Leading Your People Through AI Change  /  Module 5
Module 5 of 5

From Rollout to Embedded

Your launch is done. This module focuses on what comes next — building the habits, conversations, and conditions that turn adoption into the way your team actually works.

~25 minutes · 5 sections + worksheet
1
Understand
Module 1
2
Plan
Module 2
3
Listen
Module 3
4
Build
Module 4
5
Embed
Module 5
Progress
0 / 5
How this module works
01Read the contents of the section
02Answer the quick check and worksheet question
03Your answers save automatically as you go
04Print your worksheet anytime from the bottom of the page
1
What Full Adoption Actually Looks Like

Most managers celebrate go-live. Full adoption looks different. Adoption is when people reach for the tool before reaching for the old way — unprompted, in their normal workflow, as the obvious choice.

The gap between "we've launched" and "they've adopted" is where most AI projects lose momentum — through gradual drift back to old habits, compliance that looks like engagement, and usage that plateaus at 20% of its potential.

That gap often traces back further than the rollout itself. Business readiness — whether the organisation had the clarity, leadership alignment and capability in place before go-live — usually explains more than anything that happened during it.

What full adoption looks like:

People use AI across tasks, on their own initiative
They talk about it with colleagues without a manager in the room
They find new use cases you never suggested
The tool becomes part of how they think about their work — the first move, not an afterthought

The signal to watch for: people reaching for the tool between check-ins, on tasks you haven't specifically prompted. Usage that is self-directed, varied, and growing.

Quick check
What's the clearest signal that your team has reached full adoption?
Your worksheet
Be honest. Which description fits what you actually observe?
2
Measuring Adoption Honestly

Login counts tell you who opened the tool. Real adoption data tells you whether using it changed how people work. Measure behaviour, not access.

What to look for:

Which tasks the team is using AI for, self-reported
Whether time or quality on those tasks has shifted
Whether people are finding new use cases independently
Whether they're talking about it with colleagues on their own initiative

These signals tell you far more than a dashboard. Translating this for leadership matters too — business language lands better than usage percentages:

Usage percentage

"We're at 60% usage" — invites follow-up questions you'd rather answer yourself.

Business language

"Three of my five team members are using AI for their first draft on client reports, cutting average prep time from 4 hours to 90 minutes" — a leadership-ready adoption update.

When adoption is lower than expected, name it with a diagnosis and a plan. "We're at 40%, here's what's driving it, and here's my next move" builds more confidence than waiting until the gap is too wide to address quietly.

Quick check
A leader asks how adoption is going. You're at 40%. What's the strongest response?
Your worksheet
Login counts? Self-report? Conversation? What would a more honest measure look like?
3
What Progress Actually Looks Like

Usage numbers will tell you people are using the tool. They won't tell you whether capability is actually growing. The two are very different things, and it's worth knowing what real progress looks like so you can recognise it — and so you can tell when something needs attention.

Here's a rough shape of what you can expect across three months, when things are going well.

30 days

People are trying it on one or two tasks. Outputs are mixed. Some are delighted, some are frustrated. Questions are coming in. This is exactly where you want to be.

60 days

A few people have found their groove — specific tasks where it reliably saves them time. Others are still experimenting. The team is talking about it informally. Shared prompts are appearing.

90 days

People are reaching for it without being prompted — on their own work, not just assigned tasks. Someone's teaching a colleague something they figured out. The tool has become part of how the team works.

The most reliable signal at 90 days isn't usage frequency — it's unprompted use. When someone picks up the tool to solve a problem you didn't ask them to use it for, capability has genuinely landed.

If you're at 90 days and that's not happening yet, it's worth going back to Module 3. Something is still sitting in the way — a signal your team has been sending that hasn't been fully heard yet.

The best signal that capability is real: someone uses the tool for something you didn't suggest, on a problem they cared about solving.

Quick check
At 90 days, what's the most reliable sign that genuine capability has developed?
Your worksheet
Trying it on one or two tasks? Finding their groove? Reaching for it unprompted? Be specific about where you actually are.
4
When Progress Plateaus

Every rollout reaches a plateau. Usage levels off. Early adopters are embedded. The middle group is dipping in occasionally. And a small group is still watching. This is normal — and it's the moment that separates a managed rollout from a drift back to old habits.

A plateau signals that your current approach has reached its natural ceiling. The next phase requires a different lever — and diagnosing which one saves you from pushing harder on something that has already done its job.

The three most common causes:

Narrow use cases

People do the one thing from training and see little reason to go further. Show them one more use case that fits their own work.

Inconsistent reward

For some people AI saves time; for others it creates more editing work. Match the task to where the win is real and repeatable.

An unresolved concern

Usually around status or fairness, and it's keeping a group quiet. It responds to conversation, not more training.

How to diagnose which one: have three honest conversations with people at different points on the adoption curve. Ask them what gets in the way of using it more. The answers will tell you which lever to pull.

Don't guess — ask. Three honest conversations will tell you more than any dashboard.

Quick check
Adoption has plateaued. What's the right first move?
Your worksheet
Use cases not expanded, inconsistent reward, or unresolved concern? Or a combination?
5
From Rollout to Business as Usual

There's a moment in every successful rollout when you stop managing adoption and start managing the work. AI is just there. It's part of how the team operates. Nobody calls it a rollout anymore.

Getting there requires a deliberate handover. Build AI into the structure — the templates, the processes, the onboarding for new team members. When it lives in the workflow itself, it runs without anyone having to champion it.

The signs you've reached embedding:

New team members pick it up as part of how the team works
People use it in ways you never suggested
Usage is steady and self-directed rather than peaking around check-ins
The conversation shifts from "should we use AI" to "how do we use it better"

Your job is to create the conditions where using AI becomes the obvious, natural choice — and then step back. That's the work of embedding. The tool handles the task. You handled the people. That's the harder job, and the one that actually determines whether this works.

Quick check
What's the most reliable sign that AI has become business as usual for your team?
Your worksheet
Describe what embedded looks like. Then name the gap.
Module 5 Worksheet

Your Embedding Plan

Your five answers, gathered from each section above. Edit them up there — this is a read-only copy for review, saving and printing.

Course complete
You've finished all five modules.

You now have a framework for leading people through AI change — from the first announcement to genuine embedding. The work is in the conditions. You've built them.