We have run data engagements with twenty housing associations between 2023 and the end of 2025. The board-level conversation has shifted. Five years ago, it was a cost-of-ownership conversation — "how much does BI cost us". Today it is a regulatory and safety conversation. This article is the pattern of questions we hear, and the posture that tends to satisfy them.
Question 1 — "Do we know, for every property, its current damp-and-mould risk?"
Since Awaab's Law and the subsequent regulatory tightening, this is the question that will not go away. The honest answer from most associations is "partially". The data exists — repairs history, surveyor notes, tenant contact logs, EPC data — but it is scattered across Orchard, NEC, QL, Active H, Aareon, and a customer contact system, and it has never been assembled into a per-property signal.
What a satisfactory answer looks like: a per-property damp-and-mould risk score, refreshed monthly, triangulating repairs, complaints, property characteristics and surveyor visits. Surfaced in the housing officer's workflow, not on a dashboard nobody reads. Auditable — every score can be explained by the underlying evidence.
Question 2 — "Can we evidence our tenant-satisfaction measures?"
The Regulator of Social Housing's TSMs are mostly measurable, but the evidence chain is brittle. Most associations can produce a number; few can reconstruct exactly how the number was computed, from which tenants, excluding which, over which window. When the Regulator asks — and they do, quarterly — the answer should not take three weeks to produce.
The posture: one source of truth for survey data, lineage captured to the final TSM, and the calculation under a semantic layer so the definition cannot drift.
Question 3 — "What is the real condition of our stock?"
Every board wants this and every board knows the answer is patchy. Stock condition surveys are typically five years old. The component-renewal database is maintained to varying standards. The EPC estate is a statutory record but not operationally useful.
The movement we have supported: a consolidated asset register, updated quarterly from the operational systems, with a confidence score per attribute. Not a one-off survey; a living record. Add a 1% sample quality audit and you have something the board can trust.
Question 4 — "Are we confident about our repairs-service performance?"
Responsive repairs KPIs — appointment kept, first-time fix, average completion — are measured by every association. What is less common: the ability to ask "why did these 400 jobs take longer than the SLA" with a clear, drillable answer in the repairs flow. The data is there; it is almost never joined to the rest of the estate (stock, tenant, contractor).
Question 5 — "What is our exposure to the AI-and-automation narrative?"
The newer board question. Boards want to know three things: first, which of our AI use cases are live and what is their residual risk; second, what is our vendor dependency exposure; third, what is the pace at which we can adopt safely. There is no single answer; what boards respond well to is an AI register, aligned to the risk appetite, reviewed quarterly.
The pattern behind all five questions
Each question is nominally about a different topic. Each has the same underlying answer: data assembled from many operational systems, governed, owned, surfaced in the flow, auditable. The housing associations doing this well spend modest money and staff it with a small, permanent data team. The ones struggling have bought a platform and called that the strategy.
What we have learned running these engagements
- Start with one live question. Damp and mould, usually. A visible, urgent question unlocks the governance and platform conversation.
- Build on what exists. The finance and housing systems are not the enemy; the patterns we ship layer on top, not replace.
- Do not let the board see a raw dashboard. Board papers need narrative and caveat. A dashboard without interpretation causes more confusion than it resolves.
- Stewardship beats tooling. A named data steward per domain (repairs, income, voids, asset condition) is worth more than another platform.