Discovery-Driven Planning for New Markets
Plan for learning rather than execution when entering markets that do not yet exist
Discovery-Driven Planning, championed by Christensen for disruptive innovation contexts, inverts the traditional planning process for entering new markets. Conventional planning assumes that market characteristics can be researched and forecasted, leading to detailed business plans with revenue projections, cost structures, and ROI calculations before significant investment. This works in established markets with known customer behaviors but fails catastrophically in new or disruptive markets where the customer, the value proposition, and the business model are all unknown. Discovery-Driven Planning instead assumes that the initial plan is based entirely on assumptions—and the primary goal of early investment is to test those assumptions as cheaply and quickly as possible. Rather than committing resources to execute a plan, you commit resources to learn whether the plan could work. Investment is staged around assumption-testing milestones: invest enough to test the most critical assumption, evaluate results, then either pivot or invest enough to test the next assumption. This approach matches the inherent uncertainty of disruptive markets with a planning methodology designed for learning rather than execution.
- In new markets, all projections are assumptions—plan accordingly
- The primary goal of early investment is learning, not execution
- Investment should be staged around assumption-testing milestones
- Failing fast and cheap on wrong assumptions is vastly better than failing slow and expensive
- Plans should be living documents that evolve with each learning cycle
- Create a Reverse Income StatementStart with the financial results required for the venture to be worthwhile (minimum acceptable revenue, profit margin, return on investment) and work backward to identify what must be true for those results to materialize. How many customers would you need? At what price point? With what cost structure? What market share? This reverse engineering reveals the critical assumptions buried in every business plan—assumptions that are typically hidden in forward-projected spreadsheets but become visible and testable when you work backward from required outcomes.Pro tipHighlight the three or four assumptions that most determine whether the venture succeeds or fails—these are your highest-priority testsWarningDo not confuse the reverse income statement with a real forecast—its purpose is to surface assumptions, not predict outcomes
- Rank Assumptions by Impact and UncertaintyList every assumption in your plan and rank each one on two dimensions: how much does the venture outcome depend on this assumption being correct, and how confident are you that it is correct. The assumptions that score highest on impact and lowest on confidence are your priority tests. These are the make-or-break unknowns that will determine success or failure—and they should be tested before you invest significant resources in execution.Pro tipBe brutally honest about confidence levels—the most dangerous assumptions are the ones the team has informally agreed are true without evidence
- Design Minimum-Cost Experiments to Test Key AssumptionsFor each high-priority assumption, design the cheapest possible experiment that would provide meaningful evidence for or against it. Can you test customer willingness to pay without building the full product? Can you validate market size through a small pilot in a test geography? Can you test the cost structure with a prototype rather than a factory? Each experiment should have a clear hypothesis, a minimum investment required, and predefined criteria for what constitutes confirming or disconfirming evidence.Pro tipThe best experiments test multiple assumptions simultaneously—a small pilot launch can validate demand, pricing, and cost structure togetherWarningAvoid experiments that can only confirm—design tests that could genuinely disconfirm your assumptions, or you are just seeking validation
- Stage Investment Around Learning MilestonesRelease additional funding and resources only when previous assumptions have been tested and validated. Create explicit checkpoints where the team presents what they have learned (not just what they have built) and the leadership decides whether to invest further, pivot, or exit. This staged approach limits downside exposure while preserving upside optionality. Each stage should answer a fundamental question: is there a real customer need? Can we serve it profitably? Can we scale the solution?Pro tipFrame checkpoints as learning reviews rather than performance reviews—the team should be rewarded for discovering that an assumption was wrong, not punished for itWarningPressure to show progress can cause teams to skip assumption testing and jump to execution—this defeats the entire purpose of discovery-driven planning
Honda entered the US motorcycle market with a plan to sell large motorcycles competing with Harley-Davidson and European brands. Their critical assumption—that American riders wanted large Japanese motorcycles—proved wrong immediately as dealers showed no interest. However, because Honda had not committed all resources to the original plan, they were able to pivot when they noticed that their small Super Cub bikes, which employees had been riding around Los Angeles, attracted unexpected interest from everyday consumers who would never enter a traditional motorcycle dealership. Honda shifted to selling small bikes through non-traditional channels like sporting goods stores.
Christensen advocated for Discovery-Driven Planning (originally developed by Rita Gunther McGrath and Ian MacMillan) after observing that the most common cause of failure in disruptive markets was not bad technology or insufficient resources but bad forecasting presented as fact. Companies would build detailed spreadsheets projecting revenues for markets that did not yet exist, and then commit massive resources based on these projections. When the projections proved wrong—as they inevitably did in new markets—the companies either doubled down on the wrong plan or abandoned the opportunity entirely. Neither response was appropriate. Christensen saw that what these companies needed was a planning methodology that assumed uncertainty rather than pretending it away, and that converted each dollar invested into learning rather than into premature scaling.