The Hybrid Workplace Measurement Gap

The practical conclusion

Do not ask whether “office attendance is up.” Ask which population attended, when, for how long, how the space performed, what the experience was, and whether the portfolio outcome improved.

Research context

The office has become a variable-demand environment. Before widespread hybrid work, capacity planning could rely more heavily on assigned populations and relatively stable schedules. Today, employee choice, team norms, policy mandates, travel, leave and event programming all influence demand. The same weekly attendance average can therefore produce very different operational realities.

At the same time, leaders are under pressure to reduce occupancy cost, improve employee experience, support collaboration, advance decarbonisation and prepare data for AI. These objectives can conflict. Consolidating space may reduce cost and energy use, but can create peak-day crowding. Adding amenity space may improve experience, but may lower conventional desk-density ratios. Tight attendance enforcement may lift badge counts, but does not prove that employees had productive reasons to be present.

The measurement challenge is not a minor analytics problem. Definitions determine capital decisions. If attendance frequency is mistaken for space utilisation, an organisation may preserve too much space. If a weekly average masks peak congestion, it may remove too much. If policy compliance is treated as a proxy for performance, it may optimise visible presence instead of contribution.

Evidence review

Office use is rising, but benchmarks are not interchangeable JLL's public benchmark page reported global utilisation of 56% in 2026. The study covered 84 organisations and 716 million square feet in portfolios supported by JLL Occupancy Planning and Management.[1] JLL also reported that the gap between actual and target utilisation narrowed from 25 to 18 percentage points. That improvement consisted of a two-point rise in actual use and a five-point reduction in the target. The result matters, but it is partly a story of recalibration rather than demand alone.

CBRE's 2026 global study reported utilisation of 53%, compared with 38% in 2024 and 35% in 2023, from client data covering 303 million square feet and an average portfolio size of five million square feet.[2] The increase is directionally consistent with JLL's findings. It cannot, however, be merged with JLL's 56% figure without access to common definitions, sampling frames and sensor or observation protocols.

For decision-makers, the lesson is positive but bounded: sampled portfolios are being used more intensively, while substantial capacity remains unused during measured periods. A benchmark should initiate a local diagnostic, not settle it.

Policy clarity is increasing; effectiveness still requires proof JLL reported that 62% of benchmarked organisations required fixed in-office days in 2026, up from 49% a year earlier.[1] Its separate Workforce Preference Barometer found that 66% of workers said their company set clear on-site expectations and 72% viewed structured hybrid policies positively.[3] These results challenge the simplistic idea that all structure is unwanted. They do not show that every form of mandate is effective.

CBRE's Americas survey of more than 180 CRE leaders found that 72% of organisations reported achieving attendance goals, up from 61%, while 69% measured policy compliance and 37% enforced it.[4] The expected-versus-reported gap— 3.2 days versus 2.9—shows that aggregate goal achievement can coexist with a residual shortfall. The unit of analysis also matters: an organisation reporting that it achieved a goal is not the same as an employee complying on a given week.

Peak demand is the operational centre of gravity CBRE's European survey reported 46% average weekly utilisation and 71% peak-day utilisation.[5] That 25-point difference demonstrates why a weekly average is insufficient for space planning. A building can look inefficient over a week while producing queues for rooms, noise, poor choice and neighbourhood crowding on preferred days.

The Americas survey similarly reported that 73% of responding organisations fell into an effectively-at-capacity band on peak days, compared with 34% at average attendance.[4] These figures are shares of organisations in defined capacity bands, not utilisation rates. They should not be compared numerically with Europe's 71% peak-day utilisation. Both studies nevertheless identify the same operating pattern: temporal concentration can create scarcity inside an underused portfolio.

Experience and purpose remain independent outcomes JLL's workforce survey found that almost 40% of respondents believed their office experience could improve; 65% prioritised work-life balance, 57% wanted flexible hours and 49% reported access to them.[3] CBRE's 2026 global study found that 68% identified collaboration as the most important reason to attend.[2] Together, these findings suggest that presence must be connected to a credible employee proposition.

Higher utilisation may demonstrate demand recovery. It does not establish that the workplace is supporting focused work, learning, belonging or collaboration. Those outcomes require separate evidence, ideally combining experience surveys, task or team patterns, service data and business indicators.

Findings

1. A single utilisation number cannot describe a hybrid portfolio

A portfolio average suppresses day, hour, building and team variation. Leaders need distributions: median and percentile use, peak duration, frequency of constrained periods and the share of space below a chosen threshold.

2. Target attainment can improve because targets change

JLL’s narrower actual-to-target gap included both higher actual use and a lower target.[1] Governance should therefore show the movement of actuals and targets separately.

3. Peak-day stress is compatible with weekly underuse

The European 46% weekly and 71% peak figures show why consolidation based on averages alone can degrade the workplace.[5] Peak smoothing and service orchestration may create value before lease action.

4. Policy acceptance, clarity and compliance are different constructs

A worker can value clarity without agreeing with the number of required days; an organisation can meet a goal while individual compliance varies. These measures should not be collapsed.

