Applying Monte Carlo simulation to model a sales process for forecasting future sales: a case study for a Finnish recruitment consulting company

Abstract

For years, sales forecasting has been seen as an important base for a company’s operative planning process. Effective sales forecast has the power to change the perspective of the whole company’s business model from reactive to proactive. There exists a wide and colorful literature regarding various sales and demand forecasting methods, their usage, and impact for the business from various angles. In addition, its importance for customer relationship management has been identified. However, in the academic literature there has scarcely been discussion on how to actually utilize company’s internal CRM systems to aid in sales forecasting.

In this thesis, I aim to contribute to this topic by modelling a multi-stage sales process of a Finnish recruitment company by utilizing their internal CRM data. Modelling is done by using Monte Carlo Simulation of repeated random sampling. The resulting model can be used to analyze the whole process and its uncertainties as a whole or dig deeper into a specific sales stage. Furthermore, it can also be used to forecast short-term sales volumes.

The proposed simulation model will be benchmarked to a number of commonly used quantitative forecasting methods and the results show that the simulation model outperforms them in terms of forecasting accuracy and forecasting bias. The dual-sided role of the model is to aid in forecasting and act as a decision-support-system (DSS) when the case company is assessing their resourcing alternatives and their probable outcomes throughout their value chain.