Project Background

About the Observatory

How this project was built, the methodology behind the data, and the team at H Heuristics.

01

The Project

Why This Observatory Exists

The Global Diffusion of Innovations Observatory was built to make the abstract theory of technology diffusion visible and interactive. Rogers' framework is elegant, but its power only becomes apparent when you can watch the S-curves unfold in real data, country by country, year by year.

We chose the Gapminder format — pioneered by Hans Rosling — because it reveals the relationship between economic development and technology adoption in a way that static charts cannot. The animation shows countries moving through development space, tracing trajectories that reveal convergence, divergence, and leapfrogging.

This observatory is part of H Heuristics' broader work on economic diffusion, convergence dynamics, and development economics. Our research programme includes the D-Coefficient framework for measuring institutional diffusion capacity, Convergence Maps for visualizing development trajectories, and the Carpathian Crescent simulation engine for exploring counterfactual development paths.

02

Methodology

Data & Modelling Approach

The dataset covers 100 countries across 64 years (1960–2023), generating 6,464 country-year observations. Three technology adoption indicators are tracked:

Internet Users (% of population) — from the first ARPANET connections in 1969 to near-universal access in developed nations
Mobile Cellular Subscriptions (per 100 people) — from the first commercial networks in the early 1980s to the smartphone era
Fixed Broadband Subscriptions (per 100 people) — from the DSL era (late 1990s) through fiber deployment

Modelling Approach

Technology adoption trajectories are modelled using the Rogers S-curve (logistic function), parameterized by country income group and documented adoption timelines:

High-income countries — Early adopters (internet: ~1990, mobile: ~1985), fast diffusion velocity, high adoption ceilings (internet: 97%, mobile: 145 per 100)
Upper middle-income — Mid-range adopters (internet: ~1995), moderate diffusion velocity
Lower middle-income — Later adopters (internet: ~2000), slower diffusion, lower ceilings
Low-income — Latest adopters (internet: ~2002), slowest diffusion, lowest ceilings (internet: ~55%)

Population data is based on UN World Population Prospects. GDP per capita (PPP, constant 2017 international dollars) follows World Bank trends. The model incorporates known diffusion anomalies — mobile leapfrogging in Africa, China's compressed adoption timeline, and the post-Soviet digital catch-up in Eastern Europe.

Technology Stack

The interactive visualizations are built with Observable Framework (v1.13.4) and D3.js (v7.9.0). The informational website uses the H Heuristics brand design system with Cormorant Garamond and Source Sans 3 typography. The entire site is served as a static site, deployable to any web host or GitHub Pages.

03

References

Core Literature

Rogers, E.M. (1962). Diffusion of Innovations. New York: Free Press. The foundational text — now in its fifth edition (2003) and the second-most-cited book in the social sciences.
Bass, F.M. (1969). "A New Product Growth Model for Consumer Durables." Management Science, 15(5), 215–227. The mathematical formalization of diffusion theory.
Moore, G.A. (1991). Crossing the Chasm. New York: HarperBusiness. Identified the critical gap between early adopters and the early majority.
Rosling, H. (2006). "The Best Stats You've Ever Seen." TED Talk. Popularized the animated bubble chart format used in this observatory.
World Bank. (2024). World Development Indicators. Primary source for country metadata and economic indicators.
ITU. (2024). World Telecommunication/ICT Indicators Database. Primary source for technology adoption timelines.

H Heuristics · Nottingham, UK · hheuristics.com

Explore the Observatory →