01
Step one: choose observable moments
List the moments a shopper can actually witness and record without guessing: greeting time, product availability, staff answers, fitting room state, queue length, checkout accuracy and farewell. Avoid vague items such as friendly atmosphere unless you define what counts. Example: a clothing store asks the shopper to enter, look for a specific jacket in a size, request help and note how long it takes for someone to respond. Each moment should be a yes, no or timed answer, so two different shoppers would record roughly the same thing at the same store.
02
Step two: set scoring and evidence rules
Decide how each item is scored and what evidence supports it. Use a simple scale, weight the moments that matter most to the brand, and require short comments for any low mark. Evidence might include receipt details, the time of visit, staff role description and a note on store conditions, always without recording people unlawfully or breaching policy. In the jacket example, a greeting within sixty seconds might earn full marks, product knowledge could be weighted higher, and a missing size must be confirmed by asking rather than assumed. Consistent rules keep comparisons across locations fair.
03
Step three: review, report and follow up
Before a report goes out, a reviewer should check for missing fields, unclear comments and scores that contradict the notes. Then group findings by theme so managers see patterns rather than isolated complaints. For the jacket store, a summary might show that greeting is strong but size checks are slow across three locations, leading to a coaching reminder and a stock-system check. Schedule a repeat visit after changes are made, using the same checklist, so improvement can be compared honestly. Share praise as well as problems, since recognition helps staff accept the process.