Core concepts: signals, decisions, and confidence
How cross-platform data, evidence, ranked recommendations and next steps fit together.
Writes and maintains the Paid Lens documentation, from first connection to workspace administration.
Four steps produce every recommendation you see. Understanding what each one does makes the reasoning trail much easier to read.
Cross-platform marketing data
Normalises data from every connected platform onto one standard: one definition of spend, one definition of a conversion, one currency, one timezone. Without this step, cross-channel comparison is arithmetic on incompatible numbers.
Evidence and reasoning
Joins entities across systems — campaigns, creatives, sessions, contacts, deals — so a closed-won deal can be traced back to the campaign that started it.
Ranked recommendations
Scores candidate actions on expected business impact, statistical confidence, and time-to-signal, then ranks them. This is what makes the queue an ordered list instead of a pile of alerts.
Recommended next steps
Turns a scored decision into a concrete instruction: which campaign, what change, how much, and what to expect. It also tracks the result after you approve, which feeds back into future confidence scores.
How to read a confidence score
- High (80–100): the pattern holds across the full lookback window and enough conversions support it.
- Medium (55–79): the pattern is real but the sample is thin or recent. Consider a smaller test.
- Low (below 55): shown for transparency only, and never placed at the top of the queue.
See also
Run these steps on your own accounts — request beta access
Request Beta AccessPrivate beta — no card, no self-serve signup.