Store Selection for Trade Promotion Execution Audits

A practical method for defining the outlet population, stratifying stores, separating representative sampling from risk-based checks, and governing replacements and nonresponse in a trade promotion execution audit.

Trade marketing colleagues inspect a pilot promotional shelf in a grocery store.

To select stores for a trade promotion execution audit, start with a frozen outlet population, draw a random sample within documented strata, and keep risk-based checks in a separate track. Store selection determines whether the audit can support a regional estimate. An audit team that visits 40 stores near the sales office and finds most displays in place has not measured regional execution; it has measured those outlets. Agree replacement and nonresponse rules before fieldwork so the estimate remains defensible.

Store selection for a trade promotion execution audit: outlet population, representative sample, and risk-based check
Store selection separates a representative sample from a risk-based exception check.

What is the outlet population for store selection?

The outlet population is every store the promotion was assigned or eligible to reach during the audit period, not every outlet in the territory. Freeze the list at a stated cutoff and record outlet ID, region, format, distributor or route, promotion ID, planned dates and assignment status. State exclusions, such as permanently closed, outside the trading area or not participating. Keep assigned, eligible, completed and audited outlets as separate counts; an assigned outlet is not a completed execution. For a representative estimate, each outlet needs a known chance of selection. A hand-picked list does not provide that.

How should you stratify stores by region, format, and execution differences?

Before selection, divide the population into strata that could change execution: region, store format, distributor, route or promotion mechanic. Randomize within each stratum and merge very small strata under a written rule. Do not let a field manager fill a quota with outlets that are easy to visit. Keep the strata definitions and random seed with the audit file so the selection can be repeated. Risk factors such as high claim value, past discrepancies, late evidence or a new promoter belong in a separate targeted track. They can guide follow-up, but they are not a neutral basis for an execution estimate.

How do you separate the representative sample from the exception check?

Run two tracks and report them separately. A representative sample estimates how common compliance is across the defined population. A targeted check finds specific problems that need action. Mixing them into one number hides bias.

Audit tracks and the claims each can support
Track Question it answers Selection method What you can report What to avoid
Representative sample How common is compliant execution? Random selection within documented strata An estimate for the defined population, with uncertainty calculated by a qualified analyst Claiming precision without a calculation
Risk-based check Which high-risk outlets need follow-up? Rules such as claim value, discrepancy history or late evidence Named exceptions and follow-up actions Presenting findings as an overall execution rate
Convenience check What problems are easy to spot? Readily accessible outlets Operational triage only Extrapolating to all outlets

What replacement and nonresponse rules should be set before the first visit?

Define nonresponse categories: closed, wrong store record, no longer assigned, refused, inaccessible, promotion absent or incomplete audit. Log every attempt. Allow replacement only from a randomly ordered reserve list in the same stratum, and set a maximum number of replacements. Do not let auditors swap a difficult outlet for a nearby one. For a targeted exception check, do not replace the named high-risk outlet to meet a quota; the exception is the point. Report attempted, completed, replaced and unavailable outlets. If nonresponse is high in a stratum, state that the estimate is limited rather than silently filling the gap.

What does an illustrative store selection plan look like?

Consider a hypothetical promotion in 1,200 assigned outlets across three regions and four formats. The team freezes the list, merges two very small format groups under a written rule, and creates two tracks: a random representative sample drawn within strata, and a risk-based list of outlets with late claims or prior discrepancies. For illustration only, it might audit 60 randomly selected outlets and 20 targeted outlets. Those figures are not a recommended sample size or evidence of statistical precision; the actual size must follow the decision tolerance and be calculated for the population. Report the random track as an estimate for the defined population and the targeted track as named exceptions. Do not add them into one compliance rate.

Store selection checklist

Use this checklist to keep the selection auditable.

  1. Freeze and version the outlet population with promotion dates and eligibility rules.
  2. Document strata and merge small strata before selection.
  3. Keep the random selection, reserve list and selection date.
  4. Tag every targeted outlet with the risk rule that put it on the list.
  5. Allow replacements only from the same stratum reserve; log the reason.
  6. Record nonresponse and report it beside completed audits.
  7. Separate representative estimates from exception findings in the final report.
  8. State limitations and what the audit did not measure.

Airpalm’s public description covers mobile visits, route planning, store audits, sales orders, inventory and promotion management (airpalm.com, checked 2026-10-09). If you use Airpalm or another field execution system, ask how the outlet list, strata and audit status are exported, and whether replacement and nonresponse reasons are captured at outlet level. A useful next step is to test store audit data quality before relying on the audit output: Store Audit Data Quality: Test It Before Buying SFA Software.

Frequently asked questions (FAQ)

How many stores should a trade promotion execution audit include?

There is no universal number. The required sample depends on the decision, the size and structure of the outlet population, the number of strata you need to compare, and the uncertainty the business can accept. Calculate and document the sample size with a qualified analyst. A risk-based check does not need a representative sample size because it is not designed to estimate the population rate.

Can a closed store be replaced by the nearest store?

Only under a prewritten rule that uses a randomly ordered reserve from the same stratum. Choosing the closest outlet lets convenience affect selection and can bias the estimate. Record the closure as nonresponse and report the replacement.

Can targeted exception findings be presented as overall compliance?

No. Targeted outlets are selected because they are more likely to have problems, so their results describe that group, not all eligible outlets. Report them as named exceptions and follow-up actions, separately from any representative estimate.

Put this guide into practice

This general-purpose worksheet is a starting point. Adapt its fields, responsibilities and rules to your business process.