Why do up to 70% of digital transformation projects in businesses fail? John Corley from Spanish Point Technologies on how AI might break this cycle.
Digital transformation has been a boardroom buzzword for over a decade, yet too many programmes still end the same way: a roadmap, a few pilots, and a business that feels largely unchanged. The problem is not ambition or investment, but that organisations keep confusing activity with progress, mistaking planning for momentum and pilots for real change.
A practical reset is to stop treating digital transformation as a single programme. Instead, it’s a set of investments that must be balanced across four areas.
“Leaders should ask whether effort and investment are balanced across the areas that matter”
McKinsey & Company has estimated that around 70% of digital transformation programmes fail to deliver their intended outcomes because investment is concentrated in isolated initiatives rather than across the full operating model.
It’s like continuously adding apps and extensions to an old computer system until it becomes slow and unstable. What’s really required is a modern, integrated platform built to handle future demands from day one.
Tech is not the change itself
Workplace productivity and workflow applications include the everyday tools people use to communicate and collaborate, such as email, shared documents and approval workflows, and can deliver quick gains by modernising how people work day to day.
Key business systems are the core platforms that run the organisation and support customer relationships and essential operations, helping organisations operate more efficiently and consistently. Innovation focuses on using technology to change how the business engages with customers or the market, for example through data, integration, AI and custom services.
Crucially, this is where organisations move from intent to execution, using innovation to create forward momentum rather than another layer of experimentation. Enabling technology, such as cloud, data and AI foundations, supports all of this change, but it is not the change itself and only delivers value when it improves how work is actually done.
Problems arise when organisations over-invest in one pillar and neglect the others. This imbalance is particularly visible in AI initiatives. Research published in 2025 by the MIT Media Lab’s Project NANDA found that as many as 95% of generative AI pilots are delivering no measurable business impact, largely because they are not embedded into how organisations actually operate day to day. Many organisations have modernised collaboration tools while leaving the platforms that support customers and operations largely unchanged. Others have invested heavily in infrastructure, but still struggle to identify measurable improvements in customer experience, decision-making, or overall business performance.
The dot-com era offers a useful parallel. At the start of a disruptive wave, it’s often unclear where real business value will come from. In the early days of the internet, the breakthrough was not simply having a web browser, but what organisations built on top of it, such as online banking, e-commerce and digital customer services. In the same way today, value does not come from choosing a particular cloud platform, but from the practical applications and services it enables.
Why do organisations keep investing in strategy documents and reports instead of building things that work in production? Often, because it feels safe. In many organisations, inaction carries less personal risk than making a decision that might fail.
Strategy has value when it clarifies opportunity and risk while action is taken, but a report is not a result. Talking about innovation is not the same as innovating, and if the strategy document becomes the output, rather than a step towards delivery, then nothing is actually changing. This is also where traditional consulting models can fall short, because documentation alone does not change how a business actually operates.
Will AI change everything?
AI shows the same pattern repeating. Many AI projects fail to reach production because they are designed as proofs of concept from the outset, with no clear ambition to deliver a live business outcome. A prototype that classifies data or generates insights can be useful for learning, but it does not deliver value unless it is embedded into how decisions are made or how everyday work is done. Too often, organisations invest in AI capability without changing the processes or behaviours it’s meant to improve.
A useful AI solution is practical and embedded in day-to-day work. Rather than attempting full automation, effective approaches use AI to do a large portion of the work while keeping a human in the loop. People step in where the system cannot confidently decide, and that feedback helps the system improve over time. Value is delivered incrementally, and when it works, users stop thinking of it as AI.
Organisations that succeed take action, accept iteration, and balance risk with delivery, improving how work is done in stages rather than waiting for a perfect end-to-end solution. The ones that struggle wait for perfection, and remain stuck in perpetual pilot mode.
Don’t ‘just do it’
One example is an Irish software company we helped rebuild its product around AI capabilities, leading to commercial growth and expansion into the US.
That shift involved redesigning core workflows within the product, integrating AI directly into decision points rather than as a separate feature, and rolling changes out incrementally so customers could adopt them without disruption. By embedding the transformation into the product itself, it enabled customers to change how they operate, creating a multiplier effect.
Similar principles can be seen elsewhere. In Ireland, the Passport Service has transformed the renewal and application process by digitising what was once a largely manual, in-person experience, improving turnaround times and the overall customer journey.
Nike has also built an integrated digital ecosystem around services such as Nike Run Club, connecting content, community and retail into a single experience that strengthens customer engagement while supporting commercial outcomes. In each case, the technology itself is secondary, the real change lies in how processes and experiences have been redesigned to deliver measurable improvements.
There are clear warning signs when transformation is going wrong, including an over-reliance on strategy documents instead of execution, or the belief that progress cannot begin until everything is in place. Leaders should ask whether effort and investment are balanced across the areas that matter, and whether they can point to real improvements as a result.
If a CEO were to do one thing differently in 2026 to avoid another stalled programme, it would be to look honestly at where time and energy are going, identify where little is changing, and take action to unlock value, even if it means moving beyond what feels familiar.
-
Bank of Ireland is welcoming new customers every day – funding investments, working capital and expansions across multiple sectors. To learn more, click here
-
For support in challenging times, click here
-
Listen to the ThinkBusiness Podcast for business insights and inspiration. Latest episodes are here. You can also listen to the Podcast on:
-
Spotify
-
SoundCloud
-
Apple
