What is multi-source occupancy?

How it works

Each signal sees part of the picture. Bookings show who planned to come in. Badge swipes show who entered the building, but not where they sat. Wi-Fi or office network presence shows who was on the office network, often down to the floor. Sensors show whether a desk, room or zone was in use, but not who used it.

A multi-source model matches these signals by person, place and time, removes duplicates and fills gaps. A person who badged in, connected to the floor network and checked in to a desk counts once. A person with no booking who shows up on the network still counts, so walk-in use is not lost.

The result is one figure per building, floor or team that does not depend on every employee booking, every door having a reader or every desk having a sensor. When sources disagree, the model can flag the reading as low confidence instead of hiding the conflict.

How to measure it

List every source you have at each site: bookings, check-ins, badge data, Wi-Fi or network presence and sensors. Note what each one covers and what it misses.
Match records by person and day (badge, network, check-in) and by space and hour (sensors, bookings), so one visit counts once.
Compare each single-source figure with the combined figure for the same period. Large gaps show where a source is weak.
Report the combined figure with the sources behind it for each site, so readers know how much to trust it.

Examples

Badge data shows who entered the building and network presence shows which floor they spent the day on; together they give floor-level attendance without desk sensors.
A site with sensors on one floor and none on the others uses network presence and check-ins for the rest, so every floor reports on the same basis.

Common mistakes

Adding sources together without removing duplicates, so one person counts two or three times.
Treating bookings as presence. A booking shows intent; it should confirm or explain other signals, not stand in for them.
Ranking sites with different source coverage side by side without saying so.

How Kadence helps

Kadence records presence from several signals today. The Kadence Agent checks people in automatically when it detects them on the office network, Enhanced Network Detection adds floor-level, VPN-aware presence, Automatic Bookings records people who turn up without a booking, and check-ins can also come from kiosks, room displays, badge data or the API. Workplace Insights reports attendance and occupancy from those check-ins, not bookings alone.

Workplace occupancy · Workplace analytics · WorkOps

Frequently asked questions

Why combine occupancy data sources?

Every source has blind spots. Badges miss people who follow someone through the door, bookings miss walk-ins and include no-shows, sensors do not know who used a space, and network presence misses people whose laptops are off. Combining them gives one figure that holds up when one source is weak, which matters when that figure drives lease and consolidation decisions.

Do you need sensors for multi-source occupancy?

No. Many organizations combine bookings, check-ins, badge data and office network presence without any sensors. Sensors add detail where they exist, such as how many people are in a room, but the model does not depend on them. That lets every site report on the same basis, even where installing hardware is not practical.

What is a unified occupancy model?

It is another name for multi-source occupancy: a model that merges presence signals such as Wi-Fi, badge, sensor and booking data into one occupancy figure. The word unified stresses that leaders see one number per site rather than a separate report from each system.

How is multi-source occupancy different from badge data?

Badge data is one source. It shows who entered a building and when, but not which floor or desk they used or how long they stayed. Multi-source occupancy uses badge data alongside network presence, check-ins and sensors to answer those questions.

How do I combine badge, Wi-Fi and booking data into one occupancy number?

Match each source to the same person, place and day, remove duplicates, and count someone as present when a trusted source shows them on site. Keep each source visible so you can see where they disagree. The result is one figure you can defend to leadership, with the evidence behind it.

Presence data (sensorless occupancy) The record of who was actually in the building and when, from signals you already have: network, badges and check-ins.
Badge data Access-control swipe records showing who entered a building or floor and when, often used to measure attendance.
Occupancy sensor A device that detects whether a desk, room or zone is in use, or counts the people in it, usually anonymously.
Occupancy confidence score A rating of how reliable an occupancy reading is, based on how many independent data sources agree.
Walk-in usage Desks or rooms used without a booking. If it isn't captured, utilization reports undercount real demand.
Occupancy monitoring Continuous measurement of how many people use each building, floor, room and desk, and when, from presence signals or sensors.

Kadence editorial team · Last reviewed October 6, 2026