This paper presents the forecast analysis of a seasonal product using the concept of Bayesian Dynamic Linear Model (DLM). In the dynamic linear model, the demand of a period is related to the immediate past period through a link function and the forecast is updated at each time step. The Bayesian framework permits the joint estimation of the model parameters using the prior probability distribution assigned to each parameter and the historical data. The parameters of the model are updated in a Bayesian process using the Markov Chain Monte Carlo (MCMC) sampling from the posterior distribution with an efficient Gibbs sampling algorithm. The forecast results show that the average error and tracking signals of the forecasts are within 0.1 and 0.25 limits, respectively. It is implicit that the ability to integrate the updated demand information through the Bayesian techniques reduced the forecast errors and improved the efficiency.