Sometime in the not too distant future, some new Peter Drucker is going to point out how managers in the 1990s spent way too much money on information technology because CEOs were afraid, unwilling, or untrained to manage it. Instead managers blindly put their trust in technological experts whose motto was "Infallible, inflexible, and inflatable." Until finally one day, like the television anchor played by Peter Finch in the movie Network, they stuck their heads out the window and yelled, "I’m mad as hell, and I’m not going to take this anymore!" And mad they should be.

IT capital spending per white collar worker has tripled since 1980, while overall IT spending is projected to increase by 60 percent over the next five years (Exhibit 1). Yet despite all this spending, companies that manage to extract significant business value from their IT investments are rare exceptions to the rule.¹ Many CEOs have tried to improve matters, perhaps by outsourcing parts of the IT function or repeatedly changing the CIO. But little has changed. IT expenses continue to rise, major reengineering projects still languish because key systems fail to materialize, and market share and the bottom line show little of the promised improvement.

Hopefully, that Peter Drucker of the future will also point out that CEOs finally learned to resolve the problems simply by regaining control of their IT organizations. For the truth is, IT can be managed just like any other part of your business, using the same traditional managerial toolkit to make business-based decisions and trade-offs.

The IT black box

Many of today’s business leaders view IT as a "black box," unsure of what goes on within it and not particularly anxious to find out. They don’t know the right questions they should be asking, much less the answers they should be getting. They cannot penetrate the "techno-speak" of the IT group and are often forced to defer to the technologists. In short, IT often escapes the management scrutiny which is commonplace elsewhere in the enterprise.

But think about how you manage the other major functions of your business—strategy and planning, new product development, operations, and customer service. You’re probably not an expert in all these areas, yet you have a good feel for what’s on the critical path and what’s crucial to your current strategy and plans. You have a vision of where you are trying to take the company and how these functions support that vision. You don’t micro-manage, but you know how to quickly penetrate to where the rubber hits the road. You know the crucial numbers to watch to determine if things are on track.

You realize, for example, that a key objective of strategy and planning is to focus investments on areas that will have the greatest impact on the growth and profitability of the business. You therefore use measurements such as return on invested capital, and would insist on a set of payback analyses you understand and trust, when the head of manufacturing requests funds to build a new plant.

When it comes to developing new products, no doubt you assign individual accountability, say, to a brand or product manager, whose compensation and performance assessment are based on capturing revenues or market share from new product launches within a given time frame. If a product design team were to add bells and whistles that you don’t see supported by market research and you suspect will inordinately delay your product launch plans, you don’t sit still. You challenge them, and you are comfortable in doing it. In operations, you know how maintenance relates to uptime, what the trade-offs are between customer service levels and cost, and what the three or five operationally oriented numbers are that you need to look out for.

But you have a very different approach to IT. What you do not realize is that you do not need a different approach, and, in addition, companies that have been successful with IT have been because they have used an approach that is consistent with ones they use to examine other functions. Therefore, while you and your senior management team need to be familiar with the technology of IT, you don’t need detailed knowledge in order to understand how IT investments will add value and what actions should be taken. Put a stop to or just ignore the incessant techno-babble about IT. Get people to focus on the cost, time, and results of the IT projects instead, and thereby judge the performance of the IT department in business terms. We believe you can, starting immediately, and that doing so will save you unnecessary expense, and finally allow you to get some real returns from your IT investments.

IT can be managed like traditional business functions

How can the black box be opened? Consider IT, conceptually, as a traditional business system with four similar functions: strategy and planning, applications development, operations, and user support (Exhibit 2). The key objectives and priorities for the corresponding IT and business functions are strikingly similar—which means the problems can be fixed using the skills you already have.

Strategy and planning

In any business, strategy guides the major investments and activities, and you and your senior managers take an active role in communicating those priorities to the operating functions. But few companies ensure that their business priorities guide IT decisions.

The typical IT strategic plan is only loosely connected to stated business goals. Instead, it is presented as a complex set of technical dependencies and "architectural" maps, which few but the architects can understand. The IT strategic focus, it seems, is 20 percent on the "I" and 80 percent on the "T."

Well run organizations, on the other hand, integrate their IT strategy with their overall business strategy. The decisions their IT leaders make are based on real measures of business value, not politics or pet projects. IT leaders are true partners in the planning process. They understand the fundamentals of the business, such as what really drives the bottom line, the rate at which the market changes or competitors react, and the way in which IT can help to meet overall strategic objectives.

To ensure that the priorities of the business guide IT decisions, both IT and business unit managers must be involved in the creation of IT strategy. IT managers who are held accountable for the delivery of new applications must therefore have a seat at the table when business priorities are set. Similarly, managers who will ultimately be held accountable for the performance benefits promised by IT plans must have a seat at the table when IT strategies are developed.

