Worldwide Information Technology (IT) expenditures for goods, services and IT staff this year will easily top two trillion dollars.[1] More money will be spent on IT in the US than on plant and equipment combined. Look at the capital budget of any public or private sector organization and it’s a safe bet IT is eating up most of the capital. And there’s a good reason. You can’t launch products, create value or serve customers today without involving large doses of Information Technology. To stay relevant a company or institution must invest in technology at a level calibrated by the competitive environment. And this investment has paid big dividends in terms of improved productivity across the face of our economy.  

But there is a dark side to this investment. A great deal of the money expended on IT is wasted. In fact failure and rework associated with IT related spending is the single largest source of waste in our economy. Nothing else even comes close. If you want to read some really scary stuff have a look at the IEEE’s “Hall of Shame” of spectacular IT failures over the past decade drawn from sources like the Wall St. Journal, Business Week, CEO Magazine, Computer World, Info Week, Fortune and the New York Times. The list includes a who’s who of corporate and government institutions and the waste involved is staggering. You’ll see write offs of $33 million, $45 million, $61 million, $3.4 billion, etc. It is a litany of waste on a scale seen nowhere else in our economy. 

Depending on whose figures you use or what definition of success you apply, failure rates in IT related initiatives run about 71[2] percent. For example if you define success as delivering promised value on time and on budget then most IT investments fail. If you define success as delivering some of the originally promised value on time and on budget or even late and over budget then 53%[3] of investments succeed. And if you define success by pretending that everything you do not have to completely write off is a success, then things look rosier. This may give more control of your ledger but certainly not your profit margins. But surely no one would accept a definition of success not linked to value delivered and counting only admitted write offs? Well believe it or not that’s pretty much the definition used in corporate America today. That explains why there is so much more data available on public sector IT disasters than private sector failures. It isn’t because the public sector is all that more wasteful. It’s because they have more stringent disclosure requirements backed up by Auditor Generals at various levels of government. The instinct in the private sector is to hide or obfuscate the problem. There are no regulatory or generally accepted accounting practice (GAAP) requirements—Sarbanes-Oxley notwithstanding—stipulating that the true cost of IT-related waste be reported short of large (material[4]) write downs. And not surprisingly write downs are unwelcome in most corporate settings because they are embarrassing, hurt careers and scare shareholders. When they are reported at all they are usually buried as deep as possible in shareholder or regulatory documents. For example, look at McDonald’s’ Global ERP system intended to link every outlet globally to a central resource planning system. Like so many other ERP systems it didn’t work but McDonald’s to their credit “fessed up.” According to Securities and Exchange Commission documents filed by McDonald’s, the company realized the project was a failure only after spending $170 million on consultants and initial implementation planning activities. McDonald’s ultimately disclosed the write off in a paragraph buried within a 2003 SEC filing which talked of "management's decision to terminate a long-term technology project.[5]"  At least McDonald’s came clean; most corporations don’t. A recent survey by Information Week asked respondents the following question, “Have you ever worked at a company that experienced a multimillion dollar IT fiasco”? Seventy percent, that’s right 70%, answered yes. How many of these turned up in annual reports or SEC documents? Chances are very few.

How does this slight of hand work? First the costs of a large project are buried in multiple budgets and not necessarily aggregated. So project managers can pick and choose which expense to recognize when calculating the loss. Second, the lack of authoritative accounting guidelines, particularly in the US, has led to increasing diversity in accounting for internal-use software costs. So when it comes time to determine how much to write down the choice of which budgets to attach can be highly selective. Then there is the old chestnut of “residual value,” whereby the company claims the residual value of technology and learnings for future reuse have sufficient value to reduce the write down. Finally, there is the “shrinking requirements gambit” whereby the definition of success is reduced to fit the resulting, anemic deliverables and thus declared a qualified success. The trouble with all this is that corporations are numbers driven and so long as the numbers reflecting the true nature of IT investment waste are obfuscated, it is difficult to create the focus necessary to address the problem.

So what we see is an astonishing prospect. The largest source of waste and inversely the greatest opportunity to improve shareholder value in our economy is today largely ignored because no one wants or is legally required to measure it accurately. And the problem isn’t just at the aggregate level. At an individual project level a game of “Liars Poker” takes place as investment champions—both tech and business—underestimate the likely cost in order to gain initial approval. Thus many investments start out under-funded from the get go. Add to this arbitrary and often unrealistic delivery schedules, coupled with weak governance and poor requirements planning, and you have recipe for disaster.

