Store intelligence tracks a connected funnel of metrics, starting with the people who walk past your storefront and ending with the value of what they buy. So what metrics does store intelligence track, exactly? Footfall, capture rate, conversion rate, dwell time, zone engagement, live occupancy, queue length, staff-to-traffic ratio, and repeat visit rate — each a specific measurement tied to a specific decision. In 2024, 83.7% of US retail sales were forecast to happen in physical stores, which is why measuring what people actually do inside them still matters.
What Metrics Does Store Intelligence Track: The Full List
Every metric sits somewhere on a sequence: passers-by → capture rate → footfall → conversion → transactions → average transaction value → sales. Behavioral, operational, and loyalty metrics layer on top of that backbone. The funnel connects a window display to a cash register — that's its job.
One equation ties it all together:
Sales = Traffic × Conversion × Average Transaction Value
That formula turns each metric into a diagnostic lever rather than an isolated number. A weak sales week could be a traffic problem, a selling problem, or a basket problem — the metrics tell you which. Here's what store intelligence measures, grouped by where it lands in the journey:
- Funnel metrics — passers-by, capture rate, footfall, conversion, transactions, average transaction value (ATV)
- Behavioral metrics — dwell time, zone engagement, hot/cold zones, shopper paths
- Operational metrics — live occupancy, density, queue length, wait time, staff-to-traffic ratio
- Loyalty metrics — repeat visit rate, visit frequency
What Metrics Does Store Intelligence Track: Footfall and Capture Rate
Footfall is the count of people crossing your entrance threshold during a defined interval, after staff and deliveries are filtered out. For any interval, it equals the cumulative entry count at the end minus the count at the start.
Capture rate measures the step before that — how many people who walked past actually came in:
Capture rate = (people who entered ÷ people who passed by) × 100
Say 1,000 people walk past your frontage in an hour and 80 step inside. That's an 8% capture rate. Open-mall and high-street apparel often lands between 5% and 15%; anything above 20% is rare and sometimes points to undercounted passers-by rather than a brilliant window.
Footfall is the denominator for nearly every ratio further down — conversion, staff-to-traffic, sales per visitor. A miscount at the door distorts every downstream KPI. Keep the two terms separate: footfall counts only threshold-crossers, while passers-by form the capture rate denominator and must be counted over the same time window, or the math is meaningless.
Measuring both lets you separate your storefront from macro traffic shifts. If passers-by drop but capture holds steady, the problem is outside — roadworks, a nearby anchor closing, a mall entrance blocked. If passers-by stay flat but capture falls, look at your window, your cleanliness, your greeting.
Why Counting Method Accuracy Matters
Infrared beam counters run roughly 60–80% accurate and struggle when two people enter side by side. 3D stereo and depth cameras reach 95–99% in controlled conditions. That gap isn't academic: a systematic 10% undercount inflates apparent conversion and breaks any store-to-store comparison.
Direction matters too. Legacy gross counting assumes entries roughly equal exits, which falls apart at open and close, and whenever shoppers linger in the doorway. Directional counters that separate in-counts from out-counts are the baseline for reliable KPIs. Modern systems strip staff out using worn tags, schedule-based filtering, or pattern recognition — include staff entries and you inflate footfall while pushing conversion down, a double error that compounds across every ratio.
Conversion Rate: What Metrics Does Store Intelligence Track at the Point of Purchase
In-store conversion rate is transactions divided by visitors, times 100.
Conversion rate = Transactions ÷ Visitors × 100
180 transactions across 1,200 visitors works out to 15%. Benchmarks vary by format:
- Specialty retail: 15–30%
- Grocery: 20–40% (need-based missions can push it higher)
- Big-box: 10–20%
Compare that to e-commerce, where 2–3% is the global average. The gap exists because physically walking into a store signals far more intent than a click does. Online conversion also turns on page speed and navigation; physical conversion turns on staff engagement, product availability, and the queue experience. A fix that lifts one won't touch the other.
Watch one trap: conversion often dips on your busiest days. When traffic spikes, staff-to-visitor ratios thin out, queues build, and shoppers leave without buying. Read conversion next to your staffing and queue numbers — never alone. A rolling 13-week baseline by store and daypart tells you far more than a sector average that may not describe your format.
Here's where the sales equation earns its keep. A 10% sales drop splits into a traffic contribution, a conversion contribution, and an ATV contribution. Each points somewhere different: traffic to marketing or location, conversion to selling and assortment, ATV to pricing and merchandising. The decomposition tells you which team to call.
Dwell Time, Zone Engagement, and Shopper Path Metrics
Dwell time is how long a visitor spends in the store or in a specific zone. Zone engagement combines two things — how many people enter a zone and how long they stay per visit. Retail video analytics research consistently lists zone dwell and foot-traffic-by-location among the core metrics the technology actually measures.
Dwell correlates with basket size, which makes it the raw material for merchandising decisions. The patterns map clearly:
- High traffic, low dwell — a thoroughfare. People pass through; they don't shop it.
