised benefits. These studies use different populations and definitions, so their figures are not directly comparable.
Why do most custom software projects go over budget?
Common causes include unclear requirements, scope creep, communication breakdowns, and underestimated complexity. Flyvbjerg and Gardner's 2023 research reported average IT-project cost overruns of 73%. Within the subgroup of IT projects that exceeded budgets by at least 50%, the average overrun was 447%. These historical findings describe different levels of cost risk, not guaranteed outcomes for new projects.
Can no-code platforms like Hercules build complex business applications?
Yes. The Hercules app builder supports customer relationship management (CRM) systems, enterprise resource planning (ERP) systems, and operational software. MMA Plumbing built a 46-module operating system, Petony Transportes built a logistics ERP handling approximately 400 monthly loads, and Kingdom Construction built an operations platform.
Hercules provides capabilities for building user interfaces, backend functions, databases, authentication, and hosting. Availability and commercial-use permissions depend on the applicable plan and Hercules terms.
How does agile methodology improve project success rates?
The 2015 CHAOS Report found that agile projects succeeded at 39% compared with 11% for waterfall projects. Agile uses shorter development cycles with regular stakeholder feedback, which can help teams identify problems earlier and adapt to changing requirements. For large projects, agile succeeded at 18% versus 3% for waterfall approaches. These are historical associations rather than proof that agile alone caused higher success rates.
What are the hidden costs of failed software projects?
Beyond direct spending, failed projects can waste organizational resources, delay launches, and create additional work. PMI found that organizations lost almost 12% of project investment to poor performance in its 2019 report. McKinsey's 2012 research reported that 17% of IT projects experienced problems severe enough to threaten the company's existence. Before starting a project, account for testing, maintenance, hosting, integrations, and future changes when comparing traditional development with AI-powered no-code platforms.