AI adoption starts with people: how they think and work, what matters to them and what they need when things change. Our experience in leading people and change shapes how we approach AI. The same three Working With Me principles apply: understand yourself and each other, understand the rules of the game, and establish the rhythm.
This guide pulls together what we have published on AI adoption: why rollouts stall, how to prepare, how to find the opportunities in each person's role, how to support the change, and what to measure. Read it straight through, or use the links to go deeper on any one part.
Why rollouts stall
A rollout can stop at licences and a training day without establishing whether anybody’s work changed. Clear opportunities, permissions and support help people understand what to try and how to review it. We cover this in why AI rollouts fail, and how to make adoption stick. Most organisations arrive at this from one of two places: already paying for AI tools without much evidence of value, or being deliberately careful and wanting solid foundations before they start.
Listen to your people and understand the business
Start with leadership's priorities and the experiences of the people doing the work. Ask what a better working day would look like, where effort is being wasted and what concerns people about change. Understand how they prefer to learn and receive support. Together, connect those insights to opportunities across each person's role and agree which are worth exploring.
Principle one: understand yourself and each other
People can start from different levels of experience, including no AI use at all. A working with me manual, written by the person themselves under their own sharing choices, records how they like to learn and communicate and what support helps. Managers use agreed actions and outcomes to offer support; private manual answers remain under each person’s sharing choices. It tells you who wants a worked example, who wants to try things alone, and where support should go.
The range of starting points is often wider than leaders expect. AI adoption in a neurodivergent team covers the different routes in that suit different ways of learning.
Principle two: understand the rules of the game
People need to know the business context and what they are allowed to do. How to agree the rules of AI at work with your team sets out what to cover: approved tools and accounts, what information may be used with them, how outputs get checked, who is responsible, and how the rules change. Leaders own the rules, and the people using AI should help make them usable.
The tools themselves come second. Build around the ones your organisation already approves, or assess suitable options against the work if you are starting fresh. Prioritise the opportunities that matter to your priorities and are realistic to support, without arbitrary limits on how many you take on.
Principle three: establish the rhythm
Agree who owns what, put support and review in the diary, and give managers a way to see how it is going. Change management is most of the work here: people need help, permission and time, not just access. Some things will be dropped, and dropping them is progress, because the goal is better work against the priorities that matter.
What to measure
Take a baseline before you change anything. If a monthly report currently takes a day and a half to produce and needs two rounds of correction, write that down. Then you have something honest to compare against.
- Useful adoption, meaning the work that actually changed, not logins, attendance or experiment counts.
- Time, including the review and checking effort that AI-assisted work adds.
- Quality, because faster work that has to be redone is not an improvement.
- Relevant costs, with a clear distinction between freed capacity and money actually saved.
Review agreed actions and outcomes within roles, then look for useful patterns across the team. Be clear about what is measured, who can see it and how it will inform support. Activity alone is not evidence that work improved.
Where this fits
The AI Adoption Programme runs over three months and works through each person's own role and work, the rules, the support and the measurement. It is tool-agnostic. A workshop can be a useful part of it, and the Working With Me platform connects manuals, business guidance, priorities and daily actions. We agree the platform access and support included in your engagement.
Common questions
What does AI adoption for a team actually involve
Understanding your people and your business, agreeing the useful opportunities and the rules of AI-enabled work, supporting the change through a steady rhythm, and measuring whether time, quality and costs actually improved against a baseline.
Do we need to already be using AI before we start
No. It works from wherever you are, whether that is tools you already pay for without much evidence of value, or nothing yet and a wish to get the foundations right first.
Will this push people to use AI more than they want to
No. More AI use is not the goal, better work is. The rules are agreed together, support is offered rather than imposed, and things that do not help get dropped. That counts as progress.
How is this different from a normal AI training day
Our programme starts by understanding each person and their work. Over three months, we use their shared working preferences to shape practice and support, help managers lead the change, agree clear business rules and measure the improvements. Learning to use suitable tools is part of that wider engagement.
How do you know if it is working without watching individuals
Compare agreed work outcomes with a starting baseline: useful adoption, time including checking, quality and relevant costs. Review progress within roles and across the team, with clear sharing boundaries. Managers need context to support people, while private manual answers remain under each person’s choices.