Igor Ansoff's 1957 Harvard Business Review article, "Strategies for Diversification," gave corporate growth a vocabulary that has lasted. Growth comes from selling an existing or new product into an existing or new market. That gives four cells: market penetration, market development, product development, and diversification. They carry a rising risk gradient, because each step away from "existing" takes away a body of knowledge the firm can rely on.
How does risk rise across the four cells?
The driver logic is one claim: unfamiliarity compounds, and compounding unfamiliarity compounds the risk of failure. Penetration draws on everything the firm already knows. Market or product development takes away certainty about one side and keeps the other. Diversification takes away both at once, and the risks don't simply add. A firm with no track record on either side has nothing to check its assumptions against. So it's no surprise that diversification's empirical track record is consistently the weakest of the four cells. Widely cited studies of M&A and diversification performance repeatedly put the share of unrelated moves that fail to create value at roughly two-thirds to well over three-quarters. That's stable enough to work as a standing prior whenever "diversification" appears on a slide, and Worked Example 2 puts real numbers on why.
Advertisement
How do you put numbers on each cell?
Building it correctly means going beyond the label. A quadrant name tells you the category of risk and says nothing about its magnitude. Mixing the two up is the commonest misuse of the tool. "Market development" covers everything from opening a store in the next city over, which is low risk, to entering a country with an unfamiliar regulatory regime and no local partner, which is very high risk. The fix is to attach a number to each cell. Estimate the incremental revenue potential and assign a realistic probability of success. Multiply the two for an expected value, then divide by the capital required to get a capital-efficiency ratio. Now Ansoff is a decision tool, where before it was a vocabulary lesson. It routinely reverses the naive ranking, too. The option with the biggest headline revenue number is often the worst once you price in probability and capital.
What are the blind spots of the Ansoff matrix?
Four mechanism-level blind spots follow. First, the matrix names the risk category but not its magnitude. Anyone who stops at the label will treat two "market development" moves with wildly different real risk as interchangeable. Second, it says nothing about capability, meaning whether the firm actually has the assets, brand, or skill to execute in the new cell. A cell can look attractive on paper and still be a "good market, no right to win" trap. The market-entry case format tests that same trap directly. Third, it typically ignores cannibalization. "New product, existing market" revenue often isn't additive, because some of it comes from customers who'd otherwise have bought the firm's existing product. Fourth, the matrix is a classification device and not an execution plan. It names the quadrant and never tells you how to execute inside it. That's why diversification's poor record persists despite nearly seventy years in the textbooks. Naming a risk category isn't the same as respecting it.