Grafton.
Project Barbara · The Intelligent Letting Agency

The Operational KPI Framework

The operating instrument panel · Book Health 4-RAG model · 10 June 2026

The instrument panel of the Intelligent Letting Agency — and the schema of the digital twin's first derived views. Built on Glenn Flegg; deliberately target-agnostic: run it over every acquired book from the BQ onward. Worked instrument: GF_Operational_KPI_Framework.xlsx.

1What it is

  1. Deal Snapshot — the memo front page from live data (revenue, recurring %, maintainable EBITDA, book, team, avg rent/commission, sales division).
  2. KPI Dashboard — four areas (per the Grafton Capital SaaS framework), lead and lag.
  3. Computed Now — live formulas off blue inputs: CAC/LTV, NRR formula, Rule-of-40 (~34% at GF), concentration (≤2.8%), the deal-economics block.
  4. Book Health — the 4-RAG state model (the heart of it — below).
  5. Early Warning — pre-mortem: failure modes × lead metrics × thresholds.
  6. Data Map — BQ now → Xero monthly → Veco at DD → the event log from day one; historic rebuild plan.

2Book Health — the 4-RAG state model

Four RAG states per entity — Tenant, Property, Occupancy, Landlord — with criteria derived from countable facts (= event-log queries, not opinions). The landlord RAG is revenue-weighted. The cross-reads do the diagnostic work:

Weekly snapshot grid from the DD baseline; velocity rows (new ambers/reds per week) are the trend signal. The monthly landlord sentiment pulse (process catalogue) feeds the landlord RAG directly.

3Early warning — the pre-mortem

Five failure modes, each with lead metrics and thresholds: notice count (weekly) · valuation-requests-from-the-book (a churn lead indicator) · silent-landlord % · automation coverage must lead headcount reduction · client-money reconciliation breaks (= daily red) · founder-attention cadence · GM-seat days vacant. Plus three DD-computable structural risks: landlord age/tenure mix (>30% over-65/15y+ = decay risk) · EPC D-or-below % (the MEES sell-up wave) · rent-vs-market gap (>10% under = S13 catch-up collides with tenant churn).

4Why it matters to the platform

The RAG criteria are event-log queries — which makes this framework the first concrete spec of the platform's derived views: build the mirror/event log, and Book Health is a view over it.

It is also the DD instrument: the baseline census is computed during diligence and becomes the day-one operating state. See the technology roadmap §3 and the process catalogue.