Turning occupancy data into everyday operating decisions

A parking facility produces a constant stream of small facts: a car entered, a car left, a space was taken, a level filled. One at a time they are trivia. Gathered into counts by zone and time, they become one of the most useful operating inputs a parking team has, provided everyone understands what the numbers mean and how current they are.

This guide is for commercial parking operators who want occupancy data in everyday decisions rather than only in a quarterly report. It covers what the data is, how to read it, how to keep it honest and how to build a weekly habit around it.

What occupancy data actually is

At its simplest, occupancy is the number of spaces in a defined area that are in use at a moment in time. Each part of that sentence matters.

• A defined area. Occupancy is useful by zone: a level, a row group, a reserved block, an EV charging area, a visitor section. A site-wide number hides what drivers experience, such as a full ground floor beneath a half-empty roof. Zones should match how the site is used and managed, and stay stable so this month compares cleanly with last. Mapping a vehicle lot covers how to draw them.
• Spaces in use. Decide what counts. A vehicle across two bays, a motorcycle, a coned-off space, a van in a loading bay: each needs a written rule.
• A moment in time. Every count belongs to a timestamp. A count without one is a rumour.

Keep capacity beside every count. The same number can mean a full visitor area or a nearly empty upper level, and a percentage on its own hides how many spaces are actually left. Show both.

Arrivals, departures and dwell

Counts describe the state of the site. Flows explain how it got there and where it is heading.

• Arrivals and departures by zone and time show when pressure builds and eases. A level that fills steadily through the morning needs a different response from one that fills in a rush when a nearby event lets out.
• Net flow, arrivals minus departures over a short window, is often the earliest sign that a zone will fill. A modest count with strong inbound flow deserves more attention than a high count that is holding steady.
• Dwell is how long vehicles stay. In short-stay areas, loading bays and pick-up lanes, rising dwell is a reason to review the posted rules, layout and operating pattern. In long-stay areas, dwell shows the mix of short, routine and multi-day stays.

Treat dwell as a distribution, not an average. A few vehicles parked for a week drag the average far from what most visitors experience. Look at the typical stay alongside the long tail.

Why freshness matters as much as the number

Every occupancy figure has an age. A count observed a moment ago and a count from before a camera went offline can look identical on a screen, and they mean completely different things.

The honest rule is simple: show the latest observation together with when it was made, and when a zone has not been observed recently, say so. A zone that reads “awaiting observation” or “stale” is less satisfying than a confident number, but it is truthful, and it prompts the right response: check the camera, check the network, look in person.

Never fill a gap with old data. Carrying the last count forward, smoothing over a missing period or borrowing yesterday’s value for the same hour all produce numbers that look reasonable and are wrong exactly when it matters: an unexpected surge, an incident, an event that changed the pattern. Every decision downstream inherits the error silently. A gap shown as a gap can be explained and planned around. A gap papered over becomes a false fact.

Agree on freshness thresholds per use. The age that makes a count too old for a driver-facing sign is very different from what is acceptable in a weekly planning chart. Whatever the threshold, the display should change visibly when it is crossed.

Using patterns to run the site

Once counts and flows are trustworthy and their age is visible, a few weeks of history reveal the site’s rhythm. This is where that rhythm earns its keep.

Staffing

Schedule attendants, ambassadors and cleaning crews against when arrivals actually peak in each zone, not a fixed roster. Arrival curves also show when a shift change lands in the middle of a rush.

Signage and wayfinding

If drivers circle a full ground floor while upper levels have room, the problem is information, not capacity. Use zone occupancy to decide where directional signs point at different times and, where the site has dynamic signs, which threshold switches them. Drive live signs only from fresh data; a stale count on a sign sends drivers the wrong way.

Space allocation

Reserved blocks, visitor areas, EV charging and accessible spaces drift out of step with demand. Occupancy by zone shows which sit empty while neighbours overflow, giving you evidence for a conversation with tenants or management instead of an impression.

Overflow plans

Write down what happens when a zone fills: which zone takes the overflow, who changes the signs, who needs to know. Then use arrivals and net flow to trigger the plan before the zone is full, not after drivers are already queuing.

Maintenance windows

Restriping, sweeping, lighting repairs and pressure washing all close areas. Historical occupancy shows the quietest windows by zone, so closures disturb the fewest drivers. Mark closed spaces so capacity drops with them, rather than reading as a sudden fall in demand.

Matching decisions to data

Different decisions need different data, and different tolerance for age. Making that explicit keeps live and historical views from being mixed up.

A weekly occupancy review

Patterns change decisions only when someone looks at them regularly. A short weekly review with the operations lead and whoever manages staffing and signage works well. Keep the agenda fixed:

1. Coverage first. Which zones had gaps in observation, and why? Resolve camera, network or mapping issues before reading any trend.
2. Peaks and pressure. Which zones reached capacity, when, and did the overflow plan trigger in time?
3. Turnover. Did dwell in short-stay areas, loading bays and pick-up lanes behave as intended?
4. Allocations. Are any reserved, visitor or EV areas consistently under- or over-used?
5. Unusual days. Which days (events, weather, closures) should be tagged so they do not distort comparisons?
6. One change. Agree on a single adjustment to staffing, signage, allocation or the overflow plan, and note what you expect to see next week.

Keep a running log of each change and what followed. Over a season, that log becomes the site’s operating memory.

Questions to ask before trusting a number

Run through these before any count drives a decision, especially one that is new or surprising.

• When was this observed, and is that recent enough for this decision?
• Which zone does it cover, and has that boundary changed recently?
• What capacity is it measured against, and does that reflect closed or reserved spaces?
• How were awkward cases counted: vehicles across lines, motorcycles, trailers?
• Were any cameras offline or obstructed during the period?
• Is this a typical day, or one shaped by an event, the weather or a closure?
• Does it match what someone on the ground sees? A short walk is the best calibration there is.

Common pitfalls

• Site-wide totals. One number for the whole facility hides the zone-level situation drivers experience.
• Silent gap-filling. Carrying old counts forward or interpolating missing periods turns outages into fiction.
• Averages without spread. Mean dwell and mean occupancy hide the peaks and long tails that matter most.
• Comparing changed zones. Redrawing a zone breaks comparisons with earlier weeks unless the change is recorded.
• Ignoring capacity changes. Closures and coned-off spaces look like falling demand if capacity is not updated.
• Data without a decision. If nobody can name the decision a chart supports, it does not belong on the weekly agenda.

ParkSense presents occupancy by zone with the time of each observation, and shows a zone as awaiting observation rather than guessing when it has not been seen recently. You can see that behaviour in the interactive product tour, read about the commercial parking solution, or contact us to talk through your own site.