MCMC (Markov Chain Monte Carlo)
A class of algorithms (Metropolis-Hastings, Gibbs, NUTS, HMC) that draw samples from a probability distribution by constructing a Markov chain whose stationary distribution is the target.
Last updated: 2026-05-04
Definition
MCMC sits inside Bayesian MMM as the engine that approximates the posterior over channel contributions. WatEase uses 2,000 draws after 1,000 tuning steps per chain across 4 chains — the convention from PyMC and Stan tutorials, calibrated against marketing data's typical noise level. You read the output as a histogram of plausible per-channel coefficients; the mean is the point estimate, the 2.5% and 97.5% percentiles are the 95% credible interval. Convergence is diagnosed via R-hat (should be < 1.01) and effective sample size (should be > 400).
How it applies in India
No India-specific behavior — MCMC is provider-agnostic.
Frequently asked questions
Why MCMC and not just gradient descent?
MCMC samples FROM the posterior; gradient descent only finds the maximum a posteriori (MAP) point. For Bayesian MMM you want the full distribution — that's where the credible interval comes from.
Related terms
- Bayesian MMMMarketing Mix Modeling implemented with Bayesian inference (typically MCMC sampling), producing posterior distributions over channel contributions instead of point estimates.
- Credible IntervalA range that contains the true parameter value with a stated probability (e.g., 95%). The Bayesian counterpart to a frequentist confidence interval.
- Prior DistributionIn Bayesian inference, the probability distribution that encodes belief about a parameter before observing the data.
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