Will a turnaround guarantee pay for itself?
Study design
Randomise the guarantee by counter-day. Include jobs rejected after inspection when recording conversion and analyse repair complexity by arm.
Instrument and data fields
Record counter-day assignment, quote, accepted job, ready date, refund and repeat visit.
Simulated evidence
Guarantee: 114 acceptances and 12 refunds from 160 quotes. Usual estimate: 99 acceptances and 4 refunds from 160 quotes.
Check before drawing a conclusion
Write down the comparison you would make and the unit each row represents. Identify one explanation the study design cannot rule out, then describe how it would affect the decision.
Reveal the worked analysis
After refunds, retained jobs are 102 versus 95. At ₹340 contribution on seven extra jobs, the lift is ₹2,380 before extra compensation or queue costs.
A longer trial should check whether difficult jobs select into the guarantee and whether long delays spill into later control days.