Skip to lesson
Exit
Data, SQL & Measurement1 / 2

1 min lesson

Detecting drift before it costs a quarter

Choose two examples from the table in "Detecting drift before it costs a quarter" and explain what each teaches you to do.

Step 1 of 2

Detecting drift before it costs a quarterthe silent failure mode

Pipelines rot quietly. A data provider changes its schema and match rate slips from 75% to 60% over three weeks; nobody notices until pipeline dries up. Drift detection means the metric watches itself.

Drift signal
Match rate drops week over week
Likely cause
Provider schema or coverage change
What the loop should do
Alert; fall back to next provider in the waterfall
Drift signal
Qualified rate climbs but conversion falls
Likely cause
Scoring inflated; threshold too loose
What the loop should do
Re-fit weights against recent outcomes
Drift signal
Routing SLA hit rate decays
Likely cause
Volume spike or owner capacity change
What the loop should do
Re-balance round-robin; alert on backlog
Drift signal
Enrichment cost per lead rises
Likely cause
Cheap provider missing more; falling to pricey one
What the loop should do
Re-order the waterfall by cost-adjusted hit rate

Each signal maps to an automated response, not a quarterly review.

Learn more

Full explanation

Experiments with honest readouts

Experiments with honest readoutsA/B on routing and messaging

When you A/B a routing rule or an outreach variant, the discipline is the same as any experiment. Randomize at the lead or account level, pre-register the metric and resist peeking until you have the sample to call it. A directional read stated as a directional read beats a false certainty.

  • Randomize at the right unit (account, not lead, when reps work whole accounts) to avoid contamination.
  • Pre-commit to the primary metric and the minimum sample before you launch, so you can't retrofit a winner.
  • Report the effect with its uncertainty and call out when the sample is too small to conclude anything.