Briefing
The situation
It's a year since you rolled the AI tools out broadly. Eight in ten employees use them every week. The adoption target has been met and celebrated. In the staff survey most people say they save time. The licence cost is moderate and nobody questions it. But lead times, customer satisfaction, delivery volume and quality look like last year's. Nobody has stopped doing anything. No way of working has been redesigned. Now you have to decide what next year should be about. Here is what you don't know: whether the time saving is real or a feeling, whether the time goes to better work, more work or recovery, whether the value sits in places you don't measure, whether a handful of use cases account for almost all the benefit — and whether it is simply too early to see anything.
Discussion
Questions to wrestle with
What have we actually measured?
- 1.What does high usage prove, as you as a leadership team see it — and what does it not prove?
- 2.If employees save time and the metrics stay flat: where do you as managers see the time went?
- 3.Can you as leaders name one thing someone in your own part of the business has stopped doing, or does differently, because of AI?
Who is going to realise the value?
- 1.Who owns the question of value today in your leadership team — whoever rolled out the tools, or the people who lead the work?
- 2.What would you as managers need to remove or redesign for the time saving to show up anywhere?
- 3.What do you as a leadership team want to be able to show in twelve months? Write one thing down now, in the room.
Framework · From usage to value
To lean on
Usage
People log in and try it. Always measurable.
Habit
The tool is part of the working day. Measurable, and easy to mistake for benefit.
Changed way of working
Something is done differently, or not at all. Requires someone to decide it. This is where most get stuck.
Value
The change shows up in something you already care about — lead time, quality, customer, cost. It does not happen by itself.
Decision
Possible paths
- AContinue broadly and trust that the value will come with time.
- BPick three use cases and measure them hard against the business's own metrics.
- CMove responsibility from IT to line managers, who get to show what has changed in their work.
- DStop measuring usage altogether and measure outcomes only.
Triggers
Drop in when the discussion stalls
- ▸The board asks for one example of business value. Not usage statistics.
- ▸A team manager says the team produces more text than ever, but the decisions haven't got better.
- ▸The most active user turns out mostly to have redone things that already worked.
- ▸A team that barely uses the tools delivers best in the whole organisation.
For the facilitator
Tips to get more out of it
- The group often defends the adoption figure, because it is the only success that can be shown. Let it stand — and ask what it is a measure of.
- If the conversation slides into cost or vendor: remind them that nobody questions the cost in this scenario. This is about what you want, not what you pay.
- Option A is the most comfortable and may be right. Then ask what you expect to see in a year, and what you do if you don't see it.
- Option C moves the question to where the work is done, but puts a new burden on managers who already have a full plate. Have the group say what they take away in return.
- Hold on to the question of what someone has stopped doing. If nobody can answer, you have usage, not change.
Reflection
To take with you
- "How do you act as a manager if this scenario happens in your team or department?"