The information frontier’s increased velocity has dramatically heightened expectations. Businesses are expected to deliver results ever faster and ever better. ‘Project Chunking’ responds to these rising expectations. It breaks projects into manageable chunks, each of which delivers incremental benefits. This tactic has several merits. First, risk is reduced because projects are smaller and less complex; moreover, later chunks learn from earlier ones and thus have faster response times to new information. Second, incremental benefits are realized earlier and more reliably (which is, of course, the flip side of lower risk/ better return). Finally, project Chunking provides more frequent choice points, making it easier to change project direction, scope or budget when needed. In this way, it makes the portfolio more flexible.

To better understand how Chunking does all this, consider Figure 1. It represents an ambitious, large-scale project scheduled in a traditional phased approach over several time periods.

The x-axis represents time, and the y-axis shows the net benefit of the overall project. As with most projects, this one requires some period of up-front investment. Then, over time, the benefits accrue and (expectedly) exceed the initial investment. With each phase of the project, the investment-to-benefit ratio changes, as indicated by the circles and the squares. Projects usually have one to several phases of net investment before net benefits are realized.

So what’s wrong with this picture? Well, nothing really. It’s a picture of a well-designed project that has properly scoped phases and will achieve benefits over time. But the loftier the project goal and the longer the time to implement, the greater the risk, as represented by the shaded triangles underneath the investment curve. Remember that we’re talking about life on the information frontier, where unpredictability lurks around every turn.

So what is the alternative? An incremental or “chunked” approach, as depicted in Figure 2.

Let’s take a closer look at the risk triangles in Figures 1 and 2. In the traditional scenario, depicted in Figure 1, there are only two opportunities to capture value. The long time-frame pushes the realization of benefit far into the future, and the corresponding “triangle” of risk is larger. In the ‘chunked’ diagram, each triangle is smaller than those in the traditional model because the organization is incurring less risk at each stage of investment. There is less risk in part because there is less investment at any given time and because each chunk realizes benefits – but also because there is more information available stage by stage. When Chunking is applied, information from earlier chunks is available to subsequent ones. Each chunk can also take advantage of new information the company gains – about its markets, its competitors or a new technology, for example.

Think about it. Companies are constantly gaining new data and clarity about conditions in the market. This data should be used to shape (or reshape) current projects, as well as assist in future funding decisions. Having more opportunities to act on new information improves project results.

Chunking also builds opportunities. Since Chunking has more frequent choice points, constituencies that will be affected by a proposed change can participate more often in project decisions. This is yet another reason why the triangles of risk are smaller in Figure 2 than in Figure 1.

So what does all this mean? Although there are times when the Chunking outcome would deliver a lesser net benefit than the best-outcomes result of a traditional project implementation in the same time frame, it may still be the preferred approach – especially when factoring in the probability of a best-outcomes result and the associated reduction in overall project risk. In fact, a study by The Standish Group concluded that the smaller the duration and team size of a project, the greater its chances of success (which it defined as being on time, on budget, and with all features and functions originally specified). A typical project with a team of six and a six-month time frame, for example, had a 55 percent success rate. This stands in stark contrast to the 8 percent success rate enjoyed by projects of more than 250 people and an average duration of 24 months. In other words, best outcomes are seldom achieved or may only be achieved through relatively high-risk scenarios. Most of the time, you are better off Chunking.

Two important caveats about Chunking. The first, implied previously, is that it isn’t always appropriate. For example, you generally wouldn’t use Chunking to build a large-scale transaction-processing system. This is a big bet with a long timeline where the scale needs to be built in early before benefits can be produced. However, there are probably more opportunities to employ this tool than you might think.

The second caveat: Chunking requires a clear road map and an explicit understanding of where the end-state lies. Companies that apply Chunking without well-constructed plans risk contracting “disassembly fever” and winding up with a patchwork portfolio. This constitutes a good idea gone wrong. As one auto industry executive noted, “Without a roadmap, the result is a rag-tag collection of smaller projects without common direction or outcome.”

A Chunk of Savings Example

Does Chunking really work? Does it make sense, out there in the real world, to break bigger projects into smaller ones?

The real-world evidence suggests that the answer to those questions is yes. Let’s look at the case of Carlson Hospitality Worldwide, which is the company that manages Radisson and other high-visibility hotel properties. When Carlson’s managers asked their board of directors to authorize a complete overhaul of the company’s central reservation system, the board rejected the US$15 million request as being too much, too fast. End of story? No. Carlson’s managers then adopted a Chunking approach, which featured flexible “choice” points at which the board could elect to stop the project but retain benefits already gained.

The team set up a number of guidelines to help design the chunks. Standalone benefits at the end of each one were, of course, a must. But they also focused on minimizing rework and keeping mutual dependencies to a minimum. (In other words, if one chunk was canceled, others could still go forward.) Another requirement was to work with legacy systems, since these would still be in place during some of the early chunks. With those plans in hand, the management team sought funding for the first chunk, which the board soon approved.

Chunking also let Carlson respond flexibly to new information not known at the beginning of the project – for example, how important connecting to the Internet would become. As information became available, some chunks were canceled and others added.

The payoff? The new system was voted best reservation system in the industry and handles more than 7.9 million room [bookings] per year. The voice-reservation chunk alone has already generated US$40 million in annual revenue.

So, in summary, Chunking

• Is a “get benefits as you go,” results-focused approach.

• Offers more frequent decision points at which the company can elect to change or modify the project’s course.

• Helps leverage new information and learning, thereby contributing to risk reduction.

• Also reduces risk by shortening time frames and reducing complexity.

Cathleen Benko is Braxton’s Global E-business Leader. F. Warren McFarlan is Senior Associate Dean and Albert H. Gordon Professor of Business Administration at Harvard Business School.

Reprinted with permission of Harvard Business School Press. Excerpted from Connecting the Dots: Aligning Projects with Objectives in Unpredictable Times. Copyright 2003 by Cathleen Benko and F. Warren McFarlan.