What occupancy can tell you about resource use

A question at the end of the working day

At 21:00, an office-floor people count reaches zero while electricity demand remains at 27 kW. That pattern gives the operating team a focused question: which loads are still needed? Start by confirming what the count covers and what the meter supplies.

The chart below uses synthetic data to illustrate that investigation; it is not a Good Buildings project result. Comparing space use with resource demand helps the team identify where to look. The responsible specialists then assess the loads and any proposed change.

Owners, occupiers and operators need a shared view of the question, the evidence and the response. Agree what the pattern warrants investigating, who will act and how any improvement will be assessed.

First agree what “occupied” means

Four different records can describe the same office without answering the same question:

• A booking records a reservation. It does not, by itself, establish that anyone arrived or stayed for the full slot.
• An access event records an event at a controlled point. Its meaning depends on the door, direction and event type. A building entry is not a desk-level observation .
• Measured presence or a people count describes what the sensing system detected within its defined coverage and time interval.
• Utilisation compares use with a denominator: available seats, spaces or time. The denominator belongs in the definition, not in a footnote.

These distinctions are reflected in VergeSense’s separate definitions for capacity usage, time usage and booking measures . They are examples of a supplier’s terminology, rather than universal definitions to impose on every project.

Consider a ten-seat meeting room used by two people for an entire hour. For that hour, a simple seat-capacity measure is 20%, while a simple occupied-time measure is 100%. Both can be correct. Neither establishes whether the meeting needed that room, whether an accessible alternative was available or whether a different layout would work better.

Choose the measure for the decision. Reviewing the mix of rooms needs a different view from finding a free room now. Agree capacity, working hours, minimum use duration and the treatment of walk-through traffic before comparing percentages.

JLL’s Global Occupancy Planning Benchmark Report 2026 identifies space-data accuracy as a growing priority and distinguishes building-level badge counts from the space-level evidence needed to assess focus rooms, phone booths and small meeting rooms. For your building, match the measurement to the space and decision, then check it against operating needs and people’s experience. The global study provides context, rather than a utilisation target for a UK office.

Match the place, period and unit

Match the measurement boundaries before comparing numbers. A whole-building electricity meter may include shared cooling, lifts, retail, car parks or another tenant, while the people count covers one floor. Draw the areas and supplies so the comparison has a clear basis.

Create a measurement record with the meter, supplies served, occupancy coverage, exclusions and the person who confirms them. Use the operating team’s metering records and relevant drawings to establish electrical and water boundaries. Record any gaps that affect the comparison.

NABERS’ Metering and Consumption Rules illustrate why this evidence matters: they set requirements for metering data, period alignment and the treatment of gaps and estimates. The rules apply to NABERS assessments; a general management dashboard is not a rating submission.

Also keep demand and consumption distinct. Kilowatts describe power; kilowatt-hours describe energy over time. An average demand of 20 kW sustained for half an hour represents 10 kWh. Do not add a series of power readings and label the total “energy” without accounting for their intervals.

Confirm whether a timestamp marks the beginning or end of an interval, its time zone and its daylight-saving treatment. Align the series to an appropriate common interval without pretending that monthly bills contain a half-hourly profile. For cross-country portfolios, preserve the building’s local operating day as well as an unambiguous stored timestamp.

Separate a quiet floor from a quiet data feed

Zero, missing, stale and estimated are different states. A value can be recent but incomplete, or accurate for one room while the rest of a floor is unmeasured. The dashboard should make those limitations visible before someone relies on its apparent precision.

There is an important implementation detail here. XY Sense’s explanation of occupancy processing says its analytics can omit zero-occupancy intervals; live and analytical outputs also apply different processing. A missing row in that documented dataset can therefore mean zero. It must not be confused with a failed export or an unavailable sensor.

The practical requirement is to understand each source’s contract. Confirm what an omitted interval means, how late data arrives, which filters apply and how the integration reports failure. Then test a known empty period, a known occupied period and a simulated data interruption. “No data” should not silently become “no people”.

Keep a visible coverage note and the time of the latest valid observation. Record layout changes, sensor moves, meter replacements and altered capacity. A room renamed halfway through the reporting period needs a maintained identity, not two unrelated histories with similar labels.

Give the pattern an owner and a next action

Return to the illustrative floor at 21:00. A sensible first response is to ask the facilities lead to confirm whether the people count is complete and the metered loads are correctly attributed. The team can then compare the pattern with cleaning, security, out-of-hours bookings, equipment requirements and operating schedules.

A persistent load might be legitimate. If a schedule looks inconsistent with the agreed need, the appointed engineer or controls specialist can investigate it. Any change should have an authorised owner, a record of the previous setting, operating safeguards and a plan to check the result. Good Buildings can help define that evidence and coordinate the decision; a chart is not an instruction to override building controls.

