You announce the new system, the new process, the new structure. The slides land well. Then nothing moves, or worse, people nod in the room and quietly carry on as before. The change was approved; it just wasn't adopted. Resistance is what stands in that gap, and the instinct to bulldoze through it is exactly what makes it harder.

The quick version

  • Resistance is the friction people put up against a change, slowing it, diluting it, or quietly ignoring it. Most of it is rational from where the person stands: they see a cost, a risk, or a loss you haven't addressed.
  • Adoption is the goal: people actually using the new way and sticking with it after the launch noise fades. A change isn't done when it ships, it's done when it's the default.
  • The reliable move is to treat resistance as information. It tells you what you missed, missing facts, a real downside, or plain fatigue, and points to which response will work.
  • The trap is forcing compliance. You can mandate behaviour; you can't mandate belief. Coerced change reverts the moment attention moves elsewhere.

The idea in depth: resistance is a signal, not a wall

The most useful starting point is also the oldest. In 1948, Lester Coch and John R. P. French published "Overcoming Resistance to Change" in Human Relations, reporting a field experiment at the Harwood pyjama factory. When management imposed changes to production methods, output dropped and grievances rose; when workers were involved in designing the changes, through their representatives, or as a whole group, productivity recovered fast and resistance largely disappeared. Same change, different process, opposite result. The finding, built on Kurt Lewin's force-field thinking, was that participation reduces resistance because it removes the thing people were actually reacting to: being done to.

So the first move is not a better launch deck. It is to involve the people affected early enough that the change carries their fingerprints. That doesn't mean design by committee, it means the people who will live with the change get a real say in how it works before it's locked. A team that helped shape the new process defends it; a team handed the finished article looks for the exit.

An honest limitation. Coch and French studied a single factory in the 1940s, and later researchers have questioned how cleanly its results generalise, the groups weren't perfectly comparable, and "participation" can be done as theatre, which fools no one. Treat it as a strong, much-replicated direction rather than a guarantee: involvement usually lowers resistance, but only when it's genuine and when people have real influence over something that matters to them.

Six ways to handle resistance, and when each one fits

Involvement is powerful but slow, and not every situation allows it. The most practical map of the alternatives comes from John Kotter and Leonard Schlesinger's "Choosing Strategies for Change" (Harvard Business Review, 1979; reprinted 2008). They start by noting that people resist for different reasons, they fear losing something they value, they misunderstand the change, they genuinely think it's a bad idea, or they simply have a low tolerance for upheaval, and that the right response depends on which reason you're facing. They lay out six approaches, from lightest touch to heaviest.

flowchart TD
  A(["Why is this person
resisting?"]) --> B(["Missing or wrong
information"])
  A --> C(["Loss of something
they value"])
  A --> D(["Fear, anxiety,
adjustment strain"])
  A --> E(["Speed essential,
no time to persuade"])
  B --> F(["Educate &
communicate"])
  C --> G(["Involve them, or
negotiate a trade"])
  D --> H(["Facilitate &
support"])
  E --> I(["Co-opt or compel
last resort"])
Kotter & Schlesinger's logic: match the response to the reason behind the resistance, using the lightest touch that works. Leaders Loop

The six, in order of escalation: education and communication (for resistance caused by missing or wrong information); participation and involvement (when you don't have all the answers and others have power to block you); facilitation and support (when people are struggling with the adjustment itself); negotiation and agreement (when someone will clearly lose out and can resist hard); manipulation and co-option (a fast, ethically risky shortcut for when nothing else is feasible); and explicit and implicit coercion (when speed is essential and you hold the authority, a genuine last resort). The discipline is diagnostic. Name the reason first. Then reach for the lightest approach that fits, and escalate only when you have to.

You can mandate behaviour. You cannot mandate belief, which is why coerced change reverts the moment you stop watching.

An honest limitation. Kotter and Schlesinger's framework is a clear thinking tool, not tested law, and the later approaches carry real ethical weight. "Manipulation" and "coercion" can buy short-term movement and poison long-term trust; lean on them and you'll spend the credibility you need for the next change. Use the model to reach for the lightest lever, not to justify the heaviest.

Adoption is individual: the ADKAR sequence

Handling resistance gets a change started. Making it stick is a different problem, and the most-used practitioner lens here is the ADKAR model from Prosci's Jeff Hiatt, drawn from studying change patterns across hundreds of organisations. Its premise is simple and easy to forget: organisations don't change, people do. For any individual to adopt a change they have to clear five steps in order, Awareness of why the change is needed, Desire to take part, Knowledge of how to do it, Ability to actually do it in practice, and Reinforcement to keep doing it once the spotlight moves on.

