Most of this guide can be done in a month without new software. The order matters: measure first, fix the trip second, deal with the few third, and test the expensive ideas last.

1. Week 1 · MeasureRun the audit in chapter 1. Pull twelve months of returns by product and customer, one cost-per-return number, and the four-way repeat split from chapter 5. Start the scorecard.
2. Week 2 · Fix the tripMake your terms match in all five places. Return something on your phone and fix the worst step. Reorder the outcome screen. Turn on refund-at-scan, with a random 10% held on current timing.
3. Week 3 · Reasons and the fewRewrite the reason list and hold the first weekly defect review. Run the chapter 11 tool to find your serial-returner line, look at ten customers above it by hand, and send the first plain notes. Set a keep-it rule for low-value items from the chapter 9 arithmetic.
4. Week 4 · Test and check the floorLaunch the bonus-credit holdout with the break-even share from chapter 4 written down. If you charge a fee, or plan to, run the chapter 5 tool and write the decision rule; reverse an untested fee for a random half of new customers. Send your terms to counsel. Hold the first monthly scorecard review.

Measure the next order, fix the trip, charge only the few, and test anything that costs money.

At day thirty you should have a scorecard, one set of terms, a portal that works on a phone, refunds that move at the scan, a weekly defect review, a serial-returner rule and at least one holdout running. Most brands never find out what their returns policy does to the next order. You will.

Do this

• Put the four weeks on the calendar today, with one owner for each week’s list, and book the first monthly scorecard review for day thirty.