Definition
Decision-to-Action Latency (DAL) is the elapsed time between the arrival of a meaningful signal and the departure of a meaningful action. Not the time to see the signal — the time to move on it.
Why it matters
DAL is the single number that governs the practical return of every dashboard, model and AI investment. A perfect insight arriving too late has the same value as no insight at all.
Symptoms
- No one in the room can answer "how long did the last decision take?".
- Meetings summarise the past instead of resolving the present.
- The same data quality issue appears on the agenda every quarter.
- Time to first action is not instrumented anywhere.
Typical mistakes
- Confusing dashboard refresh cadence with organisational cadence.
- Optimising the analytical stack while leaving approval chains untouched.
- Measuring throughput of decisions instead of latency of each one.
Examples
Fraud detection surfaces a suspicious pattern within seconds; the block leaves the organisation after the next weekly risk council — five days later. The model is real-time. The organisation is weekly.
How Zero-Gravity approaches it
We name a small number of decision domains, instrument DAL per domain, and set a target latency appropriate to the domain's risk profile. Interventions are chosen for their effect on that latency; results are re-measured and recorded in the Evidence Pack.