Prior Distribution
In Bayesian inference, the probability distribution that encodes belief about a parameter before observing the data.
Last updated: 2026-06-10
Definition
Priors are the BEFORE-data belief; the posterior is the AFTER-data belief. Industry-informed priors are calibrated against anonymized aggregates from comparable verticals — for a new D2C beauty brand with no historical data, the prior on Meta ROAS comes from the 25th-75th percentile of D2C beauty advertisers. After 90 days of the customer's own data, the prior is washed out and the posterior reflects only their data. Per-vertical priors are how MMM systems solve the cold-start problem so first-week recommendations are not blocked on data.
How it applies in India
India-specific priors are calibrated separately because Meta's post-iOS-14.5 attribution behavior + Indian conversion rates differ from US baselines.
Related terms
- Bayesian MMMMarketing Mix Modeling implemented with Bayesian inference (typically MCMC sampling), producing posterior distributions over channel contributions instead of point estimates.
- Posterior DistributionIn Bayesian inference, the probability distribution over a parameter after combining the prior with observed data via Bayes' theorem.
Try WatEase free
Run your WhatsApp commerce on the platform built for India — Cloud API, GST invoices, UPI checkout, opt-in tracking, and a Free Forever plan.
Start Free Today