Curve Steepness and Adoption Velocity
Curve Steepness and Adoption Velocity
One-line summary: "Some S-curves are steeper than others" — the steepness parameter k (how fast a technology crosses from niche to mainstream) is set by feedback mechanisms (learning curves, economies of scale, network effects, lock-in), and modern curves are getting steeper.
The insight
Steepness is not incidental — it is the variable that decides how violent and how brief a re-rate is, and it is governed by identifiable feedback. RMI names four mechanisms that bend adoption into a steep S rather than a line: learning curves, economies of scale, technological reinforcement, and social diffusion. Where adoption raises the value or lowers the cost of further adoption (increasing returns / network effects, à la Arthur), you get positive feedback, tipping, and lock-in — but "there is no lock-in without further stabilizing returns," i.e. the steep self-reinforcing phase requires the returns mechanism to persist. In Bass terms, steepness is dominated by the imitation coefficient q (network/word-of-mouth) relative to the innovation coefficient p (see bass-diffusion-model).
A striking empirical regularity: time-to-50%-penetration has compressed across eras — telegraph 56 years, radio 22, PCs 16, internet 7, smartphones 5, AI tools ~3 (projected). Newer technologies ride steeper curves (higher k), so the window to take a position is shorter.
Evidence
- From 2026-06-16-autoresearch-tony-seba-technology-disruption-s-curves: RMI's four feedback mechanisms — "learning curves, economies of scale, technological reinforcement, and social diffusion" — "create the characteristic S-shape rather than linear growth."
- From 2026-06-16-autoresearch-tony-seba-technology-disruption-s-curves: accelerating time-to-50% — "Telegraph: 56 years … PCs: 16 … Internet: 7 … Smartphones: 5 … AI tools: 3 years projected"; AI fit with Bass p≈0.01, q≈0.8 (very high network effect) (Dr Li).
- From 2026-06-16-academic-research-technology-adoption-s-curves: network effects / increasing returns drive steepness and lock-in; "no lock-in without further stabilizing returns" (Roedenbeck 2007); lock-in mechanisms enumerated for energy/transport transitions (Klitkou et al. 2015).
Design implications
- The presence and durability of feedback mechanisms (learning rate, scale, network effects) is the leading indicator of a steep curve and of a defensible eventual share — exactly what an investor wants to gauge before a re-rate.
- Steepness compounds urgency: RMI — "the fastest movers will stand to capture the biggest rewards, and the slowest movers will be left with the biggest losses."
Contradictions / tensions
- High measured steepness in cost does not guarantee steep adoption timing — autonomy/robotaxi had a steep narrative but slow real adoption (see adoption-underestimated-or-overhyped and tony-seba's track record).