Workplace occupancy: measure how your offices are really used
You are paying for every square meter, every day, whether anyone sits there or not. Most workplace decisions are still made on badge swipes, booking systems and the impression of whoever walked the floor on a Tuesday. We help you measure workplace occupancy with counts that are anonymous by design, so that you can decide about floors, desks and meeting rooms with evidence instead of anecdotes, without turning your office into a surveillance project.
Worried about how employees will react? That objection is the first thing we deal with, not the last. See the privacy section below.
The problems occupancy data is bought to solve
Nobody buys occupancy monitoring because it is interesting. They buy it because a decision is due and the data to make it does not exist. These are the situations we hear most often.
Desks and rooms you pay for and nobody uses
Rows of desks assigned to people who come in a couple of days a week. Meeting rooms booked all day and empty half of it. The lease is paid on the floor area, not on the part of it that gets used.
Space decisions made on impressions
Renew the lease or give back a floor? Move from assigned desks to shared ones, and at what ratio? Today the answer often comes from a manual walk round, a badge report that only shows the front door, or the loudest opinion in the room.
Cost per square meter that nobody can defend
Finance asks what the office costs per person actually present. Facilities can give the cost per square meter, HR can give the headcount, and the number that connects the two does not exist.
The hybrid return that is not a flat line
Attendance is not evenly spread. Tuesday and Wednesday overflow, Friday is empty, and the same building is under pressure and underused in the same week. Averages hide exactly the peaks you need to plan for.
Cleaning and services on a fixed schedule
Floors, rooms and washrooms are serviced on a calendar rather than on use. Areas that were empty all week get the same attention as areas that were full every day.
Three questions to answer before you buy anything
These questions decide whether the project is worth running at all, and they decide which technology fits. Answer them before you look at any product.
What decision will this data change?
Write the decision down first: give back a floor, change the desk ratio, resize the meeting rooms, move the cleaning budget. A hypothetical example: a company on three floors with a lease break in nine months, deciding whether to hand one of them back. That is a decision with a date, and it already tells you which floors to measure and for how long. If no decision depends on the numbers, measuring will produce a dashboard that people look at twice. The strongest occupancy projects have a date and a decision attached to them.
Do you need counts, or do you need identities?
Almost every real workplace question can be answered by counting: how many people were in this zone, at this hour, on this day. Knowing which named person sat at which desk is a different project, legally, technically and culturally, and it is rarely needed. Our advice is to design for counting, and to be explicit that identification is out of scope.
What infrastructure already exists?
If the building already has CCTV cameras with usable angles, measuring occupancy can often ride on hardware you already own. If there are no cameras, or if cameras are not acceptable in that space, sensors are the alternative. The answer changes the budget more than the choice of vendor does.
How we measure occupancy
We come at this from two sides. We are a technology provider, with our own products, the Sentinel video tracker and the Bluetooth Beacon Tracker, and we are a consultant on indoor geolocation projects, which is how we can say out loud that no single sensor answers every question in a building. Where the answer is a third-party technology, RFID, UWB or Wi-Fi positioning, we supply and integrate it as well: we do not build those, we do provide them. A marketing company that analyzes how people move through spaces is one of our clients. These are the options we compare, and what each one is honestly good for.
Video counting on existing cameras
Our Sentinel video tracker works with the CCTV cameras already installed, counts with 98% counting accuracy and reaches 20 cm spatial precision. We still measure the count against a manual count in your building during the pilot, because a number from someone else's building is not a number you can plan with. For occupancy work it is configured to count people and measure flows and dwell time per zone, not to recognize anyone. It suits open plan areas, entrances, canteens and common spaces where a camera already looks at the right part of the room.
Room and desk sensors
Where cameras are not present or not appropriate, for example inside meeting rooms, a sensor in the space reports whether it is occupied and how many people are in it. This is what most buyers mean when they search for a workplace occupancy sensor. Room and desk sensors are a third-party category rather than something we built. We assess them, specify them against the alternatives and manage the project end to end. For this category, which parts we supply directly and which come through a partner depends on your site and your country, and we confirm that on the discovery call.
BLE for the equipment, not the people
Our Bluetooth Beacon Tracker locates tagged objects with 1 to 3 meter accuracy in 90% of cases, with nodes that detect beacons up to 30 meters away, mounted on ceilings up to 7 meters, updating in the order of seconds. In an office we use it for shared equipment: AV trolleys, test devices, laptops from a pool, spare monitors. We do not recommend putting beacons on employees to measure space utilization: counting answers the question without the tag and without the argument.
