Imagine you are a company selling a fast-moving consumer good in the market.<\/p>\n
Let\u2019s assume that the customer would follow the given journey to make the final purchase:\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 \u00a0 \u00a0 \u00a0 Now, how to find out in which state the customers would be after 6 months?<\/p>\n Markov Chain comes to the rescue!!<\/p>\n \u00a0Let\u2019s first understand what Markov Chain is.<\/p>\n \u00a0<\/p>\n Markov Chains:<\/strong><\/span><\/p>\n A\u00a0Markov chain<\/strong>\u00a0is a\u00a0stochastic model\u00a0describing a\u00a0sequence of possible events in which the probability of each event depends only on the state attained in the previous event<\/p>\n \u00a0<\/strong><\/p>\n Markov Chain \u2013 States, Probabilities and Transition Matrix<\/strong><\/p>\n Let\u2019s delve a little deeper.<\/p>\n A Markov Chain provides:<\/p>\n \u00a0\u00a0 \u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0&\u00a0<\/p>\n \u00a0<\/p>\n Using the above two information, we can predict the next state.<\/p>\n \u00a0<\/p>\n In mathematical terms, the current state is called Initial State Vector.<\/p>\n \u00a0<\/p>\n So, what we get is:<\/p>\n Final State = Initial State *Transition Matrix<\/strong><\/p>\n \u00a0<\/strong><\/p>\n Classic Example<\/strong><\/p>\n A classic example of Markov Chain is predicting the weather. We have two different weather conditions: Sunny and Rainy. Let\u2019s assume today it is sunny. We have the following probabilities:<\/p>\n \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 Source: Wikipedia<\/a><\/p>\n<\/p>\n Here the initial vector is =\u00a0<\/p>\n Transition Matrix = \u00a0\u00a0\u00a0\u00a0<\/p>\n <\/a> \u00a0<\/p>\n \u00a0<\/p>\n Recollect the\u00a0Final State = Initial State *Transition Matrix?\u00a0<\/strong>The above represents the same.<\/p>\n So, what is the inference?<\/p>\n There is a 90% chance that the weather will be sunny on Day 2 and 10% chance that it will rain.<\/p>\n \u00a0<\/p>\n Back to the Problem<\/strong><\/p>\n Coming back to the problem where we need to know what the state the customer is after 6 months of launching the product.<\/p>\n We can assume there are 4 states in which the customer can be at any point in time.<\/p>\n<\/p>\n We have the following information:<\/p>\n \u00a0<\/p>\n The Marketing Analytics objective:<\/strong><\/p>\n \u00a0<\/p>\n So, lets dive into the math part.<\/p>\n Note:\u00a0 A \u2013 Awareness, C \u2013 Consideration, P \u2013 Purchase, NP \u2013 No Purchase<\/p>\n \u00a0<\/p>\n Initial State Vector =<\/p>\n <\/a><\/p>\n Transition Matrix =\u00a0 It would be clearer to see the movements among all 4 states diagrammatically.<\/p>\n<\/p>\n Final States of Customers = Initial State Vector * Transition Matrix<\/p>\n Evaluation of the Result<\/strong><\/p>\n Now let\u2019s evaluate our results.<\/p>\n Initial Vector<\/strong><\/p>\n<\/p>\n Final Vector<\/strong><\/p>\n Overall our analysis goes to show that the campaign\/ads has worked!!<\/span><\/strong><\/p>\n Markov Chain has many other applications in Marketing Analytics and other fields such as NLP.<\/span><\/p>\n Stay tuned for more articles\u2026.<\/span><\/p>\n <\/a><\/strong><\/p>\n <\/a><\/p>\n<\/p>\n<\/div>\n Go to Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<\/a>\u00a0These are the states at which the customer would be at any point in the purchase journey.<\/p>\n
\n
\n
\n
\n
<\/a><\/p>\n
<\/a><\/p>\n
<\/a>Weather on Day 2 =<\/p>\n
<\/a><\/a><\/p>\n
\n
\n
\n
<\/a><\/p>\n
<\/a><\/p>\n<\/p>\n
<\/a><\/p>\n<\/p>\n
<\/a><\/p>\n
<\/a><\/p>\n
<\/a><\/strong>We can notice that the number of people under \u2018Awareness\u2019 and \u2018Consideration\u2019 have decreased. This is a good thing because, the people actually shifted from the state of \u2018Awareness\u2019 and \u2018Consideration\u2019 to the state of \u2018Purchase\u2019 (an increase of nearly 34%\u00a0!!) Also notice that the number of people in \u2018No Purchase\u2019 states decreased (a decrease of 11%).<\/span><\/p>\n