Skip to content

Sustaining vs. Disruptive Innovation

Definition

A sustaining technology improves the performance of an established product along the dimensions that mainstream customers have always cared about — it can even be radical or technologically difficult, and incumbents still tend to win at it. A disruptive technology instead brings "a very different value proposition than had been available previously": it underperforms on the traditional metric but is cheaper, simpler, smaller, or more convenient, which appeals first only to a fringe of new or undemanding customers. Christensen is explicit that this is not the same split as "incremental versus radical" — sustaining technologies can be enormously difficult, and disruptive ones can be technologically trivial.

In the Book

The book's founding case is the disk drive industry's succession of smaller drive formats — 14-inch, then 8-inch, then 5.25-inch, then 3.5-inch, then 1.8-inch. Each new, smaller architecture was rejected by the leading manufacturers' existing customers (mainframe makers wanted more capacity at lower cost per megabyte, not a smaller drive) and so was commercialized instead by entrant firms selling into new applications — minicomputers, then desktop PCs. In every one of these six transitions but two, the industry's dominant firm lost its leadership position, even though, as the 8-inch case shows, the incumbents were technologically fully capable of building the new drives once they finally tried: "the established firms... were held captive by customers." Christensen contrasts this with thin-film read/write head technology, an extremely difficult and expensive sustaining innovation that incumbents nonetheless led without exception, because their mainstream customers wanted it.

Why It Matters

The distinction explains a pattern of failure that "incumbent inertia" or "bad management" cannot: companies fail not despite good management but because of it, when the innovation in question scores worse on the metric their best customers currently pay for. It gives a diagnostic question that generalizes past technology products — is this change something my most important constituents are asking for and will pay more for, or does it look worse by their current standards while opening a new value proposition for someone else — that applies to any domain where an incumbent's success metric can blind it to a rising alternative measured on different terms.