Use Leading and Lagging Indicators

A lagging indicator tells you what happened. It is the truth and it arrives too late to act on.

A leading indicator moves earlier and is a bet: you believe it predicts the lagging one. It is actionable precisely because it is not yet the truth.

Neither is useful alone. A lagging indicator alone gives you an accurate autopsy. A leading indicator alone gives you something to optimize with no evidence it matters.

State the lag

If you claim one predicts the other, say how long the gap is. Without it nobody can tell whether this month's leading number should have shown up yet, so every conversation about whether it is working is unresolvable.

The lag is also the thing most likely to be wrong, and stating it is what lets somebody find out.

Be honest about the relationship

The pairing is a hypothesis, not a fact. The honest form of the claim is: we believe this leads that, here is the lag we expect, and here is what we would see if we were wrong.

Two metrics can also both be correct and simply answer different questions with no predictive relationship at all. A record can be maximal on a momentum measure and poor on a qualification measure, and pairing those as leading and lagging would be inventing a relationship that does not exist.