A GENERALIZED FLOW FOR B2B SALES PREDICTIVE MODELING

Abstract

The paper and packaging company that provided the data for this research has a long history of sales expertise.
This expertise is captured predominantly in the intuition of sales representatives, many of whom have worked in
the industry for 20 years or more
AIM:
Intuition is not easy to record and disseminate across an entire sales force, however, and thus one of the
company’s most valuable resources is inaccessible to the broader organization. As a result, the company tasked
this team with extracting the most important factors in driving sales success and modeling win propensities
using data from their customer relationship management (CRM) system.
OBJECTIVE
Most prior work in this space has been performed by private companies, both those that have developed
proprietary technologies for internal use and those that sell B2B services related to predictive sales modeling. As
a result, research in the field is typically unavailable to the public. Some examples include Implisit —a company
recently acquired by Salesforce.com that focuses on data automation and predictive modeling—and Insight
Squared, which sells software that includes a capability to forecast sales outcomes.
ABSTRACT:Predicting the outcome of sales opportunities is a core part of successful business management.
Conventionally, making this prediction has relied mostly on subjective human evaluations in the process of sales
decision making. In this paper, we addressed the problem of forecasting the outcome of business to business
(B2B) sales by proposing a thorough data-driven Machine Learning (ML) workflow on a cloud-based
computing platform: Microsoft Azure Machine Learning Service (Azure ML). This workflow consists of two
pipelines: (1) An ML pipeline to train probabilistic predictive models on the historical sales opportunities data.
In this pipeline, data is enriched with an extensive feature enhancement step and then used to train an ensemble
of ML classification models in parallel. (2) A prediction pipeline to utilize the trained ML model and infer the
likelihood of winning new sales opportunities along with calculating optimal decision boundaries. The
effectiveness of the proposed workflow was evaluated on a real sales dataset of a major global B2B consulting
firm. Our results implied that decision-making based on the ML predictions is more accurate and brings a higher
monetary value.