The US Department of Energy’s re-tuning process combines building information, trend analysis and a walk-through before corrective actions and follow-up analysis. That sequence is a practical way to organise an investigation: identify the pattern, verify it against the building, then agree and assess the response with the responsible specialists.

Use the same discipline for other resources. An unusual overnight water flow warrants checking the supply boundary, meter behaviour and expected activities before calling it a leak. For waste, ask which disposal points and periods are included, how quantities were obtained and which collections are absent. A low recorded weight may mean less waste, or simply less complete reporting.

A connected water system may also expose controls. Watergate describes a local API for Sonic telemetry and valve control. Receiving a consumption figure and authorising a valve command are different permissions. Define the reporting requirement, the permitted actions and the response owner with the appointed water-services and controls specialists before selecting the connection.

Waste information needs a defined collection process as well. WRAP’s workplace guidance uses waste audits and action plans to connect measurement to responsibility and review. For a building dashboard, establish whether a weight represents a bin, a tenant, a shared collection or an estimate. Preserve that distinction when comparing it with attendance.

A useful operating view brings together the question, aligned measures, coverage, relevant events and next action. Give each exception an owner, a check to carry out and a place to record the outcome.

Bring indoor conditions into the same conversation

A room that fills regularly and shows a recurring rise in CO₂ raises a different question from an empty room with persistent electricity demand. Occupancy can help interpret the timing, but it does not establish the ventilation rate, pollutant source or cause of a complaint on its own.

HSE’s guidance on using CO₂ monitors recommends repeated occupied-period measurements and recording occupancy and ventilation conditions. It also explains that placement affects readings and CO₂ is a broad ventilation guide, not proof of safe conditions.

Choose the information for the question: CO₂, particles, temperature, relative humidity and VOC measurements describe different things. Confirm the device’s actual measurements, siting, maintenance and limitations. Avoid compressing them into a single “healthy building” score.

Agree who reviews a recurring exception and who has the expertise and authority to respond. An air-quality view is more useful when it links to a facilities investigation, occupant feedback and a recorded resolution than when it ends at a coloured indicator.

Define how you will measure improvement

Agree the evaluation before making a change. Record the baseline conditions, intervention, review period and other influences, including weather, attendance, operating hours, equipment and measurement coverage. This gives the team a sound basis for explaining the result.

The US Department of Energy’s account of measurement and verification explains the role of documented baselines, responsibilities and adjustments. Its performance-contract procedures are not a universal requirement for every dashboard, but they show why a defensible savings claim needs more than a before-and-after screenshot.

Use a proportionate test: confirm the authorised change, check that the service still meets its operating need and compare sufficiently complete information. Report the observed difference with its measurement limits. Make a savings claim only when the evidence supports it.

Choose the question before the denominator. In a hypothetical comparison, daily electricity rises from 100 to 120 kWh while attendance rises from 10 to 20 people. Electricity per attendee falls from 10 to 6 kWh, but total consumption rises by 20%. Both measures matter. Show the absolute and relative figures together, with the conditions of comparison, so the reader can see what improved and what increased.

Agree information rights before connecting systems

Bringing sources together can change what they reveal. A count intended for space planning should not quietly become a named attendance record. Decide the required level of detail, permitted purpose, access, retention and export rights before designing the integration.

For UK projects involving personal information, the ICO’s worker-monitoring guidance explains the importance of necessity, proportionality and transparency. Involve the client’s privacy and security leads; an aggregate display alone does not establish that the underlying collection is anonymous.

Start with a short brief covering the intended outcome, measurement definitions and boundaries, source and quality checks, responsible teams and next decision. Use it to guide technology selection, information requirements and the review of operational results.

Sources and further reading

• JLL — Global Occupancy Planning Benchmark Report 2026. Global research on data accuracy, reporting and space-level utilisation; interpret its sample and geography before applying a benchmark locally.
• US Department of Energy — re-tuning investigation sequence. Connects building information, trend analysis, site observation, action and follow-up.
• Watergate — local telemetry and valve-control interfaces. An example of why a reporting connection and a control permission need separate requirements; product suitability remains a project-specific decision.
• WRAP — planning workplace recycling measurement and action. Explains the role of a waste audit and an action plan with responsibilities and review dates.
• VergeSense — occupancy intelligence overview. The supplier’s explanation of occupancy information and its workplace uses.
• VergeSense — definitions of capacity usage, time usage and booking measures . Useful when comparing the meaning of workplace metrics.
• XY Sense — how live detections become occupancy analytics . Documents filters, omitted zero intervals and time handling for that platform.
• NABERS — Metering and Consumption Rules . Formal requirements for data used in its assessments.
• HSE — siting, using and interpreting CO₂ monitors .
• US Department of Energy — baselines and verification in energy performance contracts .
• ICO — data protection and monitoring workers . Check its current status when planning a UK deployment.