How to run an office utilization study that planners can use is a question that sounds straightforward until you are sitting in front of a lease renewal decision, a portfolio consolidation, or a capital request to retrofit three floors — and realize the headcount reports, badge data, and anecdotal feedback in the room tell completely different stories. A utilization study done well resolves that conflict by producing a single, defensible dataset that connects how space is actually occupied to what the organization needs to decide next.
Why Most Utilization Studies Fail Before They Start
The most common failure in office utilization work is not bad data — it is a misaligned brief. Teams launch a study because a lease is expiring, a CFO asked a question, or a redesign project needs a number to anchor its scope. Without a written statement of what decisions the study must support, every subsequent data choice becomes arbitrary.
A brief should state the decision horizon, the asset or portfolio in scope, the audience for the findings, and the minimum confidence threshold the organization requires before acting. Those four elements force every later choice — what to measure, how long to observe, what baselines to use — to serve the same purpose.
The second structural failure is treating utilization as a single metric. Peak occupancy, average daily occupancy, seat utilization, and space utilization are related but distinct. Confusing them produces recommendations that do not match the actual constraint. A building may show high peak occupancy for two hours a day and very low average utilization across the week, which implies a scheduling problem, not a supply shortage.
Getting the definitions right before data collection begins is non-negotiable. Document each metric name, the calculation method, the unit of observation, and the time window in a measurement protocol that every data collector signs off on. That document becomes the audit trail when findings are challenged later.
Defining the Scope of the Study
Scope decisions in a utilization study cascade quickly. A study covering a single floor for a lease renewal decision requires different rigor than a multi-site portfolio rationalization supporting a ten-year capital plan. Underscoping wastes an opportunity; overscoping produces findings too diffuse to act on.
The practical starting point is to map every distinct space type in scope: enclosed offices, open workstations, conference rooms by seat count, focus rooms, collaborative areas, amenity spaces, and support areas. Each type may have a different utilization pattern and a different planning implication. Collapsing them into a single number destroys the information.
For each space type, determine whether the question is about supply adequacy, demand timing, or behavioral fit. A conference room inventory may be adequate in total but undersized in room configuration — too many large rooms, too few small ones. A workstation count may be theoretically sufficient but clustered in one wing while another sits empty. The scope document should state which question applies to each space type.
Scope also includes time. A four-week observation window captures less variance than twelve weeks, but a twelve-week window costs more and may span seasonal anomalies. For most planning decisions, a minimum of four to six weeks is reasonable, chosen to avoid holiday periods and known atypical events like conferences or move-in weeks.
Choosing the Right Measurement Methods
Office utilization data is collected through three primary methods: badge or access-control reads, sensor-based presence detection, and structured observation. Each has a different profile of accuracy, cost, latency, and planning suitability. The right study uses the method whose output matches the planning question, not the method that happens to be available.
Badge data answers the question of how many people entered a building or zone on a given day. It is reliable at the aggregate level and is often already available from existing security infrastructure. Its limitation is that it records entry, not occupancy: a person who badged in and spent the day in a conference room shows up the same as one who sat at a workstation all day.
Sensor-based presence detection, whether passive infrared, ultrasonic, or camera-based with anonymized detection, answers the question of whether a specific space was occupied at a point in time. The data resolution is high and the observation can be continuous, but sensors must be validated against reality before the study begins. A poorly calibrated sensor in a conference room will systematically undercount or overcount, corrupting the analysis.
Structured observation, sometimes called the walkthrough or utilization survey method, uses trained observers who record the occupancy state of each space on a defined schedule — typically every thirty to sixty minutes across the working day. It is labor-intensive but produces auditable, human-confirmed data that is easier to defend in a capital request or lease negotiation. For high-stakes decisions, structured observation is frequently the most defensible method even when sensors are available.
Designing the Observation Protocol
The observation protocol is the technical document that turns a measurement approach into repeatable data collection. It specifies the observer rotation, the observation intervals, the data recording format, the state definitions for each space type, and the exception handling rules.
State definitions are the most consequential design decision in a walkthrough study. Each space should be classified at each observation interval into one of a small number of unambiguous states: unoccupied, occupied by one person, occupied by two to five people, occupied by six or more, or in some designs, reserved but unoccupied — sometimes called ghost occupancy or no-show occupancy. That last state is particularly important for conference room analysis because it distinguishes between rooms that are truly unused and rooms that are booked but not attended.
Observer training must be rigorous enough that two observers in the same space at the same time would record the same state. Run a calibration exercise before the study begins by having pairs of observers independently survey the same floor at the same time and comparing results. A discrepancy rate above five percent typically signals that state definitions need clarification.
The recording format matters more than teams expect. Paper forms introduce transcription errors; spreadsheet apps on mobile devices introduce entry errors and version-control problems; dedicated survey tools with pre-loaded floorplans reduce errors but require setup time. Whatever format is chosen, the data schema must match the analysis structure exactly so that cleaning time is minimized.
Build exception handling into the protocol before the study starts. What happens when a conference room is under maintenance? What if a floor is closed for a half-day town hall? Define the rule — exclude the interval, flag it, or impute from adjacent intervals — in writing, so the analyst is not making judgment calls during cleaning that could bias the results.
