You deployed the platform. The executives signed off. The launch email went out. The training is live. And six months later, adoption is stuck somewhere between fifteen and thirty percent, most of it performative. Deployment is a project plan; adoption is a population of human beings choosing, every day, to do their work differently.
Most change campaigns fail because they treat adoption as a communications problem: build a deck, run a town hall, send the cascade emails, write the FAQ. Adoption is a behavioral problem, and behavior at scale follows rules that four decades of behavioral science have already documented.[1]
This playbook is what I have learned operationalizing those rules across Fortune 100 and public sector AI rollouts. It is built for change leaders, CHROs, transformation owners, and program managers: anyone who has been handed an adoption target and is wondering why the standard playbook is not getting them there. It is structured as a matrix you can use Monday morning.
The Two Segmentation Lenses Most Teams Miss
Effective adoption requires segmenting your population on two independent dimensions. Most change teams use one (usually org chart or function) and then wonder why their cascade is not landing.
Lens One, diffusion segments: who moves first. Everett Rogers’ research on how innovations spread divides any population into five segments based on willingness to adopt.[2] The first two, innovators and early adopters, are combined below.
The remaining 16 percent are laggards, who move last or never, often acting on a fear nobody has surfaced.
Lens Two, stages of change: where each person is on their own journey. James Prochaska’s transtheoretical model says any individual moves through five psychological stages when changing a behavior:[3] precontemplation (“I do not see a problem”), contemplation (“maybe there is a problem”), preparation (“okay, I am going to try”), action (“I am doing it”), and maintenance (“I have been doing it”).
These two lenses are independent. You can be an early adopter in precontemplation: the personality that moves first, but who has not yet seen why this matters. You can be early majority in preparation: a pragmatist who has seen enough peer proof to be ready to try. You can be a laggard in maintenance: someone who finally moved, and now you have to keep them from relapsing.
Segmenting on org chart alone hides all of this. Segmenting on the matrix surfaces it. And once you can see where your population actually sits, your campaign strategy gets specific instead of generic.
The Matrix: Who Needs What, When
Each cell of the matrix tells you what move to make. Read across a row to see how a segment moves through their journey. Read down a column to see how to design a stage-specific intervention across your whole population.
Before you read it, ask yourself three questions. What percentage of my population is in each diffusion segment? (Do not guess. Look at prior change initiatives. Who showed up first?) What percentage of each segment is in each stage of change? (Survey data and engagement signals beat intuition.) And where is the largest cell? That is where your campaign needs to go to work. Most campaigns spend on innovators and ignore the early-majority precontemplation cell, which is often the biggest single block.
| Segment | Precontemplation | Contemplation | Preparation | Action / Maintenance |
|---|---|---|---|---|
| 16%Innovators & Early Adopters | They would move first, but nobody has asked them. The sponsor reaches out directly: “we want you to help shape this.” | They are interested and ready to talk about it. Give them early access and talking points so they can win over peers. | They are about to start. Clear every obstacle and make their first week a success story. | They use it well, but quietly. Make them visible: a champion network and public celebration, so their proof spreads. |
| 34%Early Majority — Pragmatists | They do not yet see what this has to do with their job. A 15-minute call with a peer in their function who already moved does what no deck can. | They are open, but waiting for proof it is ready. Ship the whole product: toolkit, integrations, FAQ, success stories from their function. | They are willing to try but do not know when to start. Give a concrete cue: “when X happens this week, do Y.” | They have started and could still drift. Manager check-ins every two weeks, coaching circles, peer momentum updates. |
| 34%Late Majority — Skeptics | They think waiting is the safe choice. Show the cost of sitting out: “70 percent of your peers are already in.” | They have doubts they will not say out loud. Arrange a call with someone like them who had the same fear and moved past it. | They will try, but they are afraid of failing in public. Smallest possible first step, a buddy, and a manager script for the conversation. | They moved, and the new habit is fragile. Recognize it: visible credit, a word from the executive. |
| 16%Laggards | They are resisting, usually out of fear for the job or loss of control. One-on-one only: listen for the real fear, then address that. | They are wavering but wary. Manager and skip-level conversations: “you are not alone, here is how we support you.” | They will only move with help at their side. Office hours, a pre-built path, reassurance over inspiration. | They finally moved and want no attention for it. Quiet recognition and steady manager support. |
Reading the matrix. Most enterprise campaigns over-invest in the top-left cell (recruiting innovators) and the top-right cell (celebrating champions). The cells where adoption actually wins or loses are the middle two rows, in the contemplation and preparation columns. That is where pragmatists and skeptics live, and that is where most rollouts stall.
The Five Frameworks Behind the Matrix
Each cell of the matrix is grounded in a specific behavioral science framework. If you understand the underlying mechanics, you can adapt the matrix to your context. If you only memorize the cells, you will misapply them when reality breaks the pattern.
