Employee stalls vs. customer stalls: reading a parking lot count right

A raw vehicle count from a retail lot tells you how many cars were parked there on a given day. It does not tell you how many of those cars belonged to shoppers. If you're using lot occupancy as a footfall proxy ahead of a tenant's comp print, that distinction is the whole exercise. Get it wrong and you'll chase noise from shift changes instead of signal from foot traffic.

Why the raw count misleads you

Employee vehicles behave differently than customer vehicles, and that difference shows up in the data as a stubborn floor you have to subtract out before the number means anything.

Staff typically arrive before doors open and stay for an entire shift, sometimes two. A big-box location running a morning crew and an afternoon crew will show a near-constant employee presence across most of the business day, regardless of whether it's a slow Tuesday or a packed Saturday. Customer vehicles, by contrast, cycle. They arrive, park for forty minutes to two hours, and leave. That turnover is what correlates with transaction volume and, eventually, with the comp number the tenant reports.

If you take one count per day and treat the whole lot as "customers," you're baking in a constant that has nothing to do with sales. Worse, that constant can be large: a grocery anchor or a big-format discounter can easily have thirty to sixty staff vehicles parked at any one time, and that number barely moves week to week. Against a swing of a few hundred customer cars, it's not trivial. It's the floor your signal sits on.

Separating employee stalls from customer stalls without a staff badge list

You don't get a roster, and you don't get license plates. What you do get, from a repeat pass at sub-half-meter resolution, is a reasonably reliable stall-by-stall count on a given day. A few things to lean on:

The geography of the lot itself. Employee parking is almost never the row closest to the entrance. Store policy, loading-dock access, and plain habit push staff vehicles to the back rows, the overflow section, or the strip nearest a service door. If the imagery resolves individual stalls, you can often flag the back rows as the likely employee zone just by looking at where cars cluster on a quiet morning before the store opens to the public.

The baseline-floor method. Pull the lowest sustained count in your time series during the hours the location is open and compare it to peak counts on the same days. That floor, the number of cars that are there almost regardless of day or hour, is your best proxy for staff parking. Subtract it from the daily total and what's left is a closer read on customer traffic.

Day-of-week contrast. Staffing levels don't swing much from a Tuesday to a Saturday at most retail locations. Customer traffic does, often by a wide margin. If your weekday-to-weekend delta is small, you're probably still counting mostly employee cars. A seasoned retail analyst builds that delta into the model rather than treating one day's count as representative of anything.

Net customer parking methodology, in practice

Net customer parking is the daily total minus the estimated employee floor, indexed over time rather than reported as a one-off figure. The index matters more than any single day's absolute count, because it's the trend line, rising, flat, or falling against the same weeks last year, that tells you something before the tenant's own numbers come out.

A few caveats worth keeping in the model. Multi-tenant centers complicate the floor estimate, since you're counting everyone's staff in one lot. Big-box locations with seasonal hiring will shift the floor itself across the year, usually upward into the holiday stretch, so don't treat it as fixed. And for industrial sites, the analogous split is shift-worker parking versus visitor and delivery-driver parking, which follows the same logic but on a different clock.

None of this requires a plate reader or a badge feed. It requires a count you can trust, pulled on the same cadence, long enough to see the floor settle out. That's the series Parking Lot Counts delivers, a per-site occupancy read you can chart against a tenant's reported comps and start separating staff from shoppers for yourself. If you're already modeling footfall proxies for a retail or industrial name, it's worth seeing what a daily lot series looks like before the quarter closes.