Using Markov Chain to Forecast Sales Booking

‘We Sell Everything in Software’ – WSES Inc. is a products company and specializes in software solutions for different industries such as defence, clinical research, consumer goods, capital markets, security, banks, and insurance among others. One of the divisions of WSES focuses on enterprise software product. Every quarter, Jack Williams, CEO, had to give a forecast of sales to the stakeholders for the enterprise software product division. The forecast which he had given for the last quarter was USD 2.4 billion whereas the actual sales booking was only USD 1.48 Billion. Jack wanted more accurate forecasting of sales and he had a discussion with Michael Summers, the CFO. Michael explained to Jack that this was something which was not in his hands since he was taking the numbers from Ben Osborne, Vice President of Marketing. Ben explained to Jack that the process they were following was taking the last quarter’s sales and adding their estimate of 1.5% to it. Jack did not approve of this method. He felt that since WSES has such rich sales data, over the years, they should be having a way to hear what the data is saying. They engaged Mark, with a Ph.D. in Statistics, to understand if they could find a structured way to forecast the sales number based on historical data available with WSES.