85% satisfaction. Looks great on a board pack.
But somewhere in that 85% is a tenant who ticked "fairly satisfied" and then wrote: "I am fairly satisfied because the landlord has not been clear about my rent payments. Last year I was paying £405 and now I am paying £507."
TSM recorded a positive. A sentiment model scored that comment at around 34/100 — firmly negative. The dashboard said "okay." The words said "I'm anxious and I don't trust what's happening."
That gap is where future crises live.
Tenant Satisfaction Measures were designed for consistency and accountability. Same questions, same scale, same format — every landlord, every year. That's genuinely useful. You can benchmark, report to your board, satisfy the RSH. Fine.
But TSM only captures what tenants choose when forced to pick a box. And there's a structural flaw hiding in plain sight: most reports group "very satisfied" and "fairly satisfied" together as a single positive outcome.
They are not the same thing.
Analysis of one benchmark dataset found that "fairly satisfied" tenants often had neutral or negative average sentiment across most TSM categories. Only "very satisfied" tenants showed genuinely positive emotional tone. The "safe middle" isn't safe. It's a pool of quietly unhappy residents who haven't complained yet.
One dataset tracking both metrics from early 2023 to mid-2025 showed something striking: sentiment dipped sharply in one spring period while headline TSM scores stayed relatively steady. Six months later, TSM scores dropped to match.
Sentiment moved first.
That's not a guarantee of predictive power — but it's a pattern worth paying attention to. By the time your annual TSM tells you something's wrong, the problem has had months to compound. Tenants don't complain formally straight away. They get quietly frustrated. They stop picking up survey calls. They disengage.
The value of feedback is inversely proportional to its latency. Annual TSMs are operational hindsight dressed up as intelligence.
The combination of TSM and sentiment isn't complicated to use once you've got it. You treat the score as the headline and the sentiment as the narrative underneath.
Analysis of one tenant benchmark identified the drivers precisely: quick response times, reliable service, and staff empathy added roughly 40 points to perceived experience when present. Security issues, long appointment waits, and conflicting information dragged sentiment down by 25–36 points.
Crucially, the highest-volume theme — repairs — wasn't the worst performer emotionally. Lower-volume issues like safety concerns and waiting for appointments produced much lower sentiment, despite fewer mentions. Without sentiment, you'd be throwing resource at the noisiest category, not the most damaging one.
Tactical, not just busy. That's the shift.
When you can show a TSM percentage alongside a sentiment trend, the story becomes a lot easier to tell.
The TSM number shows how you compare to peers this year. The sentiment trend shows whether your actions are actually moving tenant experience in the right direction — and where the next risk is likely to surface. Regulators want the score. Your board wants to know what it means. Sentiment gives you the answer to the second question.
Tenants don't experience their home as a percentage. They experience it as relief, worry, irritation or gratitude. Pairing TSM with sentiment doesn't just make the data richer — it makes it honest.