Bosca / Recommendations

Personalization signals

What the model knows about a person

A signal is a rule that turns something you already know about a profile — an attribute, a segment — into a typed value the recommender can use. Signals are how you tell the model what matters, without writing any code.

Deriving a value

Attribute in, signal out

You pick a source — a profile attribute type or a segment — and write a short rule that reads the value along with how much to trust it. Gate on confidence, gate on whether it's verified, or reshape it entirely. A signal that comes back empty simply sits out; it never blocks a save.

  • Author the rule in Studio and preview it against a sample before it goes live.
  • Attribute-sourced signals are computed and cached on the profile, and recomputed automatically when the attribute changes.
  • Change a definition and the platform backfills every profile it touches in the background.
signal · favorite topic
profile attributetopic affinityruleconfidence > 60signal"nature"

Typed values

Four shapes of signal

A signal has a type, and the type decides how the model reads it — and whether it can also group people.

Categorical

one value

A single label — a favorite topic, a tier. Can also key a cohort.

Multi-categorical

many values

A set of labels — the topics someone follows. A model feature.

Numeric

a number

A measured value, bucketed and normalized before the model sees it.

Boolean

yes / no

A binary flag — subscriber or not. Can also key a cohort.

Two jobs

A feature, a cohort, or both

Every signal can do two things, and you choose which. As a feature, it feeds the model's picture of a person, sharpening their ranking. As a cohort dimension, it joins a compact key that powers "people like you" — so viewers who share signals share behavioral recommendations.

  • A viewer with no cohort signals falls back to global co-engagement — never a dead end.
  • The cohort key is deterministic and bounded, so "people like you" stays stable and cheap.
signal roles

Feature

Feeds the model — sharpens this person's ranking.

Cohort

Groups similar people for "people like you".

Learned interest

Behavior becomes a signal too

Signals aren't only about what you already store. When someone rates content highly, that interest is written back to their profile as a learned attribute — with a confidence that grows the more they repeat it, but never reaches certainty. Over time, a person's own behavior personalizes their feed.

  • Learned interests are ordinary attributes, so they can feed signals like any other.
  • Repeat engagement reinforces confidence; a single rating never dominates.
learned over time
Nature70
Photography55
Long Reads85

Keep exploring

More on Recommendations