A "transformation" usually arrives as a technology decision, a new platform, a cloud migration, an AI programme, and a budget to match. But the systems are rarely what sinks it. What sinks it is the slower, messier work underneath: getting people to change how they make decisions, serve customers and run the business, and getting them to keep doing it after the consultants have gone home. The evidence on this is unusually consistent, and it points the other way to where most of the spend goes.

The quick version

  • Business transformation is a deep, deliberate change to how an organisation operates, its strategy, processes, structure and culture. Digital transformation is the version driven by technology, but the goal is still a business change, not a software install.
  • Most efforts fall short. McKinsey's long-running research puts the success rate below one in three; BCG's study of 825 companies found only about 30% met their targets and stuck.
  • The thing that separates the winners isn't better technology, it's leadership and people. MIT's "digital masters" out-earn their peers, and the scarce ingredient is the capability to drive organisational change.
  • The practical shift: spend less of your attention on the platform and more on urgency, a guiding coalition, relentless communication, and the everyday work routines you are actually trying to change.

The idea in depth: two words, one job

Start with the vocabulary, because the two terms get used loosely and the difference matters. Business transformation is a fundamental, intentional change in how an organisation creates and delivers value, touching strategy, operating model, processes, structure and culture at once, not a single project. Digital transformation is the subset of that work where digital technology is the lever: cloud, data, automation, AI, new digital products or channels. The trap is hearing "digital" and treating it as an IT programme. It isn't. As MIT's George Westerman and his co-authors argue in Leading Digital (Harvard Business Review Press, 2014), the discipline is "even more about transformation than it is about digital", and more about leadership than technology.

So the move is to name, on day one, the business outcome you are buying, a faster cycle time, a lower cost-to-serve, a new revenue line, and hold the technology accountable to it. A transformation defined as "migrate to the new platform" has no finish line that a customer would recognise. One defined as "cut the time to onboard a new client from three weeks to three days" tells everyone what changing actually looks like.

The idea in depth: why most efforts fall short

The failure statistics deserve care, because a sloppy "70% of transformations fail" gets repeated everywhere with no source attached. Here is the defensible version. McKinsey's "Losing from day one" (2021), drawing on fifteen years of its transformation research, reports that fewer than one-third of respondents say their company's transformation succeeded at both improving performance and sustaining the gains. Boston Consulting Group's "Flipping the Odds of Digital Transformation Success" (2020), based on a survey of 825 executives, found only about 30% met or exceeded their target value and produced lasting change, another 44% created some value but missed, and 26% effectively failed. Two independent bodies, similar numbers: roughly a third succeed.

The reasons are old, which is the uncomfortable part. John Kotter's "Leading Change: Why Transformation Efforts Fail" (Harvard Business Review, 1995) studied more than 100 companies and found the same handful of errors recurring: no real sense of urgency, no strong guiding coalition, under-communicating the vision, declaring victory too early, and never anchoring the change in the culture. None of these is a technology problem. BCG's six success factors, integrated strategy, leadership commitment from the CEO through middle management, high-calibre talent, agile governance, monitoring, and the tech itself, read like Kotter updated for the platform era. The technology line item is one of six.

flowchart TD
  A(["Transformation programme launched"]) --> B{"Where does the
energy actually go?"}
  B -->|"Mostly the platform"| C(["Tech ships, behaviour
doesn't change"])
  B -->|"Urgency, coalition,
comms, new routines"| D(["People work differently
and keep doing it"])
  C --> E(["~2 in 3: value missed
or not sustained"])
  D --> F(["~1 in 3: targets met
and the change sticks"])
The split is roughly one-in-three across both McKinsey and BCG, and it tracks where the leadership effort went, not the size of the tech budget. Leaders Loop

So the move is to audit where your transformation's leadership attention is going, not just its money. If the steering committee spends its hour on system architecture and ten minutes on adoption, resistance and middle-manager buy-in, you have already told the organisation what you think the hard part is, and you are probably wrong. This is where the formal change models (Kotter, ADKAR, Lewin) earn their keep: they put structure around exactly the people-side work that technology plans quietly omit.

The idea in depth: what the winners actually have

If a third succeed, what do they hold that the rest don't? The most useful empirical answer comes from Westerman, Bonnet and McAfee's research behind Leading Digital. Studying large firms, they separated two distinct capabilities. Digital capability is using technology well, in the customer experience, in operations, in the business model. Leadership capability is the ability to envision and drive organisational change. Score high on both and you are a "digital master." Their finding: digital masters were meaningfully more profitable than industry peers, the book reports around 26% higher profitability, and, crucially, leadership capability was the harder and more decisive of the two to build. Technology you can buy; the capacity to move an organisation, you can't.

You can buy digital capability. The scarce, decisive ingredient is the leadership capability to make the organisation actually change.

