Exploring associations between micro-level models of innovation diffusion and emerging macro-level adoption patterns

A micro-level agent-based model of innovation diffusion was developed that explicitly combines (a) an individual’s perception of the advantages or relative utility derived from adoption, and (b) social influence from members of the individual’s social network. The micro-model was used to simulate macro-level diffusion patterns emerging from different configurations of micro-model parameters. Micro-level simulation results matched very closely the adoption patterns predicted by the widely-used Bass macro-level model (Bass, 1969). For a portion of the domain, results from micro-simulations were consistent with aggregate-level adoption patterns reported in the literature. Induced Bass macro-level parameters and responded to changes in micro-parameters: (1) increased with the number of innovators and with the rate at which innovators are introduced; (2) increased with the probability of rewiring in small-world networks, as the characteristic path length decreases; and (3) an increase in the overall perceived utility of an innovation caused a corresponding increase in induced and values. Understanding micro to macro linkages can inform the design and assessment of marketing interventions on micro-variables - or processes related to them - to enhance adoption of future products or technologies.

💡 Research Summary

The paper presents a novel agent‑based model (ABM) that integrates two fundamental drivers of innovation diffusion at the individual level: (1) the perceived relative utility of adopting the innovation and (2) social influence exerted by an individual’s network contacts. Each agent is assigned a utility parameter (U) reflecting the intrinsic benefit it derives from the new product. Adoption occurs either because the utility exceeds a personal threshold (the “innovator” effect) or because a sufficient proportion of an agent’s neighbors have already adopted, weighted by a social influence coefficient (the “imitator” effect). The social structure is modeled using a Watts‑Strogatz small‑world network, where the rewiring probability (p_rewire) controls the average path length and clustering, thereby shaping the speed and reach of information flow.

To evaluate whether the micro‑level dynamics can reproduce macro‑level diffusion patterns, the authors run extensive simulations across a wide parameter space: varying the number of initial innovators (N_i), the rate at which innovators are introduced (λ_i), the average utility level (μ_U), and the network rewiring probability. For each simulation, the cumulative adoption curve f(t) for the whole population is recorded and fitted to the classic Bass model

 f(t) = (1 – e^{-(p+q)t}) / (1 + (q/p) e^{-(p+q)t})

using least‑squares estimation. This yields induced Bass parameters p̂ (coefficient of innovation) and q̂ (coefficient of imitation) that can be directly compared with the underlying micro‑parameters.

Key findings are threefold. First, both the absolute number of innovators and the speed of their introduction positively affect the induced p̂. Faster seeding of innovators creates a steeper early rise in adoption, mirroring the Bass interpretation of p as the external influence term. Second, the network rewiring probability has a strong, monotonic effect on q̂ while leaving p̂ relatively unchanged. Higher p_rewire shortens the characteristic path length, facilitating rapid peer‑to‑peer transmission and thus amplifying the imitation component captured by q. Third, raising the overall perceived utility of the innovation simultaneously increases both p̂ and q̂. Higher utility strengthens the intrinsic motivation to adopt (raising p̂) and, because more agents become adopters early, it also boosts the social contagion effect (raising q̂).

These results demonstrate a quantitative bridge between micro‑level behavioral rules and the macro‑level Bass parameters that have long been treated as exogenous. Practically, the study suggests concrete levers for marketers and policymakers: (a) increase the pool and rollout speed of early adopters to lift p̂; (b) engineer or exploit network structures that reduce average distances—e.g., through influencer campaigns, viral referral programs, or platform design that encourages cross‑community links—to raise q̂; and (c) enhance the product’s perceived utility through design improvements, clear value communication, or complementary services, thereby improving both p̂ and q̂.

The authors acknowledge several limitations. The model assumes binary adoption (adopt/not adopt) and static utility and influence parameters, ignoring learning, experience effects, and temporal changes in network topology. Heterogeneity in agent attributes (income, risk tolerance, etc.) is also absent, which may limit external validity. Future work is proposed to incorporate multi‑stage adoption processes, dynamic utility evolution, and empirical validation using real social‑media network data.

In summary, the paper successfully links micro‑level diffusion mechanisms to the well‑established Bass macro‑model, offering a richer theoretical foundation for designing and evaluating diffusion‑oriented interventions. By revealing how specific micro‑variables map onto Bass coefficients, it equips practitioners with actionable insights for accelerating the uptake of new products and technologies.