Diffusion of innovation

“New” can promise improvement and status, but it can also signal risk, effort and disappointment. People therefore adopt the same innovation at different times and for different reasons. Diffusion theory explains how an idea, product or practice spreads through a population and why launch tactics must change as adoption moves from adventurous users to the wider market.

When to use it

• Use it to plan the launch and growth of new products or services.
• Use it to identify opinion leaders and early users who can help credible adoption spread.

Origins

Diffusion research began in the late 19th century across anthropology, geography and sociology. Everett Rogers studied the adoption of a new weed spray among Iowa farmers for his 1957 doctoral research. He synthesised evidence from many fields in the first edition of Diffusion of Innovations in 1962 and continued refining the theory; the fifth edition, published in 2003, addressed the internet and changing communication networks.

What it is

The two-step process

Two-step flow theory proposes that mass communication often influences opinion leaders first, and those people then interpret and transmit the message through their social relationships. Their judgement can materially affect adoption: reviewers influence car buyers and theatre audiences, while respected specialists shape choices inside organisations. The pattern is not universal, but it reminds launch teams that trusted interpersonal influence differs from broad awareness.

The trickle-down effect

Some innovations enter at a high price and initially remain accessible only to wealthy or privileged users. Ownership can create status and visible proof; as production improves and price falls, the offer becomes accessible to the mass market. Early mobile phones illustrate the pattern: expensive, bulky devices used by a narrow elite eventually became everyday tools.

The diffusion of innovations

Rogers described five adopter categories, defined by when people adopt relative to others in the same social system:

• Innovators, about 2 per cent, actively seek novelty and tolerate high uncertainty.

• Early adopters, about 13 per cent, are respected users whose judgement often influences others.

• Early majority, about 34 per cent, adopt before the average person but usually want practical evidence.

• Late majority, about 34 per cent, remain sceptical and often adopt after economic or social pressure increases.

• Laggards, about 16 per cent, adopt last and may rely strongly on established practice.

These are descriptive positions in one adoption distribution, not permanent personality labels.

Crossing the chasm

Geoffrey Moore adapted diffusion thinking for technology markets. He argued that early enthusiasts and visionaries value novelty and strategic possibility, while the early majority wants a reliable, complete solution supported by credible references. The “chasm” is the go-to-market discontinuity between those groups. Crossing it often requires a focused beachhead segment, a complete use case and proof from peers—not merely more promotion to early enthusiasts.

Technology acceptance model

Technology acceptance research emphasises perceived usefulness and perceived ease of use. A person is more likely to adopt when the technology appears to improve performance without disproportionate effort. Awareness, interest, desire and action still matter, but adoption also depends on evidence of relative advantage, confidence, compatibility and manageable learning cost.

The diffusion of innovations

Rogers represented adoption timing as a bell-shaped distribution and cumulative adoption as an S-curve. The categories are uneven because very early adopters can be meaningfully distinguished, whereas the later tail is more homogeneous. Many innovations never reach a self-sustaining level of social proof. A frequently cited threshold lies near the transition from early adopters to the early majority, when approximately 16 per cent of the population has adopted, but it is not a universal law.

Diffusion speed depends on the innovation and its system. Toys and digital products can spread rapidly when they are observable, easy to try and supported by networks. Industrial materials can take far longer. Carbon fibre was invented in the nineteen fifties but remained concentrated in sport and specialised uses for nearly 30 years before aerospace adoption expanded. Graphene, isolated in 2008, still requires application proof and scalable, affordable production before broad diffusion.

Developments of the model

The framework has been applied far beyond agricultural pesticides to software, electronics, public health and organisational practice. Later work emphasises that initial adoption is not the same as continued use: innovators may also be the first to abandon one novelty for another.

Diffusion describes a population pattern; it does not by itself supply the motivation or ecosystem required for adoption. Technically superior keyboard layouts have not displaced QWERTY at scale, and the logic of a shared international language has not made Esperanto widespread. Installed habits, complements, standards and network effects can outweigh product advantage.

How to use it

First decide whether the offer is a meaningful innovation for the customer, not merely old content in new packaging. Then estimate where target customers sit on the adoption spectrum and identify the evidence, risk reduction and support each group needs.

A European paper merchant applied the model while testing service innovations with printers. Proposed offers included rapid or night-time delivery, timed delivery, consignment stock, additional products and advisory services. The survey also asked respondents to choose a neutral statement describing their approach to new technology:

• New technology is important, and we would be among the first to use a new product or service.

• We may not be first, but we adopt before most others.

• We wait until early problems have been resolved.

• We are in no hurry to deploy the latest technology.

• We wait as long as possible and change only when the existing infrastructure is redundant or no alternative remains.

The observed distribution resembled the categories Rogers described in his 1962 work. More importantly, the results connected adoption readiness with company type and size, allowing the merchant to target service innovations where they were most likely to be valued.

Some things to think about

• Segment customers by adoption readiness when that dimension changes the proposition, proof or launch sequence.

• Expect radical innovation in business markets to require sustained evidence and ecosystem change.

• Seek credible opinion leaders, but do not assume that popularity among enthusiasts proves mainstream fit.

Top practical tip

Match the launch to the next adopter group. Early users may respond to novelty and possibility; the mainstream usually needs a complete solution, trusted references, low switching risk and clear economic value.

Top pitfall

Do not treat trial as sustained adoption or adopter categories as fixed identities. Measure continued use, context and switching behaviour, and revise the diffusion plan when the innovation or social system changes.

Further reading

• Rogers, E.M. (two thousand and three). Diffusion of Innovations. Free Press.
• Moore, G.A. (nineteen ninety-one). Crossing the Chasm. HarperBusiness.