FM contract KPI reporting: what public buyers actually ask for

What this guide covers

Most FM bid teams write their KPI response from memory of the last contract. That works until a buyer defines "complete" differently, measures attendance from a different timestamp, or ties money to a KPI you treated as a nice-to-have. This guide does three things:

1. Compiles the KPIs that UK public buyers have actually written into published FM and building maintenance documents, with each target quoted as published and linked to its source.
2. Sets out the data model a contractor needs to calculate those KPIs from job, asset, helpdesk and contract data, with a formula for each.
3. Works one fictional month through to a service credit, and shows the matching monthly report.

It does not cover the detail of KPI publication under the Procurement Act 2023. The Act brought in publication of KPI performance for some contracts; which contracts, and when, is a separate topic.

The KPIs public buyers publish

The table below is compiled from six published documents: a Crown Commercial Service framework specification, four buyer specifications or KPI schedules, and one council performance report. Targets are quoted as the document states them. Where a document gives a measure but no target, the target column says so.

Three patterns are worth taking from this into a bid.

Completion includes the paperwork. City of London's building repairs contractor failed its PPM KPI because of "the inconsistency of compliance documentation loaded onto the City's CAFM system", not because visits were missed. If the certificate is not uploaded, the task is not complete in the buyer's eyes.

Money is usually tied to a minority of KPIs. The FM24 table marks some KPIs as carrying a financial penalty and leaves others (routine PPM, portal take-up, social value) unpenalised. Kent uses failure deductions counted per incident, weighted by severity. National Museums Liverpool assigns each KPI a service priority from Low to Super, with the deduction values left for the buyer to insert.

The buyer expects the report from your system. The RM6232 specification requires the supplier to input operational KPI data into the CAFM system, measure each KPI monthly, calculate payment mechanism scores, and report at region, establishment and service level, including reasons for failure.

The data model behind the report

A monthly KPI report is a set of counts over four kinds of record. Get the joins right and every KPI is a numerator over a denominator.

The joins:

The formulas, for a reporting month M:

Two definitions cause most disputes. Is a P3 job "due in M" when it was logged or when its deadline falls? Does "working day" exclude bank holidays? Write the answer into contract_kpi.exclusions and the report becomes repeatable.

Worked example: one month, end to end

This example is illustrative. The council, its four sites, every job and the service credit mechanism are fictional. The priority bands copy the Penlee House structure and the weighted percentage structure echoes the FM24 table, but the at-risk pot is our assumption: published schedules we read listed per-KPI percentages without always stating what they apply to.

Hypothetical mechanism: monthly charge £84,000; 10% at risk, so £8,400; each KPI carries a weight; a KPI below target forfeits its full weight that month.

Credit: £8,400 × (25% + 20%) = £8,400 × 45% = £3,780. Net payable: £84,000 − £3,780 = £80,220.

The two statutory misses show why the certificate field matters: one legionella temperature check was done three days late, and one emergency lighting test was done on time but had no test record attached at month end. Under a "completion and document upload" definition both count. The P1 failure was water ingress at a depot intake room, made safe 4 hours 51 minutes after the call. One job out of 18 cost £1,680, which is why P1 volumes this low deserve daily monitoring rather than a month-end surprise.

The site breakdown also shows problems the contract-level figures hide. Harbour Leisure Centre's first-time fix is 88.3% (53 of 60) and its P2 and P3 results are below target, even though every one of those KPIs passes overall.

Building it from your FSM data

The report above needs data from at least three places: the job system (PPM and reactive), the helpdesk or telephony export, and the contract schedule, which usually lives in a spreadsheet the bid team built. Survey scores often sit in a fourth tool. The manual version is a monthly export, a VLOOKUP between job and ticket references, and an argument about which timestamp counts.

The IVP.ai Data Layer connects those sources (anything with a REST API, any SQL database, and Excel files such as the KPI schedule), joins them on the keys above, and keeps the definitions in one governed place. A report or agent then runs on that data: the monthly KPI pack, or a daily Deskless by IVP.ai check that flags a P1 approaching its deadline or a statutory task without a certificate before the month closes. A national fire safety and security company used the same approach to unify HubSpot, Uptick and finance spreadsheets into one governed view, saving more than 10 hours a week on reporting (case study).

Related pages: the PPM compliance report template covers statutory PPM in more depth, field service KPIs covers the internal weekly pack, and engineer utilisation covers the resourcing side. More at field service reporting.

If you want to see your own contract's KPIs calculated from live data, the free 48-hour build connects up to three of your sources, cleans the data and sets one report or agent running within 48 hours of read-only access. You keep access for 14 days, then choose a plan or walk away.