Offer Decisioning and Personalization
Core mental model
Offer Decisioning selects the best offer for a profile using decisioning objects: placements (where an offer appears), collections (groups of offers), decision rules (eligibility constraints), ranking (priority/weighting or ranking formulas), and fallback offers. Personalization uses the same profile data through expressions in messages.
Concepts candidates must know
- Offer library: offers with eligibility rules, representations (content per placement), and constraints (frequency capping, dates).
- Decision rules evaluate profile attributes/events; ranking decides among eligible offers; fallback covers no-eligible-offer cases.
- Decision scope combines placement + collection + ranking + fallback for a given request.
- Personalization editor: expression syntax, helper functions, and previewing with test profiles.
- Integration: a journey action can request a decision for a profile; the returned offer is injected into the message.
Traps
- Assuming the highest-priority offer wins without checking decision rules and capping - an offer can be eligible but capped.
- Forgetting the fallback offer: if none is configured and no offer is eligible, the placement returns nothing.
- Confusing collection (a curated set) with decision scope (the full selection configuration).
- Personalization previews use a test profile; a blank preview often means the field isn't in the profile schema, not that the syntax is wrong.
Worked rule of thumb
For "which offer is shown," evaluate in order: eligibility (rules + dates + capping) → ranking among eligible → fallback if the eligible set is empty. If the answer options differ by ranking vs rules, identify which stage the scenario describes.