Free Resources:
Markov Chain is a stochastic process that can be used to predict future consumer choices when past purchase history of individual consumers are not available (because product purchase history is not tracked adequately, or because the costs involved in extracting such information from legacy systems are prohibitive).

For example, let's assume that your company sells two major types of products  (X and Y), plus a variety of other products, which are grouped together as Z.

About % of new customers purchase Product X during their first visit to your company, about % of new customers purchase Product Y during their first visit to your company, and the rest purchase Product Z during their first visit to your company.

Among customers who purchase Product X, % will purchase Product X again on their next visit, % will purchase Product Y on their next visit, and the rest will purchase Product Z on their next visit.

Among customers who purchase Product Y, % will purchase Product X on their next visit, % will purchase Product Y again on their next visit, and the rest will purchase Product Z on their next visit.

Among customers who purchase Product Z, % will purchase Product X on their next visit, % will purchase Product Y on their next visit, and the rest will purchase Product Z again on their next visit.

With the above information, Markov Chain can be used to predict which type of product (X, Y, or Z) customers are most likely to purchase on their n visit to your company (where n can be 2nd, 3rd, 4th, etc.). To see Markov Chain in action, press the button below. For those who wants a more technical explanation of the Markov Chain, please follow this link.

JTK Research Home Page Market Research Applied Statistics Customer Relationship Management E-Solutions About/Contact JTK