Shopper Insights Without a Survey: What Footfall, Dwell Time and Conversion Reveal

A category manager runs a two-week promotion on a new end-cap. Sales lift 11 percent. The post-campaign survey, sent to 400 loyalty members, says 68 percent "noticed the display" and 54 percent found it "very appealing". Nobody in the room can say whether the lift came from the display, the weather, or the fact that the store had 9 percent more visitors that fortnight. The survey measured opinions about a thing. It did not measure the thing.

That gap is where measured behaviour earns its budget. Entrance counts, zone dwell time, path data and conversion against POS answer the operational questions a survey cannot: how many people were actually in the store, where they went, how long they stayed, and what share of them bought. This article walks through what each of those metrics genuinely reveals, where they mislead, and how to combine them into shopper insights you can defend in a trading meeting.

Why stated preference keeps failing retail decisions

Surveys and loyalty data share three structural problems. They sample only the people willing to answer or willing to sign up, which skews towards engaged, frequent shoppers. They capture recall rather than action, and recall after the fact is generous. And they say nothing about the visitor who walked in, looked at the new fixture for four seconds and left. In most stores that visitor is the majority, and the majority is where the growth is.

Shopper behaviour data collected passively has the opposite profile. It covers everyone who crosses the threshold, it records what happened rather than what was remembered, and it is continuous, so you can see the Tuesday afternoon dip as clearly as the Saturday peak. It also has limits, which we will come to. But the starting point is the right one: count the whole population first, then ask questions of the sample.

Footfall: the denominator everything else depends on

Entrance counting is the least glamorous metric and the most important, because every ratio you care about sits on top of it. Conversion is transactions divided by visitors. Sales per visitor is revenue divided by visitors. Staff-to-traffic ratio is rota hours divided by visitors. Get the denominator wrong by 10 percent and every downstream number is wrong by 10 percent.

This is why accuracy specifications matter more than they look. Vemco's contractual minimum is 96 percent, with 98 to 99 percent typical where lighting, layout and visitor behaviour allow. The honest version of that sentence is the useful one: a glass-fronted entrance with strong backlighting, or a doorway where trolleys and pushchairs bunch up, will sit nearer the floor than the ceiling, and you should know which of your sites those are before you compare them.

Returning to the end-cap example: with reliable counts, the 11 percent sales lift against a 9 percent traffic rise becomes a 2 percent improvement in sales per visitor. Still positive, but a very different story, and a very different decision about whether to roll the display out to forty stores.

Dwell time: reading in-store customer behaviour by zone

Dwell time is where retail insights become specific to merchandising. Measured per zone rather than per store, it tells you how long people spend in front of a category, a fixture or a service point. The interpretation depends entirely on context, which is why dwell time is dangerous as a headline number and powerful as a comparison.

• Long dwell, high conversion: the zone is doing its job. Protect it and study why.
• Long dwell, low conversion: shoppers are interested but blocked. Common causes are confusing range architecture, missing sizes, or a self-service point that needs staff.
• Short dwell, high conversion: a destination category. People know what they want. Do not waste prime adjacency space trying to slow them down.
• Short dwell, low conversion: either the zone is a corridor, or the proposition is invisible. Check the path map before you blame the product.

One practitioner observation from implementations: before trusting any dwell figure, check where the zone boundary falls relative to the queue. A till queue that spills into a neighbouring category will inflate that category's dwell time every Saturday and make a mediocre fixture look like a hero. Draw the zone to exclude the queue path, or measure the queue as its own zone and subtract it.

Conversion against POS: where customer behavior analytics meets the P&L

Linking visitor counts to transaction data by hour is the step that turns a traffic report into something finance will read. Conversion by hour exposes the patterns that store managers feel but cannot evidence: the lunchtime window where traffic rises and conversion falls because two of four tills are closed, or the late-afternoon slot where fewer people come in but a higher share of them buy.

Daells Bolighus used exactly this combination of footfall and POS to understand how traffic and purchasing moved together across its furniture and home stores, so that staffing and promotional timing could follow measured demand rather than assumption. Magasin, across its department stores, works with the same principle at a larger scale: counts by entrance and floor set against sales give a picture of which areas convert the traffic they receive and which simply receive it.

The discipline here is to hold conversion and traffic apart when reporting. A store whose sales fell 5 percent while footfall fell 12 percent improved its conversion. That store manager did something right and should be asked what. Reporting sales alone would have put them on a performance plan.

Demographics without identification

Age and gender estimates add a layer that loyalty cards never can, because loyalty cards only describe members. Knowing that a children's wear brand such as Luksusbaby draws a visitor mix that shifts by time of day lets range placement and staffing reflect who is actually in the store rather than who signed up to the newsletter. These are aggregate estimates, not identities: no face is stored, no persistent identifier is created, and no individual is tracked between visits. For a shopper insights UK programme operating under UK GDPR, that distinction is the one your DPO will ask about first.

The same metrics at mall level

Mall marketing teams face the footfall problem in reverse: they generate the traffic but do not own the tills. Entrance and zone counts across the centre show whether an event moved people past the tenants who paid for it, and path data shows whether the new food court actually feeds the upper floor or just the car park. Odense Municipality applies the same measurement to a city centre, counting pedestrian flow to evidence how events and public realm changes affect footfall for the retailers along those streets. The logic is identical to a single store: count the population, then judge the intervention.

Where surveys still belong

Behaviour data tells you what happened. It does not tell you why a visitor left a zone after four seconds. The efficient sequence is to let measured data locate the problem, then spend a small, targeted survey or a shift of floor observation on the why. That is cheaper, faster and more honest than surveying four hundred loyalty members about a display most visitors never saw.

Frequently asked questions

What are shopper insights? Shopper insights are conclusions about how people behave in a retail space that can be acted on commercially: how many visit, where they go, how long they stay and what share buy. The most reliable ones come from measured behaviour rather than stated opinion, because they cover every visitor and record action rather than recall.

How do you get shopper insights without surveys or loyalty cards? Sensors at entrances and over key zones count visitors, measure dwell time and map paths, and that traffic is matched against hourly POS data to produce conversion. No participation from the shopper is required, so the data covers the whole population rather than a self-selected sample.

What does dwell time tell you about shoppers? On its own, very little; combined with conversion for the same zone, a great deal. Long dwell with low conversion usually signals interest that is being blocked by range, availability or service, while short dwell with high conversion indicates a destination category that should not be slowed down.

Can footfall data show which displays work? Yes, provided you measure the zone around the display and compare before and after with traffic as the denominator. A sales lift that simply tracks a rise in store traffic is not evidence the display worked; a rise in zone dwell and sales per visitor is.

Is behaviour measurement GDPR compliant? Counting, dwell and path measurement as described here produce aggregate data with no identification and no persistent identifier, so no individual is tracked or recognised between visits. Demographic figures are age and gender estimates at group level, which is the basis on which these systems operate under GDPR and UK GDPR.

If you have a promotion, layout change or event whose results you cannot separate from background traffic, talk to Vemco about measuring footfall, dwell time and conversion in the zones that matter. Contact the team at vemcogroup.com/contact-us to discuss a measurement plan for your stores or centre.