As mentioned above it is true that Markov Chain Method may be used quite successful in forecasting/predicting brand switching, yet it has its own drawbacks and limitation, which are :

1. Customers do not always buy products in certain intervals and they do not always buy the same amount of a certain product each time. In practical terms it means that in the future, two or more brands may be bought at the same time.
2. Customers always enter and leave markets, which imply that markets are never stable.
3. The transition probabilities of a customer switching from the  brand to the  brand are not constant for all customers. Moreover these probabilities themselves may change from customer to customer and from time to time. Thus the transitional probabilities may change according to the average time taken between two purchases.
4. The time between different purchases may be a function of the last brand bought.
5. It is best not to consider other marketing environment such as sales promotions, advertising, competition etc., as it becomes very cumbersome to handle the combined effect of these variables in our Markov Chain model.

Consider we want to analyze the brand loyalty for cosmetics used by females in urban India who are in the age group of 20 to 40 years. In order to do this study, assume we conduct a marketing survey with a set of questionnaire which has say  number of questions. Also consider the number of consumers who are asked to answer these questionnaires is  in number. Finally the number of brands of different cosmetics is . Also assume that the questions which are used to give us some information about the behaviour about the consumer is based on a Likert scale of 1 to 5, where the number 1 denotes strongly disagree while 5 implies strongly agree. The questions one can ask in the questionnaire may pertain to income, age, present brand of cosmetics used, preference and liking between the  number of brands of cosmetics, etc.