Here's what the RSH judgements are really testing.

1. Before asking what the data says, ask if it's complete.

Several C2 landlords had surveys, compliance reports and improvement work in place, however they couldn't provide assurances on the reliability of the data which sat underneath this. A dashboard can report 100% of the records it can see, whilst it may be missing homes, hazards and actions entirely.

Consider: What information could be missing here, and how would we know?

2. Swap reassurance dashboards for challenge dashboards.

Headline percentages hide the oldest cases, the repeat failures and the customers getting the worst of it. Instead, show the overdue actions, repeat visits, failed checks, vulnerability.

Consider: Could a non-exec or councillor spot the highest-risk cases without a separate briefing?

3. Turn customer engagement into a decision trail.

Everyone can point to surveys and panels. Far fewer can demonstrate what customers said, who weighed it, the decision it changed and the outcome that followed. Counting meetings is not evidence of influence.

Consider: Can we name three recent decisions that changed because of customer influence and can we evidence the difference it made?

4. Use damp and mould to test the whole machine.

It's where data quality, contractors, communication, vulnerability and legal timescales all collide, which in my opinion makes it the best stress test you can have. Eastlight's move from C2 to C1 is evidence of this with their grade shifting once they showed sustained improvement, not just a plan.

Consider: Can we evidence the timescale and the quality of the outcome on a case by case basis?

5. Test the joins, not just the middle.

The biggest risks often sit where your organisation meets someone else's, such as running a repairs service or supplying you data. You can hand over the work, but you can never hand over the responsibility.

Consider: If a key partner stopped sending us data tomorrow, or we found out it was wrong, how quickly would we actually know and act?

6. Agree what "success" looks like before you launch the plan.

Completing an action isn't an improved outcome. C1 landlords weren't flawless and plenty of C2s understood their weaknesses and had plans. The difference was whether improvement was verified and sustained. Therefore, design the closure test at the start. Think about baselines, targets, quality threshold, how long it must hold, who assures it.

Consider: What would prove this worked for customers and kept working after priorities moved on?

I believe these judgements don't sort landlords into good and bad. What they demonstrate, is the regulator testing the systems that connect homes, customers, services and leadership.

There's no such thing as a perfect inspection, therefore C1 landlords have gaps and C2 isn't crisis. The difference is whether you can identify weakness, control the risk, prove improvement and stop it recurring.