Occupancy Prediction Service Impact (OPSI)

GMI-CRE-002Corporate Real EstatePilot-Ready

Definition

Frequency and cost of building service failures (insufficient HVAC, cleaning, elevator capacity, parking, catering) caused by AI occupancy prediction errors on specific days and floors.

The Measurement Gap

AI occupancy prediction drives building service configuration. HVAC pre-conditioning, cleaning crew sizing, elevator dispatching, cafeteria preparation. When the prediction is wrong, the building is configured for a day that does not materialize. A company-wide event pushes Thursday to 95% occupancy while the building is configured for 30%: insufficient cooling, inadequate cleaning, meeting rooms fully booked, visitor parking overflowed. Standard prediction accuracy (MAPE) reports the model is accurate on average. OPSI measures the cost of the specific days where the prediction fails at the specific granularity (day and floor) where the failure causes a service gap. Average accuracy obscures the prediction misses that matter most.

Decision Supported:

How often does the building get caught unprepared because the AI predicted low occupancy on a day that turned out high, and what does each prediction miss cost in degraded service and occupant dissatisfaction?

Formula

Plain language: For each day and floor or zone, compare the AI's predicted occupancy to actual occupancy. When actual significantly exceeds predicted (triggering a service gap), estimate the cost of the service degradation: HVAC running behind, cleaning crew insufficient, meeting rooms all booked, visitor parking full.

Notation:

OPSI = Σ (ServiceGap_cost_d × Severity_d) / N_days

Variables:

Variable Definition Unit
ServiceGap_cost_d Estimated cost of service failures on day d attributable to occupancy underprediction (complaint handling + overtime cleaning + comfort recovery energy + occupant productivity loss estimate) $
Severity_d (Actual_occupancy_d − Predicted_occupancy_d) / Capacity_d, counted only when positive and above a materiality threshold (e.g., greater than 15% underprediction) Ratio
N_days Total business days in the measurement period Count

Proposed formula, pending validation pilots.

At a Glance

Unit
Currency per day
Trend Direction
Lower; lower daily cost = fewer prediction-driven service gaps
Reporting Frequency
Monthly
Category
Space Intelligence and Workplace Effectiveness
← Lease Abstraction Error Cost (LAEC)Space Value Index (SVI) →

Assess Your Organization

The AI-GPS Assessment evaluates your AI governance maturity and recommends KPIs like this one for your implementation roadmap.

Take the Free AI-GPS Assessment