Diffusion of Innovations

Deciding who to roll out to first, and who to stop arguing with

Score the five attributes and list the people others actually go to. It ranks your Early Adopters, names the spread blockers, and sequences the waves — including the check on whether you are being pushed to start with the sceptic.

Work out who first →

Everett Rogers. Answers a question every rollout faces: who first?

The five adopter groups

Group	~%	What they need	How to use them
Innovators	2.5%	Novelty, early access	Useful for testing, poor advocates — peers see them as unrepresentative enthusiasts
Early Adopters	13.5%	To be seen as ahead, and respected	🎯 Your target. Opinion leaders. Everything hinges on them
Early Majority	34%	Proof it works here, with people like them	Won’t move until Early Adopters vouch
Late Majority	34%	Norms, and the old option becoming inconvenient	Move when it’s clearly the majority behaviour
Laggards	16%	Necessity	Do not spend your energy here

The rule most rollouts break

Do not start with the loudest sceptic.

There is a persistent instinct — often from a well-meaning exec — to “win over” the most vocal critic first, on the theory that if they convert, everyone follows. In practice you spend disproportionate effort, usually fail, and hand the sceptic a public platform during the fragile early phase.

Start with the Early Adopters. Build visible local proof. The sceptic is far more movable when 60% of their peers have already moved and the argument has been settled without them.

⚠️ One exception: if the sceptic is technically right about a design flaw, that’s not resistance, that’s free consultancy. See Resistance Decoder.

Identifying your Early Adopters

Not the org chart. Look for:

People others ask before deciding — watch who gets copied into things
Respected for competence, not just seniority
Already frustrated with the current way (frustration is fuel)
Socially connected across teams, not just within one
Credible to the sceptics — a bit critical themselves

The two-question sociogram

Ask 15 people: “Who do you go to when you’re not sure how to do something here?” and “Whose opinion on this change would you actually take seriously?” Tally the names. The top 5–8 are your network, and they are frequently not who leadership expected. Bring that list to the sponsor — it’s usually a genuinely surprising and useful piece of insight.

What makes an innovation spread (Rogers’ five attributes)

Design-test your change against these:

Relative advantage — is it visibly better for the user, not just the organisation?
Compatibility — does it fit existing values and ways of working?
Complexity — how hard is it to understand and use? (Lower is faster)
Trialability — can people try it without commitment or risk?
Observability — can others see the results?

Score your change 1–5 on each. Anything ≤2 is a spread blocker you can design around before launch — usually more cheaply than any comms campaign.

Score your change

Relative advantage: ___ Compatibility: ___ Complexity (5 = simple): ___ Trialability: ___ Observability: ___

Weakest attribute: _______________ What we’ll do about it: _______________

Observability is the most commonly neglected. If nobody can see who is doing the new thing and what they’re getting from it, diffusion stalls regardless of how good it is.

Sequencing a wave plan

Wave	Who	Aim
0 — Pilot	1 team, high early-adopter density, sympathetic manager	Learn and fix. Expect to change the design.
1	Early adopters across sites	Generate local proof stories
2	Early majority, region by region	Momentum; publish adoption trends
3	Late majority	Norms + old route made inconvenient
4	Stragglers	Manager-led, individual, quiet

Never announce a wave plan that implies Wave 4 people are laggards. Label waves by geography or function, never by attitude.

Links

→ Champion Network Playbook → Social Norms and Influence → Pilot Design and Experiment Card

← Back to the Behavioural Change Toolkit

Last updated: 11 August 2026