Stations and airports - which retail space is worth its rent
Rents in stations are negotiated on location, not on what the location delivers. The landlord says tens of thousands pass by every day. The tenant says little of that reaches the shop. Both are right - and neither can prove it.
A station, an airport or a large interchange is two things at once: a transport facility and a place to shop. The transport side is covered in Passenger counting in public transport, including platform thresholds and cleaning by usage. This article takes the other side: the spaces between the platforms that pay rent, and the question of which of them are worth it.
Not all footfall is the same
The figure in every leasing brochure is the number of people who use the station each day. For a single retail unit it says little. A kiosk by the stairs to the main platform and one at the end of a side passage share the same station footfall and have completely different businesses.
The difference becomes measurable once each retail unit is given a zone of its own. Then there are three figures that mean the same thing to landlord and tenant:
| Metric | What it measures | What it means in a negotiation |
|---|---|---|
| Capture rate | Visits to the shop zone divided by visitors to the site | What share of the flow the space actually draws in |
| Dwell time | How long someone stays in the zone | Whether the space is a detour or a destination |
| Engagement | Share of visits that last longer than walking past | Whether attention turns into interest |
How these metrics are calculated, why the capture rate can exceed 100 per cent, and what sample a comparison needs, is covered in Measuring dwell time. Here the question is what they mean in a station.
A figure both sides accept
The value of measured footfall per retail unit isn't that one side wins the argument. It's that both sides see the same figure.
- For the landlord, it shows which spaces benefit from the passenger flow and which don't. That is the basis for a rent that reflects the location - and for knowing whether a weak unit has a tenant problem or a location problem.
- For the tenant, the feeling that little gets through becomes a number. If the capture rate falls while station footfall holds, it's the business. If both fall, it's the station.
- For both, a change becomes measurable: new signage, a moved staircase, an altered entrance. Before and after, zone by zone.
That only works if the figure is transparent to both sides. A comparison between shop zones therefore needs a minimum sample and a clean “no comparison period” state, rather than setting a zone with three visits against one with three thousand.
Which zone feeds which
In a station, customers come from routes. Someone walking from the platform to the exit passes different spaces from someone changing trains. Zone sequences show which transit zones and which retail zones belong together: where did most of the people entering the bakery come from?
That isn't a movement track on a floor plan but a sequence of zones - and that is exactly enough for the question that matters. If a shop sits on a route hardly anyone takes, a better display won't help; a different entrance or a different offer will. How zone sequences are read inside a shop is covered in The visitors who don't buy.
Marketing a vacancy with a profile
An empty unit in a station is usually advertised with the station's footfall. More convincing is a profile of the people who pass that particular unit: at which hours, on which weekdays, and how many of them stop. A café needs the morning commuter flow; a fashion store is more about Saturdays.
These profiles are collected anonymously and without images. They describe the flow, not individuals. Where passing by ends and stopping begins is explained in Measuring dwell time.
Demand profiles for fares and promotions
The same counting series answer a question that has nothing to do with rent but is often asked in the same building: did a fare change, a promotion or a new service shift demand?
That takes three things:
- Profiles at the right resolution. Hour, weekday and season per site. A shift from the peak hour into the shoulder doesn't show up in the daily total.
- A clean comparison of two periods. Period A against period B with daily values and a block that shows whether the difference goes beyond normal variation.
- The calendar alongside. Public holidays, school holidays and your own dates, so a holiday week isn't read as the effect of a promotion. How context sits next to the figure is covered in Context: weather, holidays and staffing.
A comparison like this can be exported - for boards and transport authorities that want to follow a decision without digging into the data themselves.
The airport: the same question with a security checkpoint
At an airport there's an extra bottleneck every passenger knows: security. Whoever is waiting before the checkpoint isn't shopping after it. Every minute lost there is missing from the retail space beyond.
So both sides belong in the same measurement: the security lanes with people queuing and being served, and the spaces behind them with their footfall. How a checkpoint is steered live - throughput per lane, the worst lane as status, a board that lets passengers spread themselves - is covered in Queues in operation.
Two years of data instead of a counting week
The biggest decisions in a station are investments: an extra entrance, a longer passage, a new row of shops. They need demand data over years, not a counting week. When vehicle boardings and station counts sit in the same database, on a one-minute grid and over two years, an application rests on a series with seasons and a previous year.
How stations and lines are compared fairly, with median, percentiles and coverage, is covered in Comparing sites.
What belongs in the tender
- Can each retail unit be run as a zone of its own, with capture rate, dwell time and engagement?
- Is there a minimum sample for zone comparisons, and how does the platform show a zone without a comparison period?
- Does the analysis show zone sequences between transit and retail zones?
- Does the comparison of two periods show whether a difference goes beyond normal variation?
- Are vehicle and station data in the same analysis, and for how long?
How we do it
In the ANALYSIT Counting System, every retail unit in a station becomes a zone: capture rate as zone visits divided by site visitors, average dwell time, engagement and visits, with comparison to the previous period and sample protection. Zone-to-zone sequences show which transit and retail zones belong together.
For fare and promotion decisions there are counting series at any resolution with peak call-outs, weekday and seasonal patterns per zone and the campaign comparison of period A against period B with daily values and a significance block, plus a calendar with public holidays, school holidays and your own dates. Vehicle boardings and station counts sit in the same database on a one-minute grid for 730 days, with a cluster benchmark across lines or stations and export as CSV and XLSX.
At the security checkpoint, ACS shows people queuing and being served for each lane, the site status is the worst lane, and the public waiting-time board without login directs passengers to the shorter lane with “Best Choice”.
If you want to know which space in your station is worth its rent, talk to us. Four weeks of measurement show where the flow goes past.