70% of Innovation Transformations Fail
It’s optimistic to think that all software, technology, innovation, and transformation projects intended to improve an organisation actually deliver their outcomes. After all, significant time, effort, and money are spent to make an organisation, or its products, services, and experiences, higher performing, efficient or more valuable. But after decades of observation, ~70% of transformation projects still fail to deliver their intended value. And the common reason they fail is because teams keep misunderstanding the Human Factor.
In 1994, The Standish Group, a software analysis consultancy, analysed 3,682 software projects (Chaos, 1994). Their surprising conclusion, released in the CHAOS report, has become industry legend, with the insight that ~80% of software development projects failed to deliver value.
Specifically, only 16.2 were on time, on budget and delivered with full features. 52.7% were challenged (over budget/time, reduced features) and 31.1% failed outright (or were cancelled).
Their top four reasons for failure included:
Lack of user involvement (12.8%)
Incomplete requirements (12.3%)
Changing requirements (11.8%)
Lack of executive support (7.5%)
Though people have challenged the original Chaos report’s methodology, the finding hinted at a greater turmoil in innovation and transformation project delivery. And these insights were, in-part, responsible for pushing more software projects to an Agile style of delivery, in an effort to reduce risk.
Fast-forward several decades, and a 2020 CHAOS report analysing 50,000+ global projects found that the situation had improved, a little. Now:
31% were successful (on time, budget, scope)
50% were challenged (overrun, delayed, incomplete)
19% failed (cancelled)
Large projects: <10% success rate
So, only ~70% software projects are considered fully successful. It’s better than 1994, but still worrying.
Spreading the lens a little wider, beyond direct software development, a 2020 KPMG survey of 400 US technology executives, found “51 percent of respondents on average have not seen an increase in performance or profitability from digital transformation investments” (KPMG, 2020).
In 2023, McKinsey, in partnership with the Centre for Major Programme Management at the University of Oxford, analysed 5,400 IT projects. They concluded that 56% of projects had a benefits shortfall and “17 percent of IT projects go so bad that they can threaten the very existence of the company” (McKinsey, 2023).
Bain & Company research from 2024 found that 88% of business transformations failed to achieve their original ambitions (Bain, 2024).
An analysis by the Boston Consulting Group in 2024 found that 74% of transformation projects failed to create short and long term value (BCG, 2024).
Finally, lest we think that the new and exciting Artificial Intelligence technologies help organisations defeat these trends, a survey by MIT, released in 2025 suggests:
“Despite $30–40 billion in enterprise investment into GenAI [Generative Artificial Intelligence] , this report uncovers a surprising result in that 95% of organisations are getting zero return. The outcomes are so starkly divided across both buyers (enterprises, mid-market, SMBs) and builders (startups, vendors, consultancies) that we call it the GenAI Divide. Just 5% of integrated AI pilots are extracting millions in value, while the vast majority remain stuck with no measurable P&L impact.”
Here’s the rub; reading between the lines, and aggregating across these different pieces of research, we propose there’s a common thread of challenges driving these results, which echoes the original observations in 1994 by the The Standish Group and these include:
Lack of user involvement (What are the needs and for Who?)
Incomplete requirements (What are we building and Why?)
Unclear scope or vision (What does good look like?)
Planning confusion (How will we deliver?)
Solving the wrong problem (Where’s the value?)
Notice the trend here?
These are all human-centred challenges, not technological limitations.
A bold hypothesis is that the success of an innovative transformation, whether focused on software, hardware, people, process, policy or experience, rarely fails because of technological limitations.
Instead, it struggles to deliver value because the teams involved don’t fully understand the Human Factor, the needs, requirements, and human-centred vision of why a project exists and how it delivers value to users, stakeholders, and the organisation that supports it.
It doesn’t matter if the project is building something that faces external ‘open-market’ customers, users or citizens or if it exists deep ‘back of house’, to improve the employee experience. Human needs, experiences, and the value gained from delivering on them are the binding threads that run through every major transformation.
