1
And Why Do I Care?

Technology. It's everywhere. In our homes, in our hands, running our corporations, healthcare and banking systems, militaries, and the very apparatus of government itself. Simply put, "Our civilization runs on software."1 We are therefore very fortunate that there are a great many organizations and millions of individuals across the globe who work tirelessly to create products and services that do actually serve us well. Sometimes they do this against the odds, battling with aging or "legacy" systems to create the new features that consumers desire but which the original designers did not envisage. Yet, despite these great efforts on the part of so many, software-based technologies are increasingly failing us, with cybercrime, global system outages, and economic competitiveness becoming the norm rather than the exception. These issues stem from ever-greater consequences and implications for us all. However, you may well think, "Well . it's not my problem to solve, so why should I care about it?" After all, it's not an issue on the scale of, say, solving world hunger, right? Wrong. In financial terms, it's much, much bigger than the cost of solving world hunger, according to UN figures. Let's delve into this a bit.

As we shall learn, there is a lot of aging software that currently runs many of the systems that modern society relies upon. So much so that the aforementioned quote by Bjarne Stroustrup, creator of the widely used C++ language, which has an extensive and enduring presence across critical software systems worldwide, could reasonably be modified to state that "Our civilization runs on aging software." Not all aging software is "bad" per se, but running, updating, or improving older or "legacy" technologies can come at a significant cost, both financial and otherwise. Modernization of these technologies, sometimes expressed in terms of legacy modernization or digital transformation, is an imperative for most governments and corporations around the developed world, at least, since they have built up the largest estates of such legacy technologies and therefore face the biggest transformation challenges. We have chosen to use the term Technology-Led Transformation (TLT) throughout this book to describe change programs where digital technology plays a significant role-whether modernizing legacy systems or digitizing manual processes. This doesn't mean such initiatives must be led by the technology function; rather, TLT refers to any transformation where technology is central to the envisaged change. However, as multiple studies conducted across many countries and industries have shown, between 70% and 88% of all transformation efforts fail to deliver on their original goals. Digital transformation initiatives of any kind have the lowest success rates of all, as evidenced by the Bain survey involving 1,000 companies around the world to gauge their level of digital readiness. They found that only 5% of those undertaking some form of digital transformation initiative had achieved or exceeded the expectations they set for themselves, compared to a success rate of 12% for conventional transformations (Figure 1.1).2 A full 75% of these companies settled for dilution of value and mediocre performance, and 20% failed outright.

Figure 1.1 Conventional vs Digital Transformation.

Source: Bain risk history survey 2017 (n = 403); Digital 360 Barometer survey 2017 (n = 1012)

The range of outcomes recorded by other consultancies, including BCG and McKinsey, split broadly by total successes, partial results, and out-and-out failures, does vary somewhat, but the de facto success rates of between 5% and 30% are poor rates of return on stakeholder resources, regardless of how the overall program is viewed.

Table 1.1 summarizes the data published by three of the world's top management consulting firms. As you can see, the track record is not good. Furthermore, these figures relate to the types of companies and governments that provide us with the bulk of the products and services we rely on to live, thrive, or survive.

Table 1.1 Reported Success and Failure Rates: Sample Bain & Company, BCG, and McKinsey & Company.

Source: https://www.einst4ine.eu/digital-transformation-success-and-failure-part-i-insights-from-industry-and-grey-literature/

Source What Is Measured Achieved or Succeeded Partial Results Failed Bain & Company (2017) Success of digital transformation initiatives 5% achieved or exceeded expectations 75% settled for dilution of value and mediocre performance 20% failed to deliver, producing less than 50% of the expected results BCG (2020) Success of digital transformation projects 30% met or exceeded their targets and resulted in sustainable change 44% created some value but did not meet their targets and resulted in only limited long-term change 26% created limited value (less than 50% of the target), producing no sustainable change McKinsey & Company (2018) Success of overall digital transformation 16% have successfully improved performance and also equipped them to sustain changes in the long term 7% performance improved but improvements were not sustained 77% performance did not improve*

* Implied, but not stated

1.1 Time, Please

IT systems are composed of multiple interconnected technological components that must function together as a whole. These components typically have interdependencies, whereby component A is reliant upon components B, C, and D to function well before it can function as designed. A failure in one component can lead to cascading issues across the whole system. This fact introduces complexity and increases the risk of delays and errors during the build, implementation, testing, and integration stages of a project. This problem is particularly acute when components are tightly coupled, as tends to be the case with legacy systems. In contrast, more modern, typically "decoupled" systems mitigate such risks by isolating failures. This interdependence plays a key role in explaining why many IT projects experience significant time and cost overruns. One study of 106 software projects conducted within a single organization found that, on average, projects took 200% longer than estimated, with the worst case running nearly 700% over time.3 The data indicated that large overruns are more frequent than the company anticipated. Other studies support this conclusion, with one analysis of 72 projects across 23 organizations identifying cost overruns of up to 525% that were much more heavily skewed toward significant overruns than had been expected. The U.S. Department of Defense reports that for the fiscal year ending in 2020, IT project spending was $37 billion, and only 35% of the projects were within budget.4 A study by McKinsey and the BT Centre for Major Programme Management at the University of Oxford reports that, on average, large IT projects run 45% over budget, with some projects overrunning cost-presumably time budgets-by as much as 400%.5

One of the most comprehensive studies in this space analyzed 5,392 IT projects worth a combined $56.5 billion, measured in terms of 2015 USD.6 It found that IT project cost overruns follow a power-law distribution, meaning that while most projects experience small overruns, a few projects incur extremely large overruns, disproportionately driving up average costs and making such extreme events more common than expected under normal distributions. The findings challenge the implicit managerial assumption that overruns follow a normal distribution. This, in turn, can dangerously underestimate the likelihood and cost of large-scale failures. The study conclusively demonstrates that interdependencies among technological components cause these extreme overruns. The key takeaway is that IT projects are far riskier than traditionally assumed, and accurate risk assessment must account for the possibility of rare but devastating overruns. So if leaders assume a normal distribution of IT project cost overruns, they may be unwittingly exposing their organizations to extreme risk by severely underestimating the probability of significant cost overruns.

According to Gartner, the global technology research and advisory firm, of the $5 trillion spent globally on IT in 2024, the largest proportion of that, at $1.55 trillion, was spent on technology services.7 This aligns closely with the data from Statista of $1.51 trillion (£1.18 trillion) for 2025.8 Taking a conservative position, we will work with these lower figures. As a point of reference, by far the largest component of Global IT expenditure is accounted for by the US with 36.4%. The UK accounts for 7.5% of the total and the expenditure per capita is similar in both countries at around $1,642 per person per year; this, however, is nearly nine times the global average. So, countries such as the UK, US, and other developed nations have the most to gain by getting this...