How to Reconcile Walk-In Enquiries With Campaign Activity

A customer who walks into a shop may have seen a sign, received a referral, searched online last week or simply passed by. Asking “Where did you hear about us?” can provide useful clues, but the answer is not a perfect record of what caused the visit. A sensible measurement process collects a small amount of consistent information, distinguishes customer recall from verified campaign codes and reports uncertainty instead of turning every sale into a marketing success story.

The operational problem

Choose the measurement question before designing the form. “What first made you aware of us?” differs from “What prompted today’s visit?” A customer may name a friend as the original source but say a festival offer finally motivated the visit. If staff use these questions interchangeably, the resulting totals are not comparable. Do not force every customer to answer, and do not ask for personal details that are unnecessary for evaluating the campaign.

How to review the result

Report total walk-ins, responses collected, unknown responses, observed redemptions and any sales value that can be reliably linked to a transaction. State the denominator: “12 of 30 responding visitors mentioned the leaflet” is clearer than “40% of customers came from leaflets” if only a subset was asked. Avoid converting response shares into causal claims. If the process is too disruptive at peak hours, sample defined time blocks and disclose that the result is a sample rather than a full count.

Failure modes to watch

Staff may prompt customers with a desired answer, especially if management celebrates a particular channel. A question such as “You came because of our Instagram post, right?” biases the response. Another failure is comparing a busy holiday week with a quiet ordinary week and attributing the entire difference to marketing. Weather, local events, opening hours and stock can influence footfall. Record the campaign window and notable operating changes so later readers understand the context.

A workable procedure

Use a short set of options plus “other” and “not sure”. Keep categories stable for the duration of the campaign: storefront signage, printed leaflet, referral, social post, online search, paid advertisement, repeat visit and unknown may be enough. If a campaign uses a unique voucher or QR code, record that separately as observed evidence rather than replacing the customer’s answer. At the end of each week, reconcile paper notes or POS annotations with the campaign dates and remove duplicate records using a privacy-appropriate method.

Example from a small business

A neighbourhood café distributes a leaflet offering a limited-time breakfast combination. At checkout, staff ask what brought the customer in today and optionally note whether the leaflet was shown. One customer remembers the café from Instagram but presents the printed voucher. The log records both the recalled discovery source and the observed offer redemption. In the report, these are two different measures; the business should not count them as two separate customers or claim the leaflet alone caused the visit.

Implementation note

A small, consistent log is more useful than a complex attribution spreadsheet nobody fills in. Train staff on one neutral question, keep options unchanged during the measurement period and report verified actions separately from customer recollection. Use the results to guide the next test, not to declare a channel universally effective.

Working record to keep

Choose evidence that another staff member can independently check: a live page, a confirmed supplier reply, a dated proof or a completed test. A tick mark without a defined check is weak evidence.

Conclusion

A small, consistent log is more useful than a complex attribution spreadsheet nobody fills in. Train staff on one neutral question, keep options unchanged during the measurement period and report verified actions separately from customer recollection. Use the results to guide the next test, not to declare a channel universally effective.