A shopping mall and the stores inside it count traffic for different reasons, and their numbers rarely agree. The mall wants to know how many people are in the building, where they move, and whether an event changed anything. A tenant wants to know how many people walked past its window and how many came through the door. Mixing those two datasets — or paying for the wrong one — is one of the more expensive measurement mistakes in retail property.
Two owners, two datasets
Mall-level counting covers entrances, corridors, the atrium, vertical circulation, and parking. The owner is the property manager, and the purposes are leasing, marketing, event evaluation, and the footfall trends reported to tenants.
Store-level counting covers storefront pass-by and store entry. The owner is the tenant, and the purposes are conversion rate, staffing, and layout decisions.
Two properties follow from that split:
- Mall footfall is not store footfall. Summing tenant entry counts will not reproduce the mall's entrance count. Shoppers enter shops repeatedly, staff cross thresholds all day, and a large share of people in the building never passes a counted store door at all.
- Neither number is wrong. They answer different questions, and a chain that benchmarks its stores against a mall-level figure is comparing a window to a building.
What mall operators measure in common areas
Entrances and parking. The building total and its arrival curve. This is the denominator for every other mall metric, including footfall per square metre and tenant performance ratios.
The atrium and event spaces. Dwell during an activation compared against normal days. This is how event return gets argued without anecdote: same space, same measurement method, before and after.
Corridors and circulation. Flow volume and direction along main paths, which exposes chokepoints, dead ends, and corridors that shoppers walk one way but never back. Leasing adjacency decisions — what should sit next to an anchor — are usually built on this layer.
Anchor-to-anchor paths. Where shoppers travel between major tenants, and which secondary units sit on the natural route rather than off it.
What tenants measure at the storefront
Pass-by versus entry. Described below, and the single most under-used mall-retail metric.
Entry counts and conversion. Entries divided into transactions, which is the number a store can actually manage day to day.
Staff exclusion. Staff crossing the storefront line during a shift will inflate counts badly — in a small store, potentially by a large fraction of reported traffic. Any storefront counting configuration needs a deliberate answer for this, whether exclusion zones, shift-aware rules, or a device that does not maintain long-term identity.
Threshold coverage. A storefront wider than the camera's field of view loses entries at the edges — usually the edges nearest the corridor, which are exactly where unplanned entries happen. Wide openings need either more than one counting device or a wider-view device mounted to cover the full opening.
Pass-by and capture rate
Pass-by is the number of people who crossed in front of the store — the traffic that had the opportunity to enter. Capture rate is entries divided by pass-by.
Most tenants can measure entries and cannot measure pass-by, which leaves them unable to separate two very different stories: "mall traffic fell, our numbers fell with it" versus "mall traffic held, our window stopped working." A corridor count from the property, even a coarse one, plus the tenant's own door count, gives a workable proxy that is far better than no denominator at all.
Capture rate also matters in lease discussions, because it is one of the few store-level metrics that reflects the unit's position and the storefront design rather than the operator's marketing.
Placement and configuration notes for mall environments
- Very high ceilings. An atrium may put several metres between the camera and the people being counted, which favours depth sensing or a lens and mounting angle matched to the actual height. Test at the real height, not in a showroom.
- Reflective and glass surfaces. Polished floors and glass balustrades create reflections that some vision systems read as extra people. Where a storefront line sits in front of a mirrored surface, verify before commissioning.
- Dense, stationary crowds. Event programming fills the atrium with people who stand still, which is the hardest condition for movement-based detection and the strongest argument for depth-based sensors in event spaces.
- Cross-traffic geometry. In corridors, mount overhead with the counting line perpendicular to the direction of travel. A line drawn at an angle will register revisits as new entries.
- Escalators and stairs. Direction of travel has to be handled explicitly; a device that counts only people crossing a plane will produce a total, not a net flow.
Sharing data between landlord and tenant
When a mall installs counting infrastructure, the questions to settle before installation are commercial rather than technical:
- Who owns the data the system produces, and can figures for a named tenant be released to that tenant or to others?
- What granularity gets shared — building total only, corridor counts, or storefront-level? What format, and how often?
- Who pays for the tenant's own storefront counting, and does it have to be the same vendor as the mall's?
A common arrangement is that the property reports aggregate building or corridor footfall on a schedule, while store-level data belongs to the tenant. Where the property shares anything resembling store-level movement data, notice obligations and privacy rules apply, and those need confirmation from a qualified legal adviser in the relevant jurisdiction rather than a vendor's assumption.
Standards also matter for comparability. If two systems in the same building define "an entry" differently — one counting staff, one not; one counting children, one not — the two datasets will never reconcile, and every subsequent argument about performance will be unfounded.
How each side uses the numbers
Leasing. Evidence for rent and percentage-rent discussions, tenant-mix decisions, and where to place categories so that complementary stores sit on the same natural path rather than at opposite ends of the building.
Marketing. Event and campaign evaluation, day-of-week and weather-adjusted comparisons, and the ability to say honestly whether a promotion moved traffic or merely moved people around inside the same building.
Operations. Cleaning, security, and lift staffing scheduled against the real arrival curve; escalator and entrance load known before it becomes a complaint.
Tenants. Hourly staffing curves, storefront and display changes measured rather than asserted, and chain-wide comparisons across units in different malls using one consistent method.
Limits worth stating up front
Counting in a public mall is more politically sensitive than counting inside one store. Visible signage, anonymous aggregation, short retention, and no attempt at identification are the practical defaults; specific obligations vary by market and require local legal review.
Numbers will also never match perfectly across systems, vendors, and mounting positions. Differences of a few percent between two independently installed counters on the same doorway are normal. Anyone presenting mall footfall as an exact figure — or using a small difference to settle a large argument — is over-claiming what the technology does.
FAQ
Why do mall footfall numbers never equal the sum of store counts? Because they measure different populations. Mall totals count people entering the building once; store counts count entries into units, so one shopper visiting six stores generates one mall count and six store counts. Staff movement, repeated visits, and people who never pass a counted door widen the gap further.
What is the difference between atrium traffic and storefront traffic? Atrium traffic is passing and browsing traffic in common space; nobody has committed to entering a specific store. Storefront traffic is traffic that has arrived at a specific unit's frontage and thus has a choice to make. The first describes the building's draw, the second describes one store's opportunity.
How many cameras does a mall need? There is no general answer, because it depends on the number of entrances, corridor geometry, ceiling heights, and whether you need flow direction as well as volume. A useful approach is to specify the questions first — building total, corridor flow, event-space dwell, storefront lines — and size coverage against those, rather than starting from a camera count.
Is it legal to count people in a shopping mall? Anonymous aggregate counting in common areas is widely practised, but notice and retention obligations differ by country and by whether any personal data is involved. Thailand, Singapore, Vietnam, Malaysia, and Indonesia each have their own regimes, and the applicable requirements for a specific property should be confirmed with a qualified legal adviser.
---
Mall counting and store counting are two layers of one picture, and they only become useful when each side knows which layer it owns. The solutions overview describes how counting and zone analytics fit mall and retail-property deployments, and the product center lists the camera families built for overhead and wide-area coverage. For the metrics that sit on top of this data, see the blog. If you are planning coverage for a property or a chain, contact our team with the floor plans.