This is abundantly clear with the most recent tranche of ‘Gold Rush’ projects, all hurrying to cram Large-Model based Generative Artificial Intelligence into products, services, experiences and processes.
Technology is, and always has been, a means to an end, not the end itself.
Nowhere is this failure of Human Centred Design more apparent than the title image of this fieldnote, the Sony Betamax; famous as a failure to understand the humans at the centre.
Released in 1975 by the now electronics and games giant Sony, the Betamax was meant to revolutionise the home-video experience, but ended up falling by the wayside, replaced by a competitor cassette technology, the doughty VHS.
From a technical perspective, Betamax had a few theoretical advantages. It’s resolution was marginally sharper (250 lines vs 240 for VHS) and it had better tape winding mechanisms, run durability, colour reproduction and the cassette was smaller.
However, JVC, the creator of the VHS, released a product that could record 120 minutes of tape (240 minutes at lower quality), and was cheaper to buy. Which meant, with the home-video market appearing in the 1980s, VHS was ready from the start to play a ‘classic’ full-length feature movie.
Betamax made assumptions about what people really needed (e.g. durability and technical excellence), but failed to grasp the insight that no-one would wanted to change a tape midway through a movie.
Crucially, the studios, distributors and first video rental stores (the other parts of a home video ecosystem) didn’t want to distribute multiple tapes for the average movie. Which meant Betamax had very little content in its library. In contrast, with VHS, JVC addressed most of the key needs of both end-users and producers, covering the entire ecosystem.
Though Sony eventually released a longer-running Betamax cassette, their product had already failed to find it’s niche and VHS went on to become the defining video technology of the 1980s and 1990s.
Which is a good reminder, if you want to make things better, and make better things, you have to start with the Human Factor; the people who use the products, services and experiences. Then you have to understand the employees, and systems that deliver. Finally, you have to grasp the value that binds everything together. None of these are technology problems, they’re people or cultural challenges; they’re Human Factors.
Focusing on the Human Factor is how you increase your chance of being in the right statistic, in with the 30% of innovation transformations that succeed; delivering value for the people that use the product or service and those that offer it.
It all comes back to the humans at the centre.
References
Bain & Company (2024) Transformation Projects. Retrieved March 2026 from https://www.bain.com/about/media-center/press-releases/2024/88-of-business-transformations-fail-to-achieve-their-original-ambitions-those-that-succeed-avoid-overloading-top-talent/
Boston Consulting Group (2024) Five Truths (and One Lie) About Corporate Transformation. Retrieved March 2026, from: https://www.bcg.com/publications/2024/five-truths-and-a-lie-about-corporate-transformation
DiscoA340 (2023) CC0, via Wikimedia Commons via https://commons.wikimedia.org/wiki/File:VHS_Tape_(Front).jpg
Dr.marioli—Wikipedia Nutzername (2017), CC0, via Wikimedia Commons via https://commons.wikimedia.org/wiki/File:Sony_betamax_cassette_L-750_SD_195min_recordingtime.png
Dr.marioli—Wikipedia Nutzernname (2017), CC0, via Wikimedia Commons via https://commons.wikimedia.org/wiki/File:Sony_Betamax_Sl-8000E_Videorekorder_1979.png
KPMG (2023) US Technology Survey. Retrieved March 2026 from: https://kpmg.com/us/en/media/news/kpmg-us-tech-survey-report-findings.html%23:~:text=new%2520KPMG%2520survey%2520has%2520found.html
McKinsey. (2023) Large Scale IT Projects. Retrieved March 2026 from https://www.mckinsey.com/~/media/mckinsey/dotcom/client_service/corporate%20finance/mof/pdf%20issues/pdfs%20issue%2045/final/mof45_largescaleit.ashx
MIT (2025) State of AI in Business. Retrieved March 2026, from https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf
The Standish Group (1994) CHOAS Report on Software Development.
The Standish Group (2020) CHOAS Report on Software Development.
.jpg){kind=link}
