Rollouts begin in the most willing or least powerful units and reach the strongest resistant division last, when internal results, trained champions, and proven methods leave it little cover for holding out.
An example
Nine sites were scheduled for a new warehouse management system at a third-party logistics company, and the program director put the national distribution center last. It was the largest site, run by a general manager with the chief executive's ear, and it would fight the change. The first three sites were smaller regional warehouses whose managers had asked for better inventory accuracy, and one had a product mix close to the national center's. By month ten those three had cut picking errors by half, the integration with the carrier platform had been debugged twice over, and twenty trained floor supervisors could be lent out. When the national center's turn came, the board deck showed six peer sites live, the general manager's regional peers were quoting the results, and the cutover went ahead on schedule with supervisors from the regional sites on the floor.
Why it works
By the time the resistant unit is engaged, internal case material, peer testimonials, and published metrics make saying yes look better. Earlier waves lower the cost of adopting: integrations are tested, training exists, and champions can be borrowed. Holding out loses its main cover, since the claim that the change is unproven has been answered by the company's own results, and what remains looks like protectionism. Being the last holdout in front of the chief executive, with peers cited as successes, carries a political cost. It requires discretion over sequencing, early sites capable of results that transfer to harder cases, and enough leadership cover to survive until the hard phase.
The thinking behind it
Mark Granovetter Granovetter's Threshold Models of Collective Behavior (1978) holds that each person joins once enough others have acted, so small changes in who goes first can produce very different outcomes. Starting with willing sites lets the early waves clear low thresholds, and each success lowers the bar for the next site. By the time the resistant unit's turn comes, most of its peers have moved.
Robert B. Cialdini In Influence (1984), Cialdini describes social proof: when unsure, people look to what others like them are doing. Peer sites live on the new system, with the resistant general manager's own regional peers quoting the results, are social proof of the most persuasive kind. That is why the clean form includes an early site that resembles the resistant unit, so the proof comes from people like them.
Solomon E. Asch Asch's Studies of Independence and Conformity: I. A Minority of One Against a Unanimous Majority (1956) found that people facing a unanimous majority went along on about a third of responses, though most responses stayed correct and about a quarter of subjects never conformed. A resistant unit facing six live peer sites is close to that position. The finding also cautions against expecting surrender, since most holdouts keep their own judgment and the case still has to make itself.
When it fits
Control, or strong influence, over the order in which units adopt a change is the precondition; without it the tactic is unavailable. It fits when one or more powerful units are likely to resist. You need early sites that are willing and comparable enough to the resistant unit that their results transfer, and enough backing from top leadership to keep the program alive through the easy phase.
How to use it well
Before spending energy on messaging, confirm that you control the sequence.
- Order the early waves by willingness and readiness, and include at least one site that resembles the resistant unit, so its results cannot be dismissed as apples to oranges.
- Invest in making the early waves succeed properly, and document what was learned.
- Build a bench of trained champions from the early sites who can help the later ones.
- Keep the resistant unit informed throughout, invite its people to see the early sites, and give it a real say in how its own cutover is planned.
When its turn comes, present peer results, offer support, and let the case make itself.
What it costs and where it backfires
Sequencing takes time, and the value at the strongest unit is delayed until last. A resistant leader who sees the plan may act early, undermining the first waves from a distance or lobbying to stop the program before it arrives. Early sites that were cherry-picked or over-resourced produce results the holdout will reject, and fairly. The clean form uses real, comparable results. Once the early waves are inflated, or the resistant unit is deliberately kept uninformed to isolate it, the rollout becomes a pressure campaign and invites the counterattack.
When it is used on you
If your unit is scheduled last, test whether the early sites resemble yours, since results from hand-picked units do not transfer. Ask for the sequencing criteria in writing. Engage before your turn comes: visit the early sites, raise your specific requirements while the design can still change, and take the help on offer if the results hold up.