If you need real numbers for how full your office actually is, but you cannot deploy cameras and will not surveil people through their badges, you are asking a specific question: how to measure office occupancy without cameras. The short answer is anonymous people counting with on-device sensors, small edge devices that count bodies in a space, per zone, per minute, and discard everything else. No video is stored, no faces are processed, and no one is identified. This guide walks through the methods on the table, where each one falls short, and how a privacy-first sensor produces office occupancy data you can act on.
01 · HOW CAN YOU MEASURE OFFICE OCCUPANCY WITHOUT CAMERAS OR BADGES?How can you measure office occupancy without cameras or badges?
You measure office occupancy without cameras by deploying anonymous occupancy sensors that count the number of people present in a zone and report a count, not an image, not an identity. The sensor runs inference on-device, emits a number (for example, "Zone 3: 14 people, 10:42am"), and never transmits or stores video. That number is the occupancy figure. You get the same fact a camera-and-analyst pipeline would give you (how full the space is) without ever creating a record that points at an individual.
This matters because the question is usually forced by a constraint, not a preference. A workplace lead who has already ruled out cameras has ruled them out for a reason: works councils, a privacy review, a lease clause, staff trust, or simple common sense. The good news is that the constraint and the goal are not in tension. Occupancy is a count. A count does not require a camera feed or a name.
02 · WHAT METHODS EXIST FOR MEASURING OFFICE OCCUPANCY, AND WHERE EACH FALLS SHORT?What methods exist for measuring office occupancy, and where each falls short?
There are five common methods. Most organisations have tried at least one and found it wanting. Here is the honest comparison.
| Method | What it measures | Where it falls short |
|---|---|---|
| Badge swipes | Building/floor entries | Counts entry, not presence; misses tailgating; HR-adjacent and identifiable |
| Room/desk booking data | Intent to use a space | People book and don't show; walk-ups never appear; no-shows inflate it |
| Wi-Fi / device association | Connected devices | Counts phones and laptops, not people; misses guests; varies by BYOD policy |
| Manual walk-arounds | A snapshot count | Expensive, infrequent, and stale the moment it's recorded |
| Cameras + analytics | Actual presence | Stores video, raises privacy and consent problems, often the ruled-out option |
The pattern is that the easy methods measure a proxy, an entry, a booking, a device, and the accurate method (cameras) carries a privacy cost most offices won't pay. Badge logs are the most tempting because they already exist, but a swipe at 9:01am says nothing about whether anyone was at their desk at 2:30pm, and it can always be back-tracked to a person.
03 · HOW DO YOU MEASURE OFFICE OCCUPANCY WITH A PRIVACY-FIRST SENSOR?How do you measure office occupancy with a privacy-first sensor?
You mount one small sensor per zone, it counts people on-device, and it reports anonymous counts to a dashboard. The mechanics are deliberately boring, which is the point. Each workspace occupancy sensor runs the counting model locally on an edge processor. The raw signal (whatever the sensor's optics or radar captures) is processed in memory and immediately discarded. What leaves the device is a number and a timestamp.
The deployment looks like this:
- One floor, six to eight zones. Open-plan areas, meeting-room banks, the social space, the kitchen.
- One sensor per zone, emitting anonymous counts on a short interval.
- No faces, no badges, no per-person tracking. The architecture never produces an identifier, so there is nothing to leak, subpoena, or misuse.
This is the difference between privacy-as-policy and privacy-by-design. A camera system that "deletes footage after 30 days" still created identifiable footage first. You are trusting a setting. A system that runs inference on the edge and emits only counts never created the identifiable data in the first place. We think the architecture has to come first, because a privacy promise bolted onto a camera is only as good as its configuration.
The count leaves the sensor. The image never does. That single design decision is what makes occupancy measurement compatible with a privacy review.
04 · WHAT OFFICE OCCUPANCY DATA CAN YOU COLLECT WITHOUT IDENTIFYING ANYONE?What office occupancy data can you collect without identifying anyone?
Quite a lot, as it turns out. Anonymous counts, aggregated over time and across zones, produce every operational metric a workplace or property team actually uses:
From those counts you can derive anchor-day curves (Tuesdays and Wednesdays peak; Mondays and Fridays don't), catering-window pressure in the kitchen between 11:45 and 12:45, and which meeting rooms run at 11% utilisation and aren't earning their footprint. None of that requires knowing who anyone is. The unit of analysis is the space, not the person.
05 · DOES MEASURING OCCUPANCY WITHOUT CAMERAS COMPLY WITH THE AUSTRALIAN PRIVACY ACT?Does measuring occupancy without cameras comply with the Australian Privacy Act?
Anonymous occupancy counting sits well within the Australian Privacy Act because it does not collect personal information. The Act and its Australian Privacy Principles govern information about an identified or reasonably identifiable individual. An anonymous count ("14 people in Zone 3") is not about any identifiable person, so the data-handling obligations that make camera and badge systems heavy simply don't attach in the same way.
This is not legal advice, and your own privacy officer should review any deployment. But the architectural fact is straightforward and useful in a procurement review: if the system never produces an identifier, there is no personal information to govern, store, or breach. That is a far easier conversation than explaining footage retention, access controls, and consent signage for a camera network. It's also why occupancy counting tends to clear works-council and staff-consultation hurdles that camera proposals stall on.
06 · HOW OFTEN SHOULD YOU MEASURE OFFICE OCCUPANCY?How often should you measure office occupancy?
Continuously, at a fine interval, for at least four weeks before you draw conclusions. Occupancy is not a single number: it's a distribution across the day and the week, and hybrid work has widened that distribution. A one-off walk-around count or a single "busy day" snapshot will mislead you, because the variance between a Wednesday peak and a Friday trough is the entire story.
A four-week trailing window is the smallest sample that separates signal from noise: it captures the weekly rhythm, smooths out one-off events (a town hall, a public holiday), and gives you a defensible baseline before you change a schedule, a fit-out, or a lease position. After that, continuous measurement keeps the baseline honest as patterns drift.
07 · HOW DO YOU GET RELIABLE HYBRID-OFFICE OCCUPANCY DATA?How do you get reliable hybrid-office occupancy data?
Reliable hybrid-office occupancy data comes from measuring presence directly and continuously, not from inferring it from bookings or swipes. Hybrid work broke every proxy method at once: badge swipes fragmented across the week, room bookings decoupled from actual attendance, and Wi-Fi counts got noisier as people carried more devices. The only method that survived the shift is counting people where they actually are.
Reliability then comes down to three things: count the right thing (people, per zone), count it honestly (we publish our view on counting accuracy and ground truth rather than claim a perfect number), and feed the result into systems your team already uses. Occivar™ integrates occupancy data into the BMS, BI, and reporting tools you run today, so the number lands where decisions get made. The same approach scales beyond corporate floors. It's how universities measure space against real demand.
The result is a number a property team can defend in front of a CFO: not "who came in", but "how the floor is actually used", measured directly, without a camera or a badge log in sight.
If your office occupancy number is still a badge swipe, a booking report, or a guess, book a demo and we'll show you what anonymous, camera-free measurement looks like on your own floor plan.