Scaling a mobility pilot: check what made the first result possible

A pilot can work well and still leave the expansion decision open. A closely involved founder, a particularly experienced coordinator or unusually convenient support facilities may have helped produce the result. Those conditions are part of the finding. They should travel into the review alongside the headline numbers.

The purpose of a learning discussion with Sparrow Mobility Labs is to understand what can be carried forward, what needs further work and what the next stage would require from everyone involved.

Return to the original question

Begin with the decision agreed before launch. If the test concerned a clearer vehicle handover, review that process before claiming broader improvements in earnings, safety or customer experience. A team may notice interesting changes outside the original scope, but those observations deserve their own questions and evidence.

Keep the starting group and observation period visible. Explain changes in participants, vehicles or service demand. A result from the people who completed the test can look different from the experience of everyone who started. Missing records and early departures are part of the review, not administrative details to discard.

Separate activity from effect

A tool being used is evidence of use. Faster completion of a task may be evidence of a process change. Whether that change caused a wider outcome is another question. HM Treasury’s Magenta Book distinguishes process evaluation from impact evaluation and explains the role of credible comparison.

Use that distinction to keep the review proportionate. Describe the observed change, the comparison available and other plausible explanations. If the test cannot establish causation, it can still reveal usability problems, implementation requirements or a promising direction for the next study.

Price the support that made it work

List the people, preparation and interventions behind the result. Include training, coordination, repairs, data cleaning and the time spent explaining the new process. Record tasks handled by the building team that would eventually need a permanent owner. Otherwise a successful pilot can conceal an expensive operating dependency.

Ask which requirements will change in a different location. Energy access, local service capability, working windows and participant experience may all matter. Where the next step depends on another organisation, take that requirement into Partnerships before treating it as available capacity.

Choose the next learning step

Expansion can mean testing with a different team, running for a longer period or reducing the amount of direct support. Choose the change that answers the most important remaining uncertainty. Changing location, vehicle type and process together makes it harder to understand why the next result differs.

Write down the decision, the evidence behind it and the condition for another review. Keep unresolved questions with the record. A pilot creates durable value when another person can understand its limits and use its learning, including when the right decision is to redesign or stop.