S-Curve Analysis

Origin

The method was popularised by Richard N. Foster (Innovation: The Attacker's Advantage, Summit Books, 1986). Foster, then a director at McKinsey, showed in case studies ranging from sailing ships to semiconductors that technologies repeatedly follow an S-shaped path — and that incumbents regularly miss the switch to the next curve, while attackers start on it unencumbered. The advantage that gives the book its name lies precisely in not having invested in the old curve.

Typical use

S-curve analysis belongs in technology and R&D planning: when deciding whether further development budgets should flow into a mature technology, when assessing successor technologies, and when timing platform transitions. A typical signal of maturity is diminishing returns on development — each additional percent of performance costs disproportionately more effort, while young rival technologies make large leaps on small budgets.

Procedure

1. Define the performance parameter and its limit

   First, fix the performance parameter that customers actually reward — storage density, efficiency, cost per unit — and estimate the natural limit that physics, materials or economics impose on this technology.

2. Locate the position on the curve

   Plot historical performance data against cumulative engineering effort, not against time. Only this yardstick reveals whether the technology is still in the steep part of the curve or already approaching its limit.

3. Identify successor curves

   Look for technologies that deliver the same customer benefit on a different working principle. Their curves typically start below the established performance level but promise a higher limit or a steeper rate of improvement.

4. Derive the switching strategy

   The comparison of curves yields the allocation decision: keep optimising, develop in parallel, or shift resources. Since the transition rarely succeeds abruptly, a dual strategy with defined exit criteria is usually appropriate.

Limits and typical mistakes

The analysis stands or falls with its data: cumulative engineering effort is laborious to reconstruct, and a technology's natural limit can often be determined precisely only in hindsight. The most common technical mistake is plotting against time instead of effort — it confuses investment cycles with technological maturity. Equally widespread is a deterministic misreading: not every mature technology gets replaced, and not every young curve delivers what its initial slope promises. Finally, the S-curve says nothing about whether customers can still absorb the performance gains — that question requires the market perspective, not the technology curve.

Relation to the Innovator's Dilemma