273. Bass Diffusion

A simple SD model of new-product adoption. Frank Bass (1969). Predicts the characteristic S-shaped adoption curve from two parameters.

273.1. Model

Let $𝐴\left(𝑡\right)$ = cumulative adopters, $𝑁$ = total market size. Adoption rate:

Two parameters:

• $𝑝$: coefficient of innovation — adoption driven by external influence (advertising, news, intrinsic appeal)
• $𝑞$: coefficient of imitation — adoption driven by word-of-mouth (existing adopters influencing potential adopters)

273.2. Structure

Two interpreting terms:

• $𝑝\left(𝑁-𝐴\right)$: innovation — fraction of remaining non-adopters who’d adopt based on external info
• $𝑞\left(\frac{𝐴}{𝑁}\right)\left(𝑁-𝐴\right)$: imitation — non-adopters influenced by the proportion $\frac{𝐴}{𝑁}$ of current adopters; total imitation pressure proportional to remaining market

Together: a limits-to-growth archetype with reinforcing word-of-mouth and balancing market exhaustion.

273.3. Behavior

Early on ($𝐴\ll 𝑁$): adoption is mostly $𝑝𝑁$ — small, driven by innovation only.

Middle: $𝐴$ approaches $\frac{𝑁}{2}$; imitation term peaks, growing very fast.

Late: $𝑁-𝐴$ shrinks; growth slows.

Peak adoption rate at:

Time of peak adoption:

For typical $𝑝\approx 0.01-0.05$, $𝑞\approx 0.3-0.5$: peak around 5-10 years after launch.

273.4. Empirical fit

Bass fit the model to historical sales of color TVs, washing machines, and dozens of other durables. Excellent S-curve fits with just $𝑝,𝑞$ as parameters.

Common parameter values:

• Consumer electronics: $𝑝\approx 0.03$, $𝑞\approx 0.4$
• Pharmaceuticals: $𝑝\approx 0.001$, $𝑞\approx 0.5$
• Software platforms: $𝑝\approx 0.01-0.05$, $𝑞\approx 0.5-0.7$

273.5. Generalizations

• Generalized Bass: include effect of price, advertising → $𝑝\left(𝑡\right),𝑞\left(𝑡\right)$ time-varying
• Repeat purchase: cumulative installed base $\ne$ cumulative purchases
• Substitute / next-generation products: previous-generation forms downstream pool

273.6. Why it matters

The Bass model is the default for:

• New-product launch forecasting
• Marketing budget allocation across product lifecycle stages
• Technology adoption studies (Rogers’ Diffusion of Innovations)

Most consumer-electronics product launches use Bass as the baseline forecast.

273.7. See also

• Logistic Growth — similar S-curve, simpler model
• Feedback Loops — R + B = S-curve archetype
• System Dynamics overview