brain/
questionopentechnology-adoption-s-curves

Can the S-curve inflection be detected in real time?

Notes

Can the S-curve inflection be detected in real time?

The question

Prospectively — not in hindsight — can you tell that a technology is at the knee (penetration accelerating through the early-majority crossing), as opposed to still pre-tipping or already past saturation? This is the single most valuable signal for capturing remaining runway.

Why it matters

The entire investing value of the S-curve lens (see s-curve-position-in-stock-evaluation) depends on locating the present moment on the curve. Buying at the inflection captures the steep part; buying near saturation captures little.

What we currently believe

The ceiling L and inflection x₀ are the least forecastable parameters: Bass M/p/q are badly identified and sensitive to assumptions, real curves are asymmetric, and learning rates change. Cost curves are far more forecastable than adoption timing (see forecastability-of-technological-progress). So real-time inflection detection is hard and should be expressed as a probability/range, not a point.

Evidence we have

Evidence we need

  • A reliable leading indicator of the knee (accelerating second derivative of penetration, cost crossing a parity threshold, q-coefficient inflection) tested against history.
  • Back-tests of "we were at the inflection" calls vs. realized outcomes.

How to resolve

  • Fit logistic/Bass models to historical adoption series and test out-of-sample inflection-timing error.
  • Cross-reference cost-parity dates (forecastable) as a proxy trigger for the adoption knee.

Related

Referenced by