Solution · Continuous discovery

Continuous discovery on a panel that remembers.

Keep a standing synthetic panel and run weekly touchpoints. Because personas carry bi-temporal memory, you can see a stance shift over weeks — and what moved it — instead of starting from zero each time.

01Best for

Ongoing discovery, roadmap signal, tracking stance over time.

Output

A running signal: what changed, who shifted, and why — week over week.

Best for

Always-on discovery, roadmap validation, tracking belief drift.

Grounding

Memory persists; get_state_at lets you time-travel a persona’s view.

Anti-pattern

No one-shot survey — the panel accumulates and remembers.

The loop

01

Stand up the panel

A persistent population that simulates days between touchpoints.

02

Weekly touchpoint

A short council on this week’s question — grounded in their week.

03

Track the shift

Memory shows who changed their mind and what caused it.

04

Synthesize the trend

A signal you can take to the roadmap, with the receipts.

02How it works

Runs through d.school Micro-Cycle.

A job is the outcome you buy; underneath, the run follows a proven framework and reaches for one move at a time. Continuous Discovery runs through d.school Micro-Cycle, using council, prototype test along the way — and it won’t converge until the panel below has weighed in.

Coverage

At least 3 personas, spanning segment · lifecycle-stage · recency-of-use. A light, repeatable loop rather than a one-off: a small standing panel that spans lifecycle stages, so signal is comparable cadence-over-cadence.