Benchmark Network
See how you compare — without your data leaving your walls.
Most services have no honest way to know whether they are doing well. The Benchmark Network lets you compare against similar services on the measures that matter — using de-identified, aggregated figures, so your raw patient data never leaves your own tenant.
How it stays private
Compare, without pooling data.
Your data never leaves
Comparisons are computed on de-identified, aggregated figures. Raw patient records stay inside your own walls — they are never pooled or shared.
Aggregated and protected
Only summary numbers contribute, with privacy protections applied so no individual patient — or your organisation's raw detail — can be picked out.
Opt-in, and reversible
You choose whether to take part and on which measures. You can change your mind and withdraw at any time.
What you can benchmark
The measures that actually matter.
Operational
How you compare on the things that run a service well — time to first contact, pathway adherence, overdue work, time-to-action on alerts.
Engagement
How patients respond — appointment attendance, digital response rates, outreach effectiveness — against similar services.
Outcomes
Where it is meaningful and consented, how outcomes for comparable cohorts compare with peers, so you know where to focus.
Federated learning
The models get better together.
Beyond comparison, taking part can make the predictions sharper for everyone. The models learn from collective, opt-in outcomes across organisations — but the learning travels, not the data. No raw patient record ever leaves your tenant to make that happen.
It means a smaller service benefits from patterns seen across the network without giving anything away — and the whole network gets more accurate at spotting risk early. This is one of the things a single, isolated system simply cannot do.
Benchmark Network
Know where you stand — and improve with the network.
The Benchmark Network is opt-in and opens with early access. Talk to us about taking part and what you would want to measure.