What past tactic records can and cannot tell you
Navigator ranks tactics with a playbook built from Outrigger research and, where they exist, records of tactics applied in the past. This lesson explains what those records are, so you weigh them correctly.
What was recorded
Earlier versions of Outrigger saved a record when an approved tactic was applied: the client, the tactic, a snapshot of the client's scores at that time, and later a second snapshot. New tactic outcomes are not being measured at the moment.
Why it is not proof
A score that moved after a tactic does not show the tactic caused the change. Other work, model updates and seasonality all move scores, and the records have no comparison group. Small samples are shrunk so one unusual result cannot dominate, but that only reflects how many records exist, not how sure anyone can be that a tactic works.
How the agent uses it
When records exist, the agent sees them as a descriptive ranking next to the playbook. It treats them as a hint, never as a forecast, and it does not promise score or visibility gains.
What this means for you
- Judge each proposal on its evidence. The rationale should point to something observed for this client, such as a missing profile or a thin review base.
- Use your own knowledge. If you know something the records cannot show, act on it and note why.