The Gap in GAAP

Intangible assets—e.g., internally-developed or externally-acquired software, patents, trademarks, etc.—have increased over the past two decades both in significance and value, becoming much more critical today in the value equation than “hard” or tangible assets. The lack of accounting or regulatory guidance has led to diverse, inconsistent accounting for internal-use software costs and other intangibles, at a time when Boards, shareholders and corporate executives require transparency of financial information in order to make informed decisions about a company and its future performance. Today these constituents are kept in the dark about the financial performance of IT investments, they lack realistic information on what is driving a company’s management performance regarding the ROI of its IT investments, and have insufficient information to effectively benchmark a company and its delivery efficiency against its competition.

Existing regulatory and standard-setting systems are struggling to keep up with this rapidly changing environment. While these gaps are to some degree recognized by Finance and Accounting Boards and the SEC, the pace of reform lags well behind the problem. It is encouraging however to note that firms with rigorous, well-designed investment governance programs have at least 20% higher profits.[6]

New Way of Thinking About the Problem—Myth Busters

Einstein once remarked that a difficult problem cannot be solved within the context or paradigm that created it. That’s a good starting point for thinking about a way out of the present mess. We can begin by exploding some flawed assumptions about the problem:

• Myth 1: Technology itself is part of the problem. It’s risky, unpredictable and unmanageable. Wrong. Unless you’re playing around on the bleeding edge of technology (no place for most corporations) then the black box is rarely the root cause of disaster. At most it accounts for 10-15% of the risk in the majority of technology investments. The real investment killers are poor governance (especially horizontal governance), weak business cases, weak value propositions, poor requirements planning, weak management protocols and inadequate risk management.
  
• Myth 2: People are the problem particularly our technology people. Wrong again. While it’s possible to have a particularly inept IT Department, this occurs statistically no more often than with other disciplines like accounting and marketing. So firing the CIO (the normal knee jerk response) probably won’t help.
  
• Myth 3: Technology results cannot be accurately measured and therefore it is a waste of time to try. Not so. Technology results—like value delivered, on time and on budget—can be measured but up to now there has been little incentive to do so.
  
• Myth 4: Risk in technology investment cannot be measured before the fact. Wrong again. The inherent risk in any normal technology investment can be measured reasonably well thus allowing for appropriate risk mitigation. Again up to now there has been little incentive to measure risk because it exposes the Liar’s Poker game and fixing the resulting exposures can slow down delivery.
  
• Myth 5: Present protocols for governing technology investments are adequate. This is the biggest lie of all. Governance and overall management of technology investments are almost universally weak and in particular the assessment and mitigation of risk is not incorporated in most company’s processes.
  
• Myth 6: This is a tech problem; let them solve it. Wrong, wrong, wrong. This is a management problem that cannot be solved by the IT Department—not now, not ever. The solution lies in business rather than tech leadership.

If We Measure It, We Can Fix It

The truth is that tools exist today that allow us to measure the risks inherent in technology intensive investments early enough in the product development lifecycle to allow for adequate mitigation but few organizations use them. Even when they are used, they are often applied in a haphazard manner, or worse are voluntary. Thus high-risk investments get funded and pushed into the development cycle without ever being adequately pressure tested for risk. Measuring waste is another example. The capability exists to measure and track value promised verses value delivered including cost and schedule performance. It has not been done because nobody in charge (including the CEO and CFO) has insisted that it be done-- nor is there an accounting or legal mandate to do so.

So finding a way out requires us to change our beliefs and assumptions, beginning with the acknowledgement that this is first and foremost a business issue rather than a tech issue—thus necessitating strong business leadership. We must further understand that technology waste can be tracked, technology investment risk can be measured and mitigated, and that our management protocols for governing tech investment are inadequate. This means undertaking to measure things that we have not measured in the past because we thought it impossible or unnecessary. Finally, it means using these measurements to hold people accountable for the performance of tech investments. This may all sound simple enough but in truth it represents a revolution in how corporate America thinks about this problem.

As John Adams profoundly observed, “A revolution occurs in the mind before it is manifested in the street.” We need just such a revolution of the mind today but it will take a push from the top—from those accountable for protecting shareholder value—to set it in motion. This is one revolution that will only be driven top down.

[1] Forrester Research Inc., as published in CIO Insight March 2007
[2] Standish Research Report – Chaos 2004
[3] ibid
[4] The fundamental concept of Materiality is generally based on the 5% rule, which holds that reasonable investors would not be influenced in the investment decisions by a fluctuation in net income of 5% or less. Nor would the investor be swayed by a fluctuation or series of fluctuations of less than 5% in income statement line items, as long as the net change was less than 5%.
[5] Eight Expensive IT Blunders by Paul McDougall,  InformationWeek, October 16, 2006
[6] Gartner Group and the Massachusetts Institute of Technology 2003 study