5. Data readiness is the constraint on advanced analytics

JLL reported that 73% had data-governance programmes, yet 70% had not begun AI implementation in occupancy planning.[1] CBRE found 55% cited data quality or lack of expertise as AI challenges.[2] Governance activity is not the same as usable data.

6. Workplace effectiveness must be measured beyond presence

Experience, collaboration quality, access to space, service reliability, work-life balance and team outcomes are necessary counterweights to attendance and utilisation.

Osmos Global analysis

A five-layer workplace measurement stack Layer Primary question Example measures Decision supported Policy What is expected? Required days; eligible population; exemptions Policy design and communication Presence Who came, and how often? Unique attendees; frequency; duration; compliance Work-pattern diagnosis Demand When and where was capacity used?

Hourly utilisation; peak/average ratio; room demand Space and service operations Experience Did the workplace enable the work?

Ease of finding space; perceived effectiveness; service quality Workplace design and programming Portfolio Did the system create value? Cost per supported employee; energy per occupied hour; lease exposure Capital and portfolio strategy The stack makes causal reasoning explicit. A policy intervention should first change presence. Presence may change demand patterns. Demand patterns and service quality influence experience. Those effects—combined with lease and operating decisions—shape portfolio value. Skipping layers encourages false attribution.

For India and GCC portfolios, this distinction is especially important. Fast growth can make utilisation appear healthy because headcount and demand are rising. Yet growth markets can still suffer peak concentration, poor meeting-space availability and fragmented data across campuses, flex space and managed offices. A common measurement dictionary should therefore precede cross-site benchmarking.

Osmos Global infers that the next competitive advantage will come less from collecting more raw signals and more from connecting the right signals to decisions. Badge, sensor, booking, HR, service and energy data each have blind spots. A trustworthy operating view triangulates them, documents exclusions and avoids identifying individuals beyond legitimate, disclosed purposes.

Recommendations 1. Adopt a written measurement dictionary. Define numerator, denominator, population, location, time window and exclusions for every headline metric. 2. Publish average, peak and distribution measures together. At minimum, report weekly average utilisation, peak-day utilisation and the number of constrained hours. 3. Separate actual performance from target changes. Show both lines over time and document each target reset. 4. Build a balanced scorecard across policy, presence, demand, experience and portfolio value; do not allow attendance to stand in for effectiveness. 5. Triangulate data sources. Reconcile badge, sensor, booking and survey evidence before committing capital. 6. Treat privacy and governance as product requirements. Establish purpose limitation, access controls, retention rules and transparent employee communication. 7. Run reversible pilots before portfolio action. Test team scheduling, event programming, service shifts and neighbourhood changes before lease exits or major fit-outs. 8. Create decision thresholds in advance. Define what evidence would trigger expansion, consolidation, redesign or no action.

Risks, limitations and unresolved questions

• The benchmark samples are not random representations of all global offices; large JLL and CBRE clients may differ from smaller occupiers. • Public pages do not expose every methodological detail, including all sensor rules, weighting and missing-data treatments. • Self-reported attendance, organisational estimates and observed building use are different evidence types. • Utilisation gains may reflect policy changes, portfolio consolidation, measurement changes or economic conditions— not a single cause. • The relationship between presence and business performance remains context-dependent and requires team-level longitudinal evidence. • Privacy, works-council and employment-law constraints vary by jurisdiction and require local review.

Decision checklist Before acting, can we answer… Yes/No Is the metric defined with a denominator, population and time window?

Are average and peak demand both visible?

Are actuals separated from targets?

Have at least two data sources been reconciled?

Is employee experience measured independently?

Are privacy and local legal requirements cleared?

Is the capital decision reversible or staged?

Is the evidence strong enough for the size of the decision?

References

[1] Xie, W., & Holmes, M. (2026, 19 May). Global Occupancy Planning Benchmark Report 2026. JLL. https://www.jll.com/en-us/insights/occupancy-benchmark-report (accessed 27 August 2026). [2] CBRE Workplace & Occupancy Research. (2026, 20 January). 2026 Global Workplace & Occupancy Insights. CBRE. https://www.cbre.com/insights/reports/2026-global-workplace-and-occupancy-insights (accessed 27 August 2026). [3] Pradère, F., Naït-Belkacem, D., & Ruiz de Castañeda, B. (2025, 9 September). JLL Workforce Preference Barometer 2025. JLL. https://www.jll.com/en-us/insights/workforce-preference-barometer (accessed 27 August 2026). [4] CBRE Research. (2025, 11 August). 2025 Americas Office Occupier Sentiment Survey. CBRE. https://www.cbre.com/insights/reports/2025-americas-office-occupier-sentiment-survey (accessed 27 August 2026). [5] CBRE Research. (2025, 1 October). European Office Occupier Sentiment Survey 2025. CBRE. https://www.cbre.com/insights/reports/european-office-occupier-sentiment-survey-2025 (accessed 27 August 2026).

Editorial note This publication presents original Osmos Global analysis based on publicly available and cited research. Source findings and Osmos Global interpretations are distinguished throughout. Third-party trademarks and source materials remain the property of their respective owners. This publication is provided for research and professional-information purposes and does not constitute legal, financial, investment or technical advice.