01
What a facility management KPI dashboard should achieve
A facility management KPI dashboard should help an authorised reader understand what changed, why it matters, which evidence supports it and what decision is required. It is not a collection of every value available in a CAFM system or spreadsheet. A useful dashboard connects property objectives to maintenance delivery, asset condition, service demand, open exposure and agreed commercial responsibilities.
The dashboard should create a consistent management conversation without pretending that a favourable percentage proves the building is safe, compliant or free from hidden defects. Facility management systems such as ISO 41001 emphasise effective service delivery that supports organisational needs. The appropriate indicators, thresholds and evidence still depend on the property, contract, users, assets, risks and applicable requirements; this article does not claim certification or reproduce a standard.
02
Define the audience and decisions before the measures
List the people who will use the report and the decisions they control. A property owner may need trends, material exceptions, budget approvals and provider accountability. An operations manager may need overdue work, repeat faults and access constraints. A technical lead may need asset-level evidence and diagnostic follow-up. One dashboard can offer layers, but every page should have a named audience and purpose.
Write decision questions first: Which planned work was due and completed? Which critical open issues need approval? Where is service demand increasing? Which assets show repeated symptoms? Which records are missing? Then select measures that answer those questions. Avoid adding a metric merely because software can display it. Large dashboards encourage readers to scan colours instead of examining the few exceptions that require action.
03
Create a KPI dictionary and dependable source controls
Every KPI needs a written definition covering purpose, formula, numerator, denominator, inclusions, exclusions, unit, reporting period, source system, data owner, approval route and treatment of missing or late records. Define the event timestamp used for received, acknowledged, attended, restored, completed and closed. If two teams use different meanings for completion, their percentages cannot be compared fairly.
Keep source records traceable to work orders, asset references, planned schedules, inspection reports, invoices, meter records, surveys or other approved evidence. Control duplicate requests, reopened work, cancelled tasks, test records and requests held for owner approval. Show data-quality exceptions alongside performance results. A precise-looking percentage based on incomplete timestamps or an unverified asset list can mislead more than a clearly labelled provisional measure.
04
Measure service demand, backlog and ageing clearly
Begin with the flow of work: requests received, valid work orders created, completed work, open work and cancelled records by agreed category and priority. A backlog count alone is weak because ten recently raised minor items are different from ten long-outstanding issues affecting essential services. Group open work into defined ageing bands and show items waiting for access, quotation, material, specialist attendance or owner decision separately.
A completion rate may be calculated as completed valid work orders divided by valid work orders due or opened within the defined population, multiplied by 100. The chosen population and cut-off must be stated because an opened-versus-completed calculation can change when the month ends. Do not improve the result by closing unresolved work administratively. Reopened items and linked follow-up work should remain visible through an agreed rule.
05
Report planned maintenance without hiding deferred tasks
Planned-maintenance reporting should begin with the approved schedule and the assets or areas it covers. A simple completion rate is completed planned tasks divided by planned tasks due in the period, multiplied by 100. Add an on-time measure when timing matters, using tasks completed by the agreed due date divided by tasks due. Record authorised rescheduling separately from missed, inaccessible, cancelled or technically incomplete work.
Completion should require the evidence defined for the task, not only a changed software status. Depending on scope, that may include date, asset or location, observations, work performed, readings, photographs, defects and the person or provider responsible. A high completion percentage can coexist with missing assets, poor task definitions or unresolved findings. Report schedule coverage and overdue recommendations so the dashboard does not reward administrative closure.
06
Separate response, attendance, restoration and closure
Service-level reporting should distinguish each stage. Response may mean acknowledgement or triage; attendance may mean arrival or remote technical engagement; restoration may mean essential service returned; completion may mean the agreed work is finished; closure may require evidence and approval. Combining these events into one response-time figure hides where delay actually occurs and can create disputes between owners, occupants and providers.
Calculate each measure from its defined start and stop event and show exclusions approved by the contract. Consider the median and ageing distribution as well as an average where a few extreme cases can distort the picture. Never copy universal priority times or targets from another property. Contractual service levels, operating hours, access, parts, specialist dependencies and safety procedures must be agreed for the actual scope.
07
Use reliability and repeat-fault indicators with context
Asset reporting can include downtime, fault frequency, recurring symptoms, open defects and maintenance history for the systems important to the property. Define whether downtime begins at fault report, confirmed loss of function or another event, and whether it ends at temporary restoration or full repair. Record planned outages separately. Without consistent timestamps and asset identity, reliability calculations imply accuracy that the data does not support.
A repeat-fault measure may count valid work orders for the same asset and failure category within an agreed window, but it should trigger review rather than automatic blame. Repetition can reflect incomplete diagnosis, incorrect task coding, operating conditions, access limitations, ageing equipment or linked system problems. Use the facility asset register to connect incidents and evidence. Remaining-life, replacement and engineering decisions require competent assessment beyond a dashboard trend.
08
Add cost, resource, energy and user measures carefully
Cost indicators may compare approved budget, committed spend, invoiced value and forecast using a stated accounting period and responsibility boundary. Separate routine contract charges, materials, specialist work, replacement and projects where the commercial model treats them differently. A lower monthly cost is not automatically better if necessary work was deferred, evidence is incomplete or a different party carried the cost.
Energy, water, staffing, cleaning quality or user-experience measures can be useful when the owner has dependable sources and a clear objective. Normalise consumption only with an agreed basis such as area, occupancy, operating hours or weather information, and document limitations. Surveys need consistent questions and sample context. Utility interpretation, statutory reporting, financial analysis and specialist performance conclusions should remain with appropriately qualified or authorised parties.
09
Show targets, trends and exceptions without traffic-light theatre
Set targets through the contract, property objectives, technical requirements or an approved baseline. Show actual result, target, prior period and a short trend where the data is comparable. Use red, amber and green only with documented thresholds and an explanation of what action each state requires. A green total should not conceal an overdue high-consequence issue, so place material exceptions and unresolved decisions before portfolio averages.
Avoid ranking buildings or providers until definitions, scope and data quality are comparable. A residential tower, retail property and office building can have very different operating profiles. Annotate material changes such as a new provider, asset-register correction, major project, occupancy shift or revised workflow. Comparisons should help investigation and learning; they should not turn incomplete data into unsupported performance claims.
10
Run a controlled monthly review and improvement cycle
Close each reporting period through a defined process: freeze the data cut-off, run quality checks, reconcile exceptions, obtain responsible review and record approved commentary. The meeting should focus on overdue actions, recurring causes, material variance, missing evidence, provider or owner dependencies and decisions with named owners and dates. Carry unresolved actions into the next review instead of replacing them with a fresh slide deck.
Control changes to formulas, categories, targets and sources. Record who approved the change, when it applies and whether prior periods were restated. Retain the detailed evidence behind summary charts and protect confidential property, access, commercial and personal information through suitable permissions. Spreadsheets exported from a live system should be dated and controlled because copies can become outdated.
Start with a small dashboard that people use, then add measures only when they support a real decision and the data can be maintained. Friends Facilities Management can discuss reporting requirements for a Dubai maintenance or facility management proposal. Final scope, data availability, formulas, responsibilities, targets and reporting deliverables must be confirmed in an accepted written agreement. This guide is general information and not financial, legal, regulatory, safety, certification or engineering advice.