- High dwell, low sales — a mismatch. Customers linger but don't buy, so check pricing or assortment.
- High reach, high dwell — premium real estate for high-margin or new products.
- Low reach, low dwell — consider relocating or trimming the assortment.
Shopper path metrics add the sequence — which zones a visitor hits, in what order, how deep into the store they go, and how they flow between categories. That's how you find out whether a promo end-cap actually pulls traffic into the adjacent aisle, or just sits there looking busy.
Productive vs. Unproductive Dwell
Longer dwell isn't automatically a win. Time spent stuck in a checkout queue or waiting at a service desk is unproductive dwell — it pulls satisfaction down, not up. Partition dwell by area before reading anything into an aggregate increase. A rising store-level dwell figure driven by longer queues is a warning, not a trophy. Because dwell distributions are long-tailed — a few very long stays skew the average — median and percentile values tell you more than the mean.
Occupancy, Queue Length, and Staff-to-Traffic Ratio
Three operational metrics tell you what's happening inside the store right now:
- Live occupancy — the real-time count of people inside (entries minus exits).
- Queue length — how many customers are waiting at a service point at a given moment.
- Staff-to-traffic ratio — staff hours divided by visitors over the same period.
Live occupancy beats raw entry rate as a service-load proxy because it shows who's actually present, not just who came through the door. It drives staffing, fitting-room management, HVAC tuning, and usage-based cleaning. During COVID-19, many jurisdictions turned occupancy into a compliance number — capacity calculations often used one person per 20 m² (roughly 215 sq ft) — which pushed retailers toward directional counters and live "enter / please wait" displays.
Queue service level is usually written as a threshold: "80% of customers through checkout in under five minutes." Grocery and mass-market shoppers expect under five minutes, and satisfaction drops sharply past that point. Measure it, and you can open lanes before the queue forms instead of after.
Staff-to-traffic ratio is the backbone of labor scheduling. Forecast traffic per 15-minute slot, multiply by your target ratio, and you get the staff hours you need — then build schedules that scale with demand rather than fixed shifts. Supermarkets might run one staff hour per 20–50 visitors; luxury can approach one hour per two visitors, justified by a far higher ATV.
One technical note: occupancy drifts if entry and exit counts are even slightly off, because errors accumulate. Good systems enforce non-negative values and reset to zero during closed hours so yesterday's drift doesn't poison today's count.
Repeat Visit Rate and Visit Frequency
Most loyalty programmes tell you what people bought. These two metrics tell you whether they came back at all — including on visits where they didn't buy anything.
Repeat visit rate is the share of uniquely identified visitors who return at least twice within a period. Visit frequency is the average number of visits per unique visitor. Together they describe loyalty in behavior, not just in transactions, and they power cohort segmentation — new, occasional, regular, and heavy visitors — feeding customer lifetime value models and offer timing. A second-visit welcome offer, a reward for regulars, an assortment refresh paced to visit cadence: all of it starts here.
Two cautions. Sensing-based methods (Wi-Fi probe, BLE) catch non-buying visits that POS loyalty data misses, but MAC-address randomization in modern phones means device-based systems only track a subset of visitors — often estimated at 30–70%. Treat the output as a relative trend, not an absolute headcount. The privacy stakes are also real: device-identifier and location tracking is personal-data processing under GDPR and CCPA, which calls for anonymization, limited retention, and notice or consent where applicable.
For directional benchmarks: grocery active customers average 4–8 visits a month. Apparel and specialty land closer to 1–2 visits a month, with one-month repeat rates of roughly 20–40%.
How the Metrics Connect: The Full Sales Equation
Pull every metric into one line and the full funnel becomes arithmetic:
Sales = Passers-by × Capture Rate × Conversion Rate × Average Transaction Value
Each term is a distinct lever mapped to a specific metric — location and exposure, storefront appeal, selling effectiveness, pricing and mix. Push the decomposition further and ATV breaks into units per transaction × average selling price, telling you whether a bigger basket came from more items or pricier ones.
The equation becomes diagnostic when you treat percentage changes as roughly additive. A 10% sales decline splits into a traffic contribution, a conversion contribution, and an ATV contribution — each routing to a different part of the business. That's the difference between "sales are down" and "conversion dropped 6 points on Saturdays because we were understaffed."
One pitfall to respect: the levers aren't independent. A promotion can lift traffic while dragging conversion and ATV down, because it pulls in low-intent browsers. Without a full decomposition, the net effect stays ambiguous — you might celebrate a footfall bump that actually cost you margin.
Sources
- MRI Software — footfall counting definitions and sensor accuracy ranges
- Ariadne — capture rate formula and storefront conversion benchmarks
- TruRating — in-store conversion rate benchmarks by sector
- Network Solutions — e-commerce conversion rate averages
- Forasoft — core retail video-analytics metrics including dwell and queue length
- Contentsquare — brick-and-mortar share of US retail sales
- StoreTraffic — occupancy capacity ratios and compliance use cases
- Rebiz — traffic counting for labor scheduling and KPI denominators