Mall Analytics in Practice: The KPIs Centre Managers Actually Use

It is Monday morning, the weekly report says footfall was down against the same week last year, and the regional manager of your anchor tenant has already emailed to ask what the centre is doing about it. You have forty minutes before the operations meeting. This is the moment mall analytics either earns its place or gets ignored: not when the dashboard is demonstrated, but when a centre manager has to turn a column of numbers into an answer. This article covers the handful of KPIs experienced centre managers actually open, how they read them, and what they do when two numbers disagree. For the wider picture of sensors, platforms and rollout, see our guide to improving shopping mall analytics; here we stay with the weekly reading itself.

The four numbers a centre manager opens first

Most managers who use their data every week have settled on a short list, and it is shorter than vendors expect. First, total entries for the week against the same trading week last year, not the same calendar dates. A week that gains a Saturday and loses a Tuesday will show a swing that has nothing to do with performance. Second, entrance share: what percentage of visitors arrived through each door. A shift in share is often the first visible sign that a car park level closed, a bus route changed or a competitor opened on one side of the building. Third, the hourly curve for the busiest day, because this is what security, cleaning and parking rosters should be built on. Fourth, and almost always skipped, data completeness: how many sensor-hours were missing or flagged. A footfall drop in a week where one entrance sensor was offline for two days is not a footfall drop, and reading the number without checking completeness is how bad decisions get made.

Capture rate: the tenant KPI that replaces arguments

Raw store footfall tells a tenant very little, because it rises and falls with the centre. Capture rate separates the two: store entries divided by mall entries over the same period. If the centre recorded 100,000 entries in a week and a store recorded 6,000, its capture rate is 6%. These figures are illustrative assumptions, not benchmarks, but the logic holds at any scale. When the centre is down 5% and a store is down 5%, capture rate is flat and the store is simply riding the centre's tide. When the centre is flat and a store's capture rate has slipped from 6% to 5%, something changed at that store: a window display, opening hours, a staffing gap, a shift in the category mix nearby. That is a conversation worth having, and it is one you can only have with store-level counting alongside entrance counting. Remember that the sum of all store entries will always exceed mall entries, because one visitor enters several stores. That is not an error; it is the reason the ratio exists.

Reading mall foot traffic data against tenant sales

Footfall on its own tells you how many people came. Footfall next to declared sales tells you what they did. Two KPIs matter here. Conversion is transactions divided by store entries, and spend per visitor is sales divided by store entries. Both require tenant revenue to be in the same system as the counts, which is what VemTenant handles alongside VemCount: traffic and tenant sales run through one pipeline, so turnover rent, sales per square metre and conversion are calculated from matching periods rather than reconciled from a spreadsheet that arrived three weeks late. The reading is straightforward once the data lines up. A tenant with strong capture rate and weak conversion has an in-store execution issue. A tenant with weak capture rate and strong conversion is doing well with the people it gets and may have a location or signage problem. The distinction decides whether the right response is a marketing conversation, a leasing conversation or an operations conversation, and it is the distinction that makes mall foot traffic data commercially useful rather than merely interesting.

How to read mall analytics when the numbers disagree

Disagreement is normal and it is where most of the learning happens. Work through the usual causes in order before concluding anything about the market.

• Tenant sum versus mall total: tenant entries will always add up to more than building entries. If a tenant challenges the gap, show them their capture rate instead of the raw figure.
• Calendar effects: moving public holidays, school terms and late-night trading events shift footfall between weeks. Compare like-for-like trading weeks, and annotate the report so the explanation travels with the number.
• Entrance behaviour: a loading dock used as a staff shortcut, a cinema door open after retail closes or a car park corridor can all inflate or deflate a door. If an entrance behaves oddly, check what was agreed in the counting logic before assuming a sensor fault.
• Sensor conditions: Vemco commits to a contractual minimum of 96% accuracy and typically delivers 98 to 99% when lighting, placement and visitor behaviour allow. Revolving doors, strong backlighting from glass façades and a surge of pushchairs on a Saturday can push any sensor toward the lower end. A small weekly variance is noise; a sustained shift at one door deserves a manual spot count.
• Sales versus traffic moving in opposite directions: if footfall is up and sales are down, start with conversion by category, not with the centre marketing plan. One large tenant's late declaration can swing the whole figure.

Turning a weekly reading into a decision

The managers who get value from their data have a fixed set of actions tied to each KPI, so the meeting does not restart from scratch every week. A change in entrance share triggers a check on wayfinding, parking and external works. A change in the hourly curve triggers a roster review with the security and cleaning contractors. A capture rate decline at a tenant triggers a visit, not an email. A conversion decline across a category triggers a look at what changed in the mix around it. And a campaign is judged on whether total entries rose, not on whether one entrance got busier while another emptied, because entrance-level data makes that distinction visible. Give tenants their own login showing their store data and anonymised category benchmarks; a tenant who sees the same numbers you see stops arguing about the numbers and starts arguing about what to do, which is the argument you want. Peer-group benchmarking, meaning same category and comparable position and frontage, is what makes those comparisons fair enough to use in lease renewals.

Frequently asked questions

How often should a centre manager look at footfall data? A weekly review is the realistic rhythm for decisions about rosters, tenants and marketing. Daily figures are useful during events and campaigns, but reacting to a single day's variance usually means reacting to noise.

Which single KPI is most useful for tenant conversations? Capture rate, because it strips out the centre's own rise and fall and shows how a store is performing relative to the traffic available to it. Pair it with conversion once tenant sales are in the same system.

What should I do when a tenant disputes the counts? Agree a short manual verification window at their entrance and compare it with sensor output, documented so it can be repeated. Most disputes resolve once the tenant understands that store entries and mall entries measure different things.

If you want to see these KPIs built from your own entrance layout and tenant mix, including how VemCount and VemTenant present capture rate, conversion and entrance share in one report, contact Vemco Group for a centre-specific walkthrough.