What makes this useful is the ordering. The classic failure, which Prosci flags directly, is jumping to Knowledge, running the training, before Awareness and Desire are in place. Train people who don't yet see why, or don't want to, and you produce capable resisters. So when adoption stalls, find the first step a person is stuck on and fix that one. Don't just repeat the training louder. Someone who won't touch the new tool rarely needs another tutorial; they need a reason to care, which is two steps earlier.

flowchart LR
  A(["Awareness
why change?"]) --> B(["Desire
want to take part"])
  B --> C(["Knowledge
how to do it"])
  C --> D(["Ability
do it in practice"])
  D --> E(["Reinforcement
keep doing it"])
  E -.->|"adoption sticks"| F(["New default"])
ADKAR runs in order, a person blocked at an early step can't be fixed by pouring effort into a later one. Leaders Loop

An honest limitation. ADKAR is a vendor-developed framework built from consulting practice rather than independent peer review, and its neat linear sequence flatters reality, real people loop back, lose desire after gaining knowledge, or backslide at reinforcement. It's most valuable as a diagnostic checklist ("where is this person stuck?") rather than a stage you tick off once and leave behind.

A worked example

Take a 40-person operations team rolling out a new scheduling platform to replace a tangle of spreadsheets. (Illustrative scenario; figures are for teaching, not a real deployment.) Leadership mandates it, IT runs two training sessions, and a month later only about a third of shifts are being scheduled in the new tool. On paper the change is "live." In practice it hasn't been adopted.

The instinct is to send a sterner email and book more training. The better move is to diagnose. Pull the resistance apart using Kotter and Schlesinger: the senior schedulers aren't confused, they're losing the spreadsheet mastery that made them indispensable (loss of something valued), while newer staff are simply anxious about getting rotas wrong in an unfamiliar system (adjustment strain). Two different reasons, two different responses. For the seniors, involve them, make them the configuration owners and super-users, so the new tool becomes a source of status rather than a threat to it. For the anxious newer staff, facilitate and support, a side-by-side parallel-run week and a named person to ask, not another slide deck.

Now run the holdouts through ADKAR. Most aren't stuck on Knowledge at all; they're stuck on Desire, because nobody connected the platform to anything they care about, it was sold as an IT project, not as fewer Sunday-night scheduling clashes. Fix the early step (a one-page "here's the Sunday-night problem this kills"), keep the seniors invested, support the nervous, and reinforce by retiring the old spreadsheets entirely so there's no path back. Adoption climbs because the friction was named and matched, not overridden.

Frequently asked questions

Isn't some resistance just people being difficult?

Occasionally, yes, but treating it that way as your default is the costliest assumption you can make. Coch and French's work, and most since, suggests that what looks like obstinacy is usually a rational reaction to an unaddressed cost, risk, or loss. Start by asking what the resister can see that you can't. If, after genuinely listening, it really is intransigence over a settled decision, that's a performance conversation, but it should be the diagnosis you reach last, not first.

How is "driving adoption" different from just rolling something out?

A rollout is the launch event; adoption is the steady state afterwards. Plenty of changes launch flawlessly and then quietly die because no one planned for the gap between "available" and "used by default." Adoption work is everything that closes that gap, reinforcement, removing the old way, making the new way the easiest path. If your plan ends at go-live, you've planned the rollout and skipped the adoption.

What if I don't have time to involve everyone?

You rarely can, and you don't have to. Involvement is the strongest lever but also the slowest, which is exactly why Kotter and Schlesinger offer six options. When speed matters, involve the people with the most power to block you and the most useful information, communicate clearly with everyone else, and reserve the heavier levers for genuine emergencies. The point isn't to consult endlessly; it's to match the effort to the resistance you actually face.

How do I know whether a change has truly stuck?

Watch what happens when the attention stops. A change has been adopted when the new behaviour survives without you pushing, when nobody's reverting to the old spreadsheet, the workaround has died, and people would now find it odd to do it the old way. In ADKAR terms, that's the Reinforcement stage doing its job. Until then you have compliance, which lasts exactly as long as the supervision.

Does communicating more reduce resistance?

Only when the resistance is caused by missing or wrong information, which is one reason among several. Communication is the right tool for confusion; it does little for someone who understands the change perfectly and stands to lose from it. That case calls for involvement or negotiation, not another all-hands. Over-communicating to a problem that isn't informational just signals that you haven't diagnosed it.

Related in the Toolkit

Managing resistance sits inside the wider craft of change leadership, the structured playbooks for sequencing a change (change models like Kotter, ADKAR and Lewin) give you the stages, and how you talk about the change throughout (communication during change) is often the difference between resistance you can work with and resistance that hardens.

Where to go next