What is a footfall proxy? A term retail REIT analysts should know

A footfall proxy is any dataset that stands in for foot traffic when you can't measure foot traffic directly. For retail REIT analysts, that's most of the time. Tenants report same-store sales quarterly, sometimes with a lag of weeks after the quarter closes. Landlords get even less, often just a percentage-rent reconciliation once a year. If you're modeling a center's health between those disclosures, you need something that moves more often than the filing calendar.

The term comes up a lot in sell-side notes now, usually attached to mobile location panels or credit card spend data. The concept itself is older than any particular data source, though: anything that correlates with store visits, foot traffic, or customer counts, and that you can observe without waiting on the tenant, counts as a proxy. A proxy is only as useful as its correlation to the two numbers that actually move a model, same-center sales and occupancy cost ratio, and that correlation varies a lot by property type and data source.

Why analysts reach for proxies in the first place

Retail REIT coverage runs on a disclosure lag that doesn't match how fast tenant performance shifts underneath it. A big-box anchor can see traffic soften for two straight months before that shows up in a 10-Q, and by the time it does, the stock has already repriced. Buy-side analysts who want to front-run that print need a read on store performance that updates weekly or daily, not quarterly.

Mobile location panels tried to solve this first, and they're still the most widely used foot-traffic proxy in the space. They have real limitations that matter for underwriting. Panel size varies by carrier and operating system, opt-in rates aren't disclosed consistently, and a center with a big elderly or low-smartphone-penetration customer base can show misleadingly thin traffic even when parking lots are full. Credit card panels have their own gap: they catch spend, not visits, so a browsing trip that doesn't convert never shows up.

Parking counts as a footfall proxy

A parking lot count sidesteps both of those gaps. It doesn't sample a subset of devices or cards, it counts the vehicles that are physically there. For a suburban strip center or a power center where most customers drive, the lot fill rate is a direct, unsampled read on store visits that day. It won't catch the walk-up or transit-dependent customer, which matters more in urban infill assets than in suburban ones, so the proxy's value varies by property type the same way mobile panels do.

What it adds that a device panel can't: a count you can audit against a photo. If the lot occupancy series says traffic dropped 15% week over week, you can pull the image and see whether that's a real falloff or a resurfacing project that closed half the spaces. That auditability is the thing most foot-traffic data products don't give you, because you're trusting a model's output with no way to check the underlying signal.

This is the use case Parking Lot Counts is built around: a daily occupancy read per site, delivered as a time series so you can chart it straight against a tenant's reported comps instead of waiting for the print to tell you what already happened.

What a proxy can't tell you

No footfall proxy, parking counts included, tells you basket size, conversion rate, or margin. A full lot on Black Friday tells you people showed up. It doesn't tell you what they bought or at what price. Analysts who get this wrong tend to treat a traffic proxy as a sales proxy and build a model with more precision than the input supports. The honest use is as a leading indicator retail sales directionally follow, a tripwire for quarters that are coming in better or worse than consensus, not a replacement for the comp number itself.

Used that way, a parking lot occupancy series earns its place next to the mobile panel and the card data in a tenant monitoring file, not instead of them. It's one more read on the same underlying question: is this center busier or quieter than it was last month, and does that match what the lease says it should be doing.

If you're building that monitoring file for a handful of centers, start with a site and see what a daily occupancy read looks like against the comps you already track.