Once business and IT leaders are working from the same agenda, they can start examining their IT strategy. The first step is make sure IT spending is aligned with strategic goals. One useful approach is to map IT spending against an ROI tree. An example is shown in Exhibit 3. Here, a petroleum exploration and production company found that 75 percent of its IT budget was being spent on administrative projects, even though the company’s growth depended upon its scientific and technical capabilities. The obvious answer was to redirect funds toward more critical areas like production costs and geological surveys. This is probably not any different from what you would do today in other parts of your business. However, do you have a spending map for IT? If not, have your finance people construct a simple ROI tree and work with the IT group to identify major investments. ROI trees don’t have to go into minute detail—a broad picture of IT spending will suffice. Once you see how spending is distributed, it will be clear what action to take.

Once IT spending has been adjusted to reflect a truer picture of business strategy, IT initiatives have to be prioritized to focus on those that are critical to the bottom line. The first step here is to demand comprehensive cost and benefits assessments from IT developers and frontline users. Cost analyses must include development, start-up, and ongoing costs. Unfortunately, many of the cost estimates projected by systems developers don’t include all relevant costs driven by new projects, including ongoing maintenance and upgrade costs over the useful life of the system. Benefits should focus primarily on quantitative benefits, which can be written into the performance commitments of individual managers, but can also include qualitative benefits like measures of operational urgency or customer mandates. Measurable business value should be the driving force, not negotiations over pet projects or the need to blindly fill users requests for more systems. If you assess IT spending as rigorously as you would any other investment proposal, it is quite likely that the number of projects in the investment portfolio will shrink.

You don’t have to waste time considering the costs/benefits of every single new project. Instead, try aggregating them into logical groups—such as those that support a reengineering project in a particular business unit or a new trade promotion strategy. These groupings can then be used to assess the net payback, without getting buried in the details.

Finally, you must play a role in establishing the sequencing of major initiatives. Paybacks can be weighed against factors such as funding limitations, required sequencing to capture business benefits, organizational change readiness, and "foundation system" necessity (for example, those which must be in place before others can be developed). Your business strategy and organizational constraints (for example, people, skills, funding) will help guide you in finding the right implementation sequence.

Applications development

Like its product development counterpart, the role of applications development is to fulfill customer needs. That means customer needs have to be understood, while development cycles have to be fast enough to stay ahead of change. In many companies, it is the area of IT which most requires top management attention.

Applications development is typically the second largest expense category (behind operations) for IT departments, and consumes many of the most talented resources in the organization. But applications development is routinely plagued by cost and schedule overruns that make it impossible to capture value from IT development in today’s business environment. Worse still, much like commercial product development, poor decisions or slow development processes can impair business performance for years to come.

Twenty years ago the typical business cycle was about the same as the technology and applications cycles. Today, however, the cycles are out of sync in most organizations. Product cycle times are typically one to three years (though the rate of change in computer and network technology is a half- to two-year cycles). And though promised applications development cycles are around one to three years, the actual applications development cycle averages three to five years (Exhibit 4). In other words, business and technology move twice as fast, on average, as IS systems development groups. IT can never deliver business value under these conditions.

Fortunately, you don’t have to be a computer programmer to figure out either how to shorten development time or keep within budget. The same problems that hinder product development abound within the IT function: ineffective project management, overscoping, group stewardship rather than individual accountability for delivery and for the capture of business benefits, and a lack of critical skills required for project execution. These similarities mean that the same fixes apply.

To ensure that new applications development is responsive enough to keep pace with business changes, shorter development times have to be made an explicit performance goal from the very outset. You should routinely demand faster development alternatives to ensure that project leaders design only core functionality and make trade-offs that accelerate delivery. Don’t let the IT group become enamored with building the most elegant, gold-plated systems that meet the potential requirements of every user. Of far greater concern is the agility with which new development can respond to market and industry changes.

Don’t accept only one answer about how to build a particular system. Demand development options. Find out if a PC-based system can be used to replace the "front-end" of an existing system rather than starting from scratch just to have a better user interface. Can IT just augment the core logic of an existing system as a short-term solution, instead of taking a year or more to rewrite the programs? Are standard software packages available to accomplish some of the new project proposals?² If the IT group’s ability to deliver the proposed design within their estimated time frame is in doubt, ask it to focus on the most important 70 percent. It might mean sticking with mature technology, or a no-frills version which won’t satisfy everybody, but there is a better chance of keeping up with business developments. (To be fair, this does not mean that one should always opt for the quick-and-dirty solution; it means that one ought to take into account all the available solutions and their life-cycle implications.)

To ensure that this sense of urgency and discipline survives as the project develops, individuals have to be made clearly accountable. Consider the development of new commercial products. To ensure timely delivery to the marketplace and meet profit or market share performance targets, accountability routinely rests with a product or marketing manager, or business unit head. Often, the individual’s compensation package is tied to meeting these goals. The same practice needs to be instigated for applica-tions development. IT managers need to be held accountable for on-time delivery. Make sure your IT leaders understand the opportunity costs if a project takes 12 months to be fully operational as opposed to the originally estimated six months; and make sure they know that schedule slippage won’t be tolerated. Likewise, business unit managers must be held accountable for the capture of business benefits. Translating performance measures into real wage and promotion incentives communicates the importance of the project in unambiguous terms.