So the move is to invest in the half that doesn't come on an invoice. That means a visible, aligned coalition of leaders, not a CIO carrying it alone, communicating the same story repeatedly, and redesigning the day-to-day routines and incentives so the new way is the easy way. Buying tools while leaving the operating model, the metrics and the manager behaviours untouched is the most common and most expensive mistake in the field: you get the cost of the platform and almost none of the transformation. The deeper play here is covered in leading transformation at scale and in choosing the right leadership style for a change that is adaptive, not merely technical.

An honest limitation. These figures are real but should be read as direction, not gospel. The headline percentages come from self-reported executive surveys, where "success" is defined by the respondent; the 26% profitability gap is a correlation in a particular study, not proof that leadership capability caused the profit. And "transformation" is a slippery category, a cloud migration and a whole-business reinvention sit under the same word. The claim that survives all the caveats is a modest, well-supported one: technology is necessary and insufficient, and the people-and-leadership side is where most value is won or lost.

A worked example

Take a mid-sized insurer, call it Meridian, that buys a modern claims platform to replace a tangle of legacy systems. (Illustrative figures throughout; this is a teaching example, not a real company.) The board approves an illustrative £8m programme with a clean promise from the vendor: claims processed in days, not weeks. Eighteen months in, the platform is live, on budget, and average claim time has barely moved. The adjusters still print forms, re-key data into the old spreadsheet they trust, and route exceptions the way they always have. The technology changed; the work didn't.

The fix is not more software. Meridian's leaders name the real outcome, cut average claim time from 15 days to 5, and make it the only number the steering group watches. They build a guiding coalition that includes the claims-floor team leaders, not just IT, because those managers are where adoption lives or dies. They communicate the why relentlessly, retire the old spreadsheet so there is no fallback, and rewire incentives so adjusters are recognised for speed-with-accuracy on the new system rather than volume on the old one. Same platform; different leadership. The number finally moves.

flowchart LR
  A(["£8m platform
bought & live"]) --> B(["Claim time barely
moves, old habits"])
  B --> C(["Name the outcome:
15 days → 5 days"])
  C --> D(["Coalition with claims
team leaders, not just IT"])
  D --> E(["Retire the fallback;
rewire incentives"])
  E --> F(["Behaviour changes,
the number moves"])
Meridian's turnaround used no new technology, only the leadership work the original programme skipped. Leaders Loop

The lesson is the article in miniature. The £8m bought a capability; it did not buy a transformation. The transformation came from the unglamorous, un-purchasable work of urgency, coalition, communication and incentive design, and it could have come eighteen months earlier if Meridian had treated those as the main event rather than the afterthought.

Frequently asked questions

What's the difference between digital transformation and business transformation?

Business transformation is any deep, deliberate change to how an organisation operates, strategy, processes, structure, culture. Digital transformation is the kind where technology is the main lever. They overlap heavily: almost every modern business transformation has a digital component, and any digital transformation worth the name is really a business change. The danger is treating "digital" as an IT project rather than an organisational one.

Is it true that 70% of transformations fail?

The "70%" figure floats around without a clean source, so be careful with it. What the credible research supports is that fewer than a third clearly succeed: McKinsey's work puts sustained success below one-third, and BCG's survey of 825 companies found about 30% met their targets and made the change stick. "Fail" is too binary, many land in the middle, creating some value but missing the goal. The honest summary is that clear, lasting success is the exception, not that everything blows up.

If most fail, why attempt one at all?

Because the alternative, letting the operating model drift while customers, competitors and technology move, has its own, quieter failure mode. The point isn't to avoid transformation; it's to avoid the avoidable causes of failure, which the research shows are mostly about leadership and adoption rather than technology. The third that succeed don't have better luck or better software; they do the people-side work properly.

Who should own a transformation, IT or the business?

The business, with IT as an essential partner. When a transformation is owned by the CIO alone, it tends to be scoped as a technology delivery and stalls at adoption. BCG's success factors put leadership commitment "from the CEO through middle management" near the top precisely because ownership has to sit with the people who run the operation, not only the people who run the systems. Middle managers, in particular, are where change is realised or quietly resisted.

How long does a transformation take, and how do I keep it from fizzling?

Longer than the launch, that's the part people underestimate. Kotter's enduring warning is against declaring victory too early; the gains have to be anchored in the culture and the everyday routines before the energy fades. Plan for sustained communication, short-term wins that prove the change is working, and a deliberate handover from "programme" to "how we now work." If the new way still depends on a project team to keep it alive, it hasn't transformed anything yet.

Related in the Toolkit

Transformation is where the rest of the Change & Transformation toolkit converges, the formal change models give you the scaffolding for the people-side work, and managing resistance & driving adoption is where most transformations are actually won or lost.

Where to go next