Your data, in your tools
Occupancy data is only useful next to the lease, the headcount and the booking system. Our BLE platform exposes a REST API, push callbacks and daily CSV files, so the numbers can go into your BI tool, your facilities platform or a simple spreadsheet, cloud by default and on-premise where your policy requires it. The video tracker runs cloud or on-premise as well, so video processing can stay inside your building where your policy requires it. Output formats and integration are defined during the camera assessment, because those are your constraints before they are ours.
Counting people is not watching people
This is the part of the project that decides whether it succeeds. Occupancy measurement in a workplace touches employees, and in the EU it touches the GDPR and, in several countries, formal consultation with worker representatives. Treating that as a legal formality at the end is the most reliable way to have the project stopped. We treat it as a design constraint at the start.
Aggregated counts, not individual tracking
The purpose is to know how many people used a zone and when. Not who, not for how long each person stayed at their desk, not who spoke to whom. A system designed to count does not need identification to answer any of the questions in the section above, and we configure it that way. Facial recognition is not required to measure occupancy, and we do not recommend it for this use case.
Tell people first, in plain language
Publish what is measured, why, at what level of detail, where the data goes and how long it is kept. Post signage where cameras or sensors are used. Brief managers before deployment, not after someone notices a new device on the ceiling. Employees object to being watched, far less to a count of how full the third floor was on Wednesday.
Do not repurpose the data
The fastest way to lose trust is to use occupancy data for something it was not announced for, for example individual attendance or performance. Decide in writing that the data is used for space and services decisions only, and give it a retention limit. If your HR team wants attendance records, that is a separate system with a separate legal basis, and we will say so.
What your DPO will ask for
Expect a data protection impact assessment, a defined legal basis, a retention period, a description of where processing happens and a list of who can access what. We provide the technical facts that go into that assessment. Where the processing happens, what leaves the building and what is retained are settled in the site assessment and the first call, with your IT team and your DPO in the conversation. On this point your policy is the binding constraint before any product is, so bring it to the call and we design to it rather than around it.
What occupancy data will not tell you
The things this project does not deliver. If one of them is what you actually need, the honest answer is that you are looking at the wrong system.
Who sat where
By design, and we consider that a feature. If your requirement is individual attendance, per person desk history or time at desk, this is not the system, and in most European jurisdictions it is a significantly harder project than the vendor demo suggests. We would rather lose that project than deliver something your employees discover as a surprise.
Why people do not come in
Occupancy data tells you that the fourth floor is empty on Fridays. It will not tell you whether that is commuting, the quality of the desks, team agreements or a manager who works from home. The numbers show the pattern, a short survey and a few conversations explain it. Buying a sensor to answer a cultural question is a common and expensive mistake.
Room bookings and the no shows
This is not a room booking system. It can show that a room booked for a large group was used by a couple of people, or that a booked room stayed empty, which is exactly the evidence needed to change the booking rules. Acting on it, releasing rooms automatically or changing policy, happens in your booking platform, which we connect to through the API.
A layout that does not work
Measurement shows where the pressure is. It will not design the workplace. If everyone crowds the same corner because it is the only quiet area with daylight, the data will show you the crowding, and the fix is furniture, acoustics and rules, not another sensor.
Attendance and time tracking
We do not sell an HR time tracking tool, and occupancy data should not become one. Different purpose, different legal basis, different system.
What it costs
We work with a USD 5,000 minimum commitment. It covers a project we can run properly, assessment and pilot included. What sits above it depends on the building, not on a price list.
What drives the cost
The number of zones and rooms to measure. Whether usable CCTV already exists, and whether the angles and coverage are good enough to count reliably. The number of floors and buildings. Rooms that need a dedicated sensor because a camera is not an option. Integration into your BI or facilities tools. Cloud or on-premise. A hypothetical example: two floors with usable cameras in the open areas plus eight meeting rooms that each need their own sensor is a different project from counting one entrance, even though both are bought under the word occupancy.
What keeps it down
Measuring the floors where a decision is pending instead of the whole estate. Re-using existing cameras. Choosing zone level counts rather than detail nobody will act on. Running a pilot on one floor and pricing the rollout on what it actually showed.
