Retail foot traffic analytics: visits and conversion

Retail foot traffic analytics explains store revenue through visits, transaction conversion and average ticket. It uses existing aggregate visitor counts and POS summaries for the same store and period. The useful question is whether sales changed because more opportunities arrived, more visits produced transactions, or each transaction generated more revenue.

Define the visit and the transaction

Foot traffic is a count of store entries over a period, as described in Shopify's foot traffic guide. An entry count is not automatically a unique shopper count. Keep the source's definition alongside every comparison, including how it treats employees, repeat entries, pickup visits and groups. Covirage is not affiliated with Shopify.

Use completed merchandise sale receipts for transactions. Exclude voids and return-only receipts from that count. Use revenue for the same channel and business dates. If revenue includes refunds from earlier purchases, label average ticket as period net revenue per sale transaction; a purchase-only basket measure requires sale receipt revenue instead.

For this analysis, net merchandise revenue excludes sales tax and deducts refunds. That is consistent with the relevant treatment in the Census Bureau's retail sales definition. Keep online orders shipped from a store separate when the buyer never entered the store. Retail POS analytics covers the transaction reconciliation that supplies these totals.

Measure the three parts of revenue

Transaction conversion = Completed sale transactions / Recorded visits

Average ticket = Merchandise revenue / Completed sale transactions

Revenue per visit = Merchandise revenue / Recorded visits

Revenue = Visits × Transaction conversion × Average ticket

The revenue identity is exact when every component uses the same rows and definitions. Conversion is a transaction-to-entry ratio here. It does not claim to measure the share of unique people who purchased.

Work the same periods for two stores

An illustrative US home-goods retailer compares April 4, 2026, with April 11, 2026. Both dates are Saturdays, and both stores have the same opening hours on both dates. Their existing aggregate counts use an unchanged entry definition. No returns or online sales are included in this example, so revenue per transaction is also the sale basket value.

Divide each transaction count by its visit count, then revenue by transactions and visits:

Percentages are displayed to one decimal and monetary ratios to two decimals. Calculate with unrounded values. Atlanta's visits rise by 1,320 / 1,200 − 1 = 10.0%, and revenue rises by the same 15,840 / 14,400 − 1 = 10.0%. Conversion and ticket hold steady, so the revenue increase is entirely explained by visits in this calculation.

Charlotte's visits rise by 1,000 / 800 − 1 = 25.0%, but transactions and revenue do not change. Conversion falls from 25.0% to 20.0%, a decline of 5.0 percentage points. That is the store requiring follow-up even though its revenue is flat.

Prove what changed with a sales bridge

Hold Charlotte's original conversion and ticket constant, change visits first, then change conversion. Average ticket stays unchanged throughout:

The bridge reconciles: $14,000 + $3,500 − $3,500 = $14,000. Changing the order of a bridge can allocate interaction effects differently, so use the same order each period. These are mathematical contributions; they do not establish why conversion changed.

For the retailer total, sum counts before dividing. On April 4, visits are 1,200 + 800 = 2,000 and transactions are 240 + 200 = 440, giving 22.0% conversion. On April 11, visits are 1,320 + 1,000 = 2,320 and transactions are 264 + 200 = 464, giving 20.0%. Averaging store percentages gives unequal stores the same weight.

Assign the follow-up to an owner

The store operations team owns the conversion review. For Charlotte, compare count-quality flags, stock availability, service coverage and checkout conditions for the same intervals. A daily total cannot establish that a particular hour was understaffed; bring hourly summaries before making that claim. Retail operations analytics covers the separate comparison of store labor and operating costs.

Marketing owns questions about attracting visits. Merchandising owns the availability and assortment review. Finance owns the revenue control. Record an action, an owner and the period to review again, while distinguishing an observed pattern from a demonstrated cause. The wider retail analytics framework connects those responsibilities.

Where it goes wrong

• Changing the count definition. Removing employees in one period but retaining them in another changes conversion without a sales change.
• Different coverage. Missing traffic intervals should be flagged as missing, not replaced with zero. Restrict comparisons to intervals with usable counts.
• Combining channels. Adding shipped online revenue to in-store transactions and visits breaks the relationship being measured.
• Treating transactions as people. Repeat entries, shopping groups and split receipts limit the interpretation of conversion.
• Inferring cause from a ratio. Lower conversion identifies a question. It does not prove staffing, pricing or assortment caused the change.

Compare your store summaries in Covirage

Bring a sample of your own aggregate traffic counts and POS summaries to Covirage, with matching store and period columns. Covirage maps the columns, its tools compute and check the figures, and the external AI model explains the result with every figure traced to its rows. Create a free Covirage account, then set up your business and bring your files.

See sales insights for the supporting business approach.