Part of ESM Institute Programme in development
← All examples

Explore a better customer offer.

Find a useful new possibility, not simply more output.

Illustrative example · no measured results claimed

The starting challenge

Develop a small, testable service idea from the fictional situation below.

A small service business provides one-off installations. Some customers have asked what help is available afterwards. No evidence of willingness to pay, frequency of demand or delivery cost has been gathered.

A useful redesign

Choose one idea and define the evidence that would justify continuing or stopping.

Make the source visible, keep assumptions separate and agree who checks the result before it is used.

Where AI can help

Try this instruction with the fictional sample in a tool you are permitted to use.

Propose three modest ways to support these fictional customers after installation. For each, state the customer need, assumptions, human contribution and smallest sensible test. Do not invent demand, revenue or market statistics. Include a non-AI option.

The human responsibility

  • Is the idea linked to a real customer question?
  • Are assumptions and evidence clearly separated?
  • Is there a way to test demand before committing significant effort?

Evidence to gather

Look for evidence of demand and achievable delivery cost. Treat projected revenue as a hypothesis.

Limits to keep in view

Use fictional or approved public information for this starter exercise. Do not upload confidential, personal or sensitive business information. Do not use the output to make high-stakes or regulated decisions.

Stop or change the approach if checking effort, errors or downstream effects outweigh its usefulness.