Licensing
Two models exist, and you choose the one that fits the case: a cloud subscription, or a one-time license for an on-premise installation. Which one is right depends on where the data has to be processed, how long you expect to run the system and how your organization prefers to carry the cost, so we settle it with you once the assessment has told us how many zones and how many buildings are in scope. Either way the floor is the same USD 5,000 minimum commitment.
How we work
Four steps, and a clear exit point after each one.
1. Discovery call
30 minutes, free, with an engineer. Which decision is pending, which spaces matter, what already exists in the building and what your position is on employee privacy. Sometimes the conclusion is that a short manual count is enough to make this particular decision, and we will tell you that.
2. Site assessment
Floor plans with the zones marked, the camera list with positions and angles, the rooms where cameras are not acceptable. We check what can be re-used, what needs to be added and what the counting will and will not see.
3. Pilot on one floor
One floor or one set of zones, measured for a defined period, with the privacy communication in place from day one. For BLE equipment tracking the pilot uses at least 6 nodes covering roughly 300 square meters. The pilot is also where the works council conversation happens with a real example rather than a hypothesis.
4. Rollout and integration
Extend to the remaining floors and send the data where it will be used: REST API, push callbacks or daily CSV. Then the reporting cadence that matches the decision cycle, monthly or quarterly rather than a live dashboard nobody opens.
Frequently asked questions
Can the system identify individual employees?
For occupancy work it is configured to count, not to identify, and facial recognition is not required to answer any occupancy question. We recommend stating that explicitly in your internal communication and in your privacy notice, because the objection your employees will raise is not technical, it is about trust. If a stakeholder in your organization asks for individual identification, treat that as a different project with a different legal basis, and expect a much harder conversation with your DPO and your works council.
Do we have to tell employees and consult worker representatives?
Assume yes, and do it early. In the EU, workplace monitoring generally requires transparency towards employees and, in several countries, consultation with worker representatives before deployment. Even where it is not strictly mandatory, announcing it after installation is the surest way to turn a facilities project into a dispute. Your legal and HR teams own the formal requirement. We provide the technical description they need for it.
Can we use the CCTV cameras we already have?
Often yes, and that is usually the cheapest route, because the hardware is already installed and powered. What matters is coverage and angle: a camera aimed at a door for security reasons may not see enough of a room to count people in it reliably. We check that during the site assessment and tell you which cameras are usable, which need repositioning and which zones need something else.
How accurate are the counts?
The video tracker counts with 98% counting accuracy and reaches 20 cm spatial precision. We still validate the count against a manual count on your floor during the pilot, because a number from someone else's building is not a number you can plan with. Camera angles, lighting, occlusion and how people move through the space all move the result, and the conditions that apply to your site are defined during the site assessment. Treat any accuracy figure you are shown, ours included, as a starting point until it has been measured where you intend to use it.
What about meeting rooms where a camera is not acceptable?
That is the normal case, and it is why occupancy projects usually mix technologies. Rooms get a sensor that reports occupancy without producing an image. Open areas and circulation spaces are covered by the cameras that already exist. Which sensor we recommend for the rooms depends on ceiling height, room size and what you need to know, which is presence only or an actual headcount.
Is this GDPR compliant?
Compliance is a property of your deployment, not of a product, so no honest vendor can tick that box for you. What we can do is make compliance achievable: counting without identification, a clear description of the processing for your DPIA, and on-premise processing when your policy requires that video data does not leave the building at all. The legal basis, the assessment and the employee communication stay with you, and we will support them with the technical detail.
Specifications
Video precision: 20 cm, on existing CCTV cameras
Counting accuracy, video: 98%, both at a crossing line and per zone, validated against a manual count in the pilot
What is measured: Counts, flows and dwell time per zone, anonymous
BLE accuracy, for equipment: 1 to 3 meters in 90% of cases
Node range and mounting: Beacons detected up to 30 meters, ceilings up to 7 meters
Update rate: In the order of seconds
Pilot, BLE equipment tracking: At least 6 nodes, about 300 square meters
Integration, BLE platform: REST API, push callbacks, daily CSV
Deployment: Cloud by default, on-premise supported, for both the BLE platform and the video tracker
Minimum commitment: USD 5,000
Bring the decision, not the shopping list
Tell us which space decision is pending and what your building already has. We will tell you what can be measured, what it will cost and how to introduce it without losing the trust of the people who work there. If the honest answer is that you do not need a system for this, we will say that too.