Six-month plan → Principles → Test and Scale

Test and Scale

Explore options cheaply, then commit decisively — especially when reporting logic, definitions or service models are changing.

The principle explained

Cartoon: route-testing expedition metaphor for Test and Scale

Test and Scale means exploring multiple possible routes to success in a cheap, safe, repeatable way, then moving decisively once the evidence is clear.

In the expedition metaphor, the team does not send the whole group onto dangerous ice to find out whether it is safe. They use drones, probes, route markers and small tests to identify the best route. But testing is only half the principle. Once the right route is clear, the team must be ready to move quickly and commit properly — promoting prototypes into products, not leaving pilots running forever.

I would not treat a new data model as automatically meaning the same thing in a different format. I would test whether it still answers the service question — then scale what works.

The principle in practice

In the Band 8a role I would use Test and Scale during system change: run old and new logic side by side, pilot with service managers, park or dismiss what the data cannot yet support, and promote winners into the performance framework. That supports job duties on process design (1.4, 1.6), compliance targets (1.7), pathway improvement (5.5) and person-spec requirements on change management (2.2), developing frameworks (2.3) and leading staff development (3.1).

MHSDS-like submission test

National-style return built from the same episode fact as local products — test inclusion rules before submit, not after NHSE query.

JD 1.7 compliance targets; person spec 1.6 NHS reporting requirements.

Source-to-report map filters

Presets and filters to test lineage views — mandatory vs local, low-confidence rows, migration-risk fields — without rebuilding the map each time.

Demand/capacity pack

Weekly prototype with two CMHT leads for six weeks → promoted to deliverable #4 (IDEA-01). INS-03 dismissed a naive forecast that overshot 18% during definition change.

Reporting assurance merge

Three overlapping registers tested as separate pages, then scaled into one assurance narrative with a clearer publish gate (IDEA-02).

Ideas under test

Portfolio of promoted, parked and dismissed ideas — including manager 1:1 guides (IDEA-05), PIP pathways (IDEA-08), DNA recovery pilots and statistical forecasting parked until data is clean.

Further ideas for a real secondment — line management, training needs from productivity gaps, pathway redesign signals.

← All three principles · Back to six-month approach