Innovation-adoption diffusion model
Bass Diffusion Model
New adoption can combine an external innovation effect with an internal imitation effect, producing the familiar rise, peak, and decline of a product-adoption wave.
f(t) / (1 - F(t)) = p + qF(t)
F(t) is cumulative adoption as a fraction of the eventual market, f(t) is the new-adoption rate, p is the innovation coefficient, and q is the imitation coefficient. The right side is the adoption hazard among remaining non-adopters.
The wave instrument begins at p = 0.03 and q = 0.38, then lets both coefficients vary. Actual time units, market potential, and coefficients must be estimated for a defined market and product.
(%/period)
The same parameters drive the curves, peak marker, hazard readout, and population wave. The particles are a proportional teaching display, not individual-level data.
- CHANGE
- Cumulative adoption
- WATCH
- new-adopter hazard
- MEANING
- The wave instrument begins at p = 0.03 and q = 0.38, then lets both coefficients vary. Actual time units, market potential, and coefficients must be estimated for a defined market and product.
Innovation starts the wave; imitation steepens it.
Cumulative adoption follows an S-shaped path while new adopters form a single wave. The position of the peak depends on p, q, market potential, and the model assumptions.
What it actually says
The Bass model describes first purchases by separating two routes into adoption. Innovators adopt independently of how many others have already adopted, represented by p. Imitators become more likely to adopt as cumulative adoption F(t) increases, represented by qF(t).
Multiplying the hazard by the remaining market produces the new-adoption curve. Early on, few previous adopters exist; later, imitation accelerates diffusion; eventually, few potential adopters remain and new adoption declines. The model can fit useful aggregate patterns without identifying the actual communication mechanism.
"A useful law compresses a pattern. It does not erase the conditions that make the pattern true."
How the idea developed
The modern form emerged through observation, argument, and later refinement. The timeline separates the first insight from the version now used in textbooks and practice.[1]
Everett Rogers synthesizes research on diffusion of innovations and adopter categories.
Frank Bass publishes the new-product growth model and tests it on eleven consumer durables.
Extensions introduce price, advertising, replacement, competition, and international diffusion.
Diffusion models support launch planning, scenario analysis, technology transitions, and comparisons across markets.
How the pattern works
The relation becomes useful only when its mechanism, measurement process, and operating range are visible.
A baseline hazard p allows adoption without prior adopters, often associated with external information.
The qF(t) term raises adoption propensity as the installed base grows.
Only the remaining fraction 1 - F(t) can become first-time adopters.
A fitted p and q summarize the curve; they do not directly count innovators, word-of-mouth contacts, or causal exposures.
F(t) is cumulative adoption as a fraction of the eventual market, f(t) is the new-adoption rate, p is the innovation coefficient, and q is the imitation coefficient. The right side is the adoption hazard among remaining non-adopters.
Where it earns its keep
Applications are strongest when the law changes a decision, measurement, model, or experiment rather than merely providing an analogy.
Build adoption scenarios for a launch
ApplicationA calibrated curve can estimate peak timing, cumulative penetration, and demand under a stable market definition.
Report uncertainty and compare against simple benchmarks; early data often identify parameters poorly.
Compare diffusion across markets
ApplicationCoefficients and peak timing can organize differences in infrastructure, compatibility, regulation, and social influence.
Cross-market coefficient comparisons require consistent units, definitions, and observation windows.
Separate exposure from remaining opportunity
ApplicationThe innovation-imitation structure clarifies why spread can accelerate and then slow.
Ideas, behaviors, and memes may involve repeated adoption, forgetting, mutation, and network structure outside the classic model.
Where it stops working
The basic model assumes a fixed eventual market, one adoption event per unit, homogeneous mixing, constant p and q, no supply constraint, no explicit price or competition, and reliable aggregate sales. These assumptions often fail during long product cycles or changing market definitions.
Good in-sample fit does not identify word of mouth as the cause. Logistic, Gompertz, epidemic, marketing-response, and heterogeneous-adopter models can produce similar curves. Forecasts from the earliest observations are especially unstable because peak position and market potential are weakly constrained.
"p is the percentage of people who are innovators"
Better: It is a hazard coefficient, not an adopter-category share."q directly measures word of mouth"
Better: It summarizes endogenous-looking acceleration and does not identify a channel by itself."An S-curve validates the Bass mechanism"
Better: Several mechanisms and functional forms can create S-shaped adoption."The market potential is known and fixed"
Better: It is often estimated and can change with price, entrants, population, or product definition.Sources and further reading
Original publications and serious secondary scholarship are prioritized over summaries.
- Bass - A New Product Growth for Model Consumer DurablesThe original 1969 model and tests on consumer-durable sales; the published title contains a noted word-order error.https://doi.org/10.1287/mnsc.15.5.215
- Bass - Comments on A New Product Growth for Model Consumer DurablesRetrospective clarification of the model, its title, development, and extensions.https://doi.org/10.1287/mnsc.1040.0300
- Mahajan, Muller, and Bass - New Product Diffusion Models in MarketingMajor review of estimation, extensions, and marketing applications.https://doi.org/10.1177/002224299005400101
- Rogers - Diffusion of InnovationsPublisher record for the broader framework covering innovations, communication, time, and social systems.https://www.simonandschuster.com/books/Diffusion-of-Innovations-5th-Edition/Everett-M-Rogers/9780743222099