1. Self-Determination Theory: why mandates fail. Deci and Ryan’s research is the academic backbone of adoption without mandates. Intrinsic motivation requires three psychological needs: autonomy (I chose this), competence (I can do this), and relatedness (I belong to others doing this). When any of the three is missing, you get compliance at best and quiet resistance at worst.[4] The nuance most leaders miss: autonomy is not the absence of structure. The question is whether the structures are autonomy-supportive or controlling. Same institutional accountability. Different psychological experience. Wildly different adoption outcomes.
2. Diffusion of Innovations: who moves first. Rogers’ work explains why the chasm between early adopters and early majority is where most enterprise rollouts fail.[2] Early adopters use the platform. Pragmatists use the platform plus the manager toolkit, plus the integration, plus the LMS hookup, plus the comms templates, plus the executive playbook. Anything missing becomes a reason to wait. Build the whole product, not the core product. And recognize that late-majority and laggard executives respond to peer references, not vision. A ten-minute call with another exec who has used the product moves more than any whitepaper.
3. Stages of Change: where each person is. Prochaska’s framework is the most useful tool for designing the right intervention at the right time. A precontemplator does not need a manager toolkit; they need awareness content that makes the cost of inaction visible. A preparer does not need a town hall; they need a friction-free first step. Mismatched intervention is the most common reason change campaigns underperform: right messaging delivered to the wrong stage.[3]
4. Fogg Behavior Model: B = MAP. BJ Fogg’s model says behavior happens when motivation, ability, and a prompt converge in the same moment.[5] The implication for change campaigns is the 80/20 inversion: most teams spend eighty percent of their effort on motivation (town halls, vision decks, exec emails) and twenty percent on ability (one-click flows, peer mentors, removing the old workflow). Best practice is to flip that ratio. In the matrix, the preparation column is almost entirely a Fogg play.
5. Implementation Intentions: closing the intent-to-action gap. Peter Gollwitzer’s research is the single most evidence-backed technique for closing the gap between “I plan to” and “I did.” The format is rigid: “When situation X arises, I will do Y.” Studies show two to three times higher follow-through than generic goal-setting.[6] An implementation intention sounds like: “When my Monday morning calendar opens, I will run my pipeline review in the new tool.” That single sentence outperforms three town halls.
Where This Lands in an AI Transformation
Most enterprise AI investments stall because the change strategy was never built for the kind of behavioral shift AI requires. AI changes more than a workflow: it changes the meaning of the work. People are asking “what does this mean for my role, my team, my future?” long before they ask how to use the tool.
Those questions cannot be answered by a launch deck. They are answered, or left unanswered, in hundreds of small moments: the manager who acknowledges the question or dodges it, the peer who has tried the tool and shares what they found, the executive who is visibly using AI themselves or quietly delegating it. Each of those moments is a cell in the matrix.
As AI takes over the optimization and execution layer, what is left is the adaptability layer, and none of those skills are mandated into existence. They develop through autonomy, competence, and relatedness, exactly the conditions the matrix is designed to create.
The organizations that get stronger from AI will be the ones whose change leaders treat adoption as a behavioral problem with a behavioral solution, not a communications problem with a communications budget.
How to Use This Playbook This Week
If you are running a live AI rollout right now, here is the sequence.
Map your population on the matrix. Pull engagement data, prior change initiative response rates, and any survey signals you have. Estimate where each segment sits. Do not aim for precision; aim for which cells are biggest.
Identify your three biggest cells. Most rollouts have two to three cells where the bulk of the population sits, and that is where your budget and manager attention should go. Choose by size, even when a smaller cell is louder.
Design the cell-specific intervention. For each of your three biggest cells, choose one intervention from the matrix. Build the artifact for it: the peer call script, the manager toolkit, the implementation intention prompt. Resist the urge to build for every cell at once.
Equip managers as the channel. Across every cell, the manager is the highest-leverage delivery mechanism. Give them three things: talking points (the rider), permission to localize (the elephant), and a one-page workflow (the path).[7] Skip the fifty-page playbook.
Measure leading indicators by cell. Are precontemplators moving to contemplation? Are contemplators signing up? Are preparers completing first sessions? Are action-stage people returning in week two? Each cell has its own signal. Track the right one.
That is the whole loop. Segment, identify the biggest cells, design cell-specific interventions, equip managers, measure leading indicators, adjust. Repeat every two weeks until adoption is sustained behavior, not reported behavior.
What This Playbook Is Not
The matrix surfaces what is happening in your population; changing it takes work the matrix can only point to. That work is in the design and delivery of cell-specific interventions, the equipping of managers, the willingness to slow down a launch to fix a wrong intervention, and the discipline of measuring leading indicators instead of vanity metrics.
Most of the frameworks here are publicly available. Adoption still stalls because the science gets lost in the gap between strategy and delivery. The matrix is one attempt to close that gap.
If your AI investment is live but not yet landing, the entry point is the Trail Diagnostic, a half-day session that maps your population onto the matrix, identifies the cells where adoption is actually stalled, and surfaces the cell-specific interventions your campaign is missing.
The session produces a written artifact: a population map and a sequenced set of first moves for the three biggest cells. It becomes the foundation for a 90-day plan built around behavioral change rather than communications volume.
If you are running a live AI rollout and the standard playbook is not getting you to the adoption target, reach out directly. The conversation starts there.