Can automating one department unbalance another?

A department automates part of its work. Its lead times fall, its indicators improve, the project is a success. A few months later, the next department is overwhelmed, another receives incomplete cases, a third has lost information it used to get through an informal exchange.

These effects cannot be seen from the automated department, because they happen elsewhere, often with a delay. They come from the connections between departments: what one passes to the other, at what pace, with what information attached to the case. Changing one step changes these connections. It is to see them before acting that we designed the AI Flow Map.

• Speeding up moves the bottleneck.The next department receives work faster than it can handle it, and the waiting shifts to it.
• Some information flowed without being written down.A phone call, a note in the margin, a habit: automation removes them without knowing it.
• Indicators remain local.Each department measures its own performance; nobody measures the flow from end to end.

1. 01Before automating a step, identify the departments that receive its output.
2. 02Ask them what they get today, including what is written down nowhere.
3. 03After launch, track an end-to-end indicator, not just the one for the department concerned.

This is the situation that led us to design the AI Flow Map: seeing the connections before changing them.

• How do you measure the return on investment of an AI project?→
• How do we choose our first use cases?→
• Why do our pilots never scale up?→
• ‘A successful pilot proves it will work everywhere.’→