Just as important as project accountability is the visibility of the delivery schedule: it is difficult to make individuals accountable for the success of the project if others cannot judge their progress. The project milestones against which the developers are measured must therefore be visible to the organization, which means linking them to releases of functionality which have direct impact on the performance of the business. Senior managers often try to keep up with development progress with written updates or periodic reviews that focus on technical problem resolution. A better method is a structured review process where trade-offs concerning major development releases are made and milestones such as critical release dates are watched closely.

Participants in the reviews should include not only a project sponsor, but also those managers whose organizations are directly affected by the initiative. These types of reviews are standard fare in the development of major new products, though they might be rejected by the IT department on the grounds that "writing major software code is unlike building a factory." But don’t be deterred. Good project management is good project management. Microsoft, arguably one of the world’s most successful software applications developers, uses very detailed milestones for every major project. True, they break up the development in ways that are particular to writing software code, but the general management principle is the same: milestones for major and minor sub-components are identified in advance; individual developers are responsible for their portion of the product being ready on time and defect free; and individual performance is tracked and has an impact on career advancement.³ Such controls prevent scope from creeping, and functionality from continually changing and expanding. They are no different from those used in developing a new commercial product.

The use of these periodic reviews—say, at quarterly intervals—should be applied not only to the development process but to piloting and rollout. Milestones can be laid down to ensure that the systems are tested and put into place rapidly and on schedule, and that the promised business benefits materialize.

Operations

As in manufacturing or production, the goal of IT operations is to keep the infrastructure up and running efficiently, for the lowest possible cost. Operating the technology infrastructure often consumes more than 50 percent of IT budgets, diverting precious resources from new applications development (Exhibit 5). It is extremely important, therefore, that operations be managed carefully to control costs.

If your company were at a cost disadvantage in its traditional operations, you would probably set up aggressive cost reduction targets based on the key cost categories. You should be setting the same standards for IT. Get a handle on where the major IT costs lie, and sit down with your IT leaders to develop goals which aren’t just one-time targets, but become part of a continual effort to reduce costs.

An electric utility found that by applying the same kind of cost discipline used in the management of its plant maintenance to the management of its IT operations, the IT budget was reduced by 30 percent without reducing the performance levels of its customer information systems. The key to success was the recognition that much of the so-called maintenance that occupied the organization amounted to systems gold-plating. Unnecessary upgrades, pre-emptive bug fixes, excess management of reports, and other noncritical work had grown to consume the organization.

The use of standards is another powerful tool to reduce operations cost and complexity while enhancing efficiency and flexibility. IT operations encompass a broad spectrum of technologies, from data and voice networks to user hardware and software, often leading to an expensive tangle of different systems with distinctive operations requirements. Companies should therefore try to standardize as many of these technologies as possible, minimizing the technology footprint of the organization while still maintaining some measure of flexibility. One way is to restrict business units and departments to pre-approved lists of PCs and desktop software. The business unit remains free to write their own purchase orders, but must purchase pre-approved products or configurations. One financial services company reduced operations costs by 25 percent by consolidating three data centers into one and reducing systems complexity through standardization. Though the decisions which led to the emergence of three separate data centers may have been correct in isolation, the degree of complexity their combined operation imposed on the IT department made cost containment impossible.

User support

Customer service is a critical function for many businesses. Excellent customer service may attract and retain a loyal customer base, while inefficiently managed functions can sap profitability. However, companies have to ensure that service costs are kept in line with the benefits. You must continually ask yourself: "What level of service am I willing to provide at what cost?" The same question should be asked when it comes to managing IT.

In the IT world, helpdesk and support costs typically account for less than 15 percent of the IT budget, which perhaps accounts for why so few organizations take the trouble to examine the IT department’s service performance. Typically, user support services are inefficient. Helpdesks and PC support staff rarely turn down user requests, and new helpdesks are set up to serve new systems, regardless of whether existing support staff could do the job adequately. Costs, as a result, are unnecessarily high, and service is not focused on the critical business functions.

Three principal tactics can improve cost efficiency. First, service functions should be consolidated as much as possible. A multi-billion dollar transportation company reduced the number of its helpdesks from fourteen to six while simultaneously improving the service quality. Second, the tools and skills of the user support staff should be upgraded to ensure problems are resolved quickly and correctly. Third, support staff should be pared back to essential levels only. Not all business units will necessarily need the same degree of attention, so service should be matched (and charged) accordingly, freeing resources that might be better used elsewhere.

That guru of the 21st century will certainly be able to look back on the late 1990s and see some signs of change. For though many IT departments remain an ineffectual drain on resources, CEOs at some leading companies around the world have opened the black box and now bring a different perspective to the management of IT. In the past, IT departments have often been judged by the number of people they employ, how much hardware the company has invested in, or whether it has the latest technology. But the best companies and management teams now understand that if IT is to deliver value, it has to be judged by its impact on the bottom line. Having dared to open the box and grapple with what lies within, they have discovered much that is familiar. The functional aspects of IT are analogous to those found in other parts of the business, which means the same management techniques can be applied to produce excellence. The best know which questions to ask. They know how to manage without being experts in every area. Let’s hope other companies and management teams arrive at the same point sooner rather than later.