Collecting Badge and Sensor Data in Parallel
Even in studies where structured observation is the primary method, collecting badge data in parallel adds an important cross-reference layer. If the structured observation count on a given day differs substantially from the badge count for the same zone, one of three things is happening: the badge system has a gap, the observers missed a population, or people are arriving without badging. Each explanation has a different planning implication.
Request badge data exports at the finest granularity available from the security or facilities management system — ideally individual badge events with timestamps, not daily aggregates. Daily aggregates prevent time-of-day analysis, which is often the most valuable output for space planning because it identifies the specific hours when demand peaks and the specific hours when space is functionally empty.
When sensor data is available alongside badge data and observation, the analysis framework should use each source for the question it answers best. Badge data anchors total daily demand. Sensor data provides continuous occupancy at the space level. Observation data provides validated spot checks and behavioral context that neither badge nor sensor data can supply. Triangulating across sources produces findings that are harder to challenge and more useful for planning.
Document the provenance of every data stream — the system it came from, the extraction date, the responsible owner, and any known data quality issues. This documentation is the foundation of the evidence trail that facility planners and capital committees need when the study findings are presented.
Cleaning and Validating the Dataset
Raw utilization data is never clean. Badge files have encoding errors, duplicate events, and unmatched card IDs. Sensor exports have dropout intervals where hardware lost connectivity. Observation sheets have blank cells from an observer who ran short on time. Cleaning is not a minor step — it is frequently where the most consequential analytical decisions are made.
The first cleaning step is to identify and document all missing intervals. A space that was not observed or not sensed for a half-day interval is not the same as a space that was unoccupied. Missing data should be explicitly coded as missing rather than defaulted to zero, which would artificially suppress utilization rates.
The second step is to flag outlier days — days where total observed occupancy is implausibly low or high relative to the rest of the study. Outliers are not automatically removed; they may be real signals about demand variability. But they must be investigated and documented before inclusion in the analysis.
The validation step involves comparing independently collected data streams against each other. If badge counts for a floor show consistently higher occupancy than observation counts on the same days, the gap needs an explanation. Common causes include multi-floor badge reads being attributed to the wrong floor, observers who skipped certain zones, or populations who work in areas outside the study boundary. Resolving the gap produces a more accurate final dataset and a more credible findings package.
Analyzing Utilization and Calculating Planning Metrics
With a clean dataset, the analysis should produce a layered set of metrics. At the asset level, calculate average daily occupancy as a percentage of capacity. At the space-type level, calculate the utilization rate for each category — open workstations, enclosed offices, conference rooms by size tier — over the full study period. At the time-of-day level, construct an occupancy profile that shows the demand curve from first arrival through end of day.
The planning metric that matters most depends on the decision. For a lease renewal where the organization is considering reducing footprint, the relevant metric is peak coincident occupancy — the highest number of people in the space at the same time, observed across the full study period. That number represents the minimum supply the organization must retain to avoid congestion on its busiest days.
For a redesign that will shift the balance between assigned seating and unassigned workstations, the relevant metric is average daily seat utilization by team or business unit, combined with the coefficient of variation for each team's demand. A team with high average utilization and low variance behaves predictably and can be planned around confidently. A team with moderate average utilization but high variance needs a planning buffer.
For a conference room inventory decision, the key metric is the distribution of room utilization by configuration — a histogram showing how often each size tier is used as a percentage of its available hours. A portfolio with many underutilized large rooms and consistently overloaded small rooms has a configuration problem, not a quantity problem. The recommendation is reconfiguration, not addition or removal of total room inventory.
Benchmarking against external references adds context but should be used carefully. Published benchmarks for office utilization vary widely by industry, work mode, and study methodology. A building with thirty-five percent average daily workstation utilization might be normal for a professional services firm with heavy client travel, or it might signal significant excess capacity in a different context. Always present benchmarks with their source methodology disclosed.
Translating Findings into Planning Outputs
A utilization study that stops at metrics has not delivered a planning output — it has delivered a dataset. The translation from dataset to actionable finding requires the analyst to connect each metric to a specific planning question and state a recommendation with the confidence level the data supports.
The standard output structure for facility planning purposes is a findings report with four components: the measurement methodology and data quality summary, the utilization metrics for each space type and time window, the planning implications tied to specific decisions, and the data limitations and recommended follow-on work. Each component serves a different audience. The methodology summary protects the analyst if findings are challenged. The implications section is what the decision-maker reads.
Scenario modeling belongs in the planning implications section. For a floor consolidation decision, model the space requirement under two or three scenarios: current demand held constant, demand reduced by a specified percentage, demand increased by a specified percentage. Apply the peak coincident occupancy metric to each scenario to determine the minimum supply threshold. Present the scenarios as a sensitivity table so decision-makers can see the consequence of each assumption. These scenarios are illustrative — they should be clearly labeled as such — but they give the committee a structure for the conversation rather than leaving the numbers floating.
