Discovery-Driven Planning

Definition

Christensen argues that "markets that do not exist cannot be analyzed: Suppliers and customers must discover them together" — the applications for a disruptive technology are not just unknown at launch, they are unknowable. Borrowing the term discovery-driven planning from Rita McGrath and Ian MacMillan, he prescribes identifying the assumptions a business plan rests on and then testing those assumptions cheaply and sequentially, before committing capital that is expensive to reverse, rather than building a single detailed forecast and executing against it. He calls the complementary market-research posture "agnostic marketing": marketing under the explicit assumption that no one — not the company, not its customers — yet knows how or how much a disruptive product will be used.

In the Book

Chapter 7 documents that Disk/Trend Report's market forecasts, built by the same experienced analysts using the same methods, were accurate to within 8% for sustaining technologies (like the 2.5-inch drive) but were off by 265% and 550% for disruptive architectures (5.25-inch and 1.8-inch drives respectively) — the 1.8-inch drive being the first generation aimed at a non-computer market nobody had yet defined. The central case study is Hewlett-Packard's 1.3-inch Kittyhawk drive: HP committed to a high-volume automated production line and a rugged, expensive design based on a single forecast (PDA sales), which locked the product out of the price-sensitive video-game market that was actually emerging as its real opportunity. Christensen contrasts this with Honda's entry into the North American motorcycle market, where the winning application (small off-road bikes) was an unanticipated success discovered by watching how people actually used the product, not a target Honda had planned for.

Why It Matters

This distinguishes two entirely different planning postures by the state of available information: when a market is known, planning-then-executing is the right discipline; when it is genuinely unknown, the same discipline produces confident, expensive commitments to the wrong assumption. It generalizes to any venture entering genuinely new territory — a research direction, a product category, an organizational form — where the correct move is to fund cheap tests of the load-bearing assumptions before funding the plan built on top of them.