Presenting Findings to Planners and Decision-Makers
The presentation of a utilization study to planners requires a different structure than the presentation to a capital committee. Planners need to see the floor-level data with enough spatial resolution to design against. Capital committees need to see the asset-level findings with enough financial context to approve a recommendation.
Build two versions of the findings package from the same underlying analysis. The planner's package should include occupancy heat maps by floor zone, time-of-day demand curves for each space type, and a comparison of observed versus design capacity for each zone. These outputs tell a designer where to add, reduce, or reconfigure space and at what times the results would be most consequential.
The executive summary package should contain the peak coincident occupancy number, the utilization rate by space type expressed as a percentage, the scenario table showing space requirement under each demand assumption, and a single-page statement of the recommended decision. The recommendation should state the basis for the confidence level explicitly: "Based on six weeks of structured observation across all occupied floors, with badge data cross-reference, the study supports a recommendation to reduce workstation inventory by approximately X percent, subject to validation of the following assumptions." Approximate ranges are more credible than false precision when the data is known to have gaps.
Anticipate objections by pre-addressing data quality limitations in the findings package rather than waiting for the question. If the study covered only weekdays and not weekends, say so and state why it does not affect the finding. If a floor was excluded due to a renovation, say so. If the badge data had a known gap in week two, say so and explain how it was handled. Transparency about limitations builds credibility rather than undermining it.
Building a Repeatable Measurement Program
A one-time utilization study is a snapshot. A repeatable measurement program is a planning asset. The difference is whether the measurement protocol, data schema, and analysis template are documented and preserved so that the next study can be run against the same baseline rather than starting from scratch.
For portfolio-level facility planning, the goal is annual or biannual utilization baselines for every significant asset, collected with consistent methodology so that trend data is meaningful. A building that showed forty percent average workstation utilization in one period and sixty percent in the next has experienced real demand growth — but only if the two studies used the same method. If the first used badge data and the second used structured observation, the difference may be methodological.
Establish a measurement calendar tied to the lease critical-date schedule. Leases with renewals or termination options in the next thirty-six months should have a current utilization baseline on file. That baseline feeds the renewal negotiation, the space planning brief, and the capital request with actual demand data rather than headcount projections that may not reflect behavior.
Document the roles responsible for data collection, cleaning, analysis, and presentation in a standing RACI. Without clear ownership, the program drifts — studies are delayed, methodology shifts between cycles, and the trend data loses its value. The RACI also ensures that when personnel change, the institutional knowledge of the measurement program survives.
Integrating Utilization Data into Broader Portfolio Decisions
Office utilization findings do not live in isolation. They interact with headcount forecasts, lease term structures, capital budgets, and workplace strategy commitments. A study that is not connected to those inputs produces a recommendation that may be analytically sound but practically unexecutable.
The integration point with headcount planning is the demand assumption. Every utilization study implicitly assumes a future demand scenario. Making that assumption explicit — stating whether the forecast headcount is flat, growing, or contracting, and over what horizon — allows the planning team to test whether the utilization finding holds across realistic futures. A finding that holds only under the most optimistic headcount scenario is a less actionable finding than one that holds across a range.
The integration point with lease term structure is the timing of decisions. A utilization finding that supports a twenty percent footprint reduction is not actionable if the lease has seven years remaining and no early termination option. The finding should be connected to the specific lease events where action is possible — renewal options, expansion rights, termination clauses — and sequenced against those events. This is the connection point where a commercial real estate CRM and lease analysis software overlap: tracking the relationship between utilization data, tenant obligations, and option windows in one place prevents the finding from aging unused until the window has passed.
The integration point with workplace strategy is behavioral. A utilization study reports what people do in a space, not why they do it. When the findings show low utilization of a specific floor or zone, the cause may be geographic inconvenience, team culture, inadequate amenities, or poor technology setup. Strategy recommendations that address only supply without addressing the behavioral driver frequently fail to produce the expected change. Pairing the quantitative study with structured interviews or workplace surveys covering the same population and time period produces findings that explain the pattern, not just measure it.
A repeatable, well-documented utilization study is one of the most durable assets a facility planning function can build. It converts intuition about how a building is used into defensible data. It converts that data into planning outputs that decision-makers can act on. And it connects those outputs to the lease, capital, and portfolio decisions where the financial consequences of getting it wrong are largest. The methodology is demanding, but the return — in credible, actionable insight — justifies the investment every time.
About Advantai
Advantai is a commercial real estate intelligence and operations platform operated by ADVANTAGE AI LLC, a Delaware limited liability company. It connects client relationships, property research, documents and financial decisions in one workspace for commercial real estate teams — advisers and brokerage teams, occupier and facility teams, and portfolio teams. The platform covers CRM and origination, requirements and site selection, Property X-Ray (an interactive 3D building workspace), financial modeling and comparison, document intelligence, transactions and diligence, client collaboration, and portfolio strategy with critical dates. The optional Super Agent upgrade adds specialist, source-backed research and automated scenario analysis.
Get Started with Advantai
Ready to see your next move clearly? Go to advantaico.com, click Request a demo and tell us about your next project. Prefer to start with a single project? Visit advantaico.com/getting-started to plan your first one.