IBM Corporation: Analytics-Driven Solutions for Increased Sales Force Productivity at IBM
The Problem
Improving sales force productivity can be an effective operational strategy for companies seeking to drive top-line revenue growth and manage bottom-line expenses in today’s challenging economic climate. At one level, driving top-line “organic” growth requires that frontline sales professionals be provided with leading-edge tools to identify better leads and hence close more deals. At a higher level, given that highly productive sales people are a constrained resource, sales executives need to optimally deploy the available sales force against the best revenue-generating accounts. Both objectives offer rich opportunities to apply leading-edge analytics to enhance decision-making capabilities at key points of the sales organization.
The Analytics Solution
In 2004, IBM initiated a broad, analytics-based initiative to improve sales productivity at both levels. The first solution, known as OnTARGET, provides a set of analytical models designed to target new sales opportunities at existing IBM accounts as well as noncustomer (“whitespace”) companies. The objective of the second stage of the initiative, the Market Alignment Program (MAP), is to optimally allocate sales resources based on field-validated analytical estimates of future revenue opportunity in each operational market segment.
The analytical models in OnTARGET and MAP are built using a large volume of historical IBM customer transactions, joined with “firmographic” information (e.g., industry, annual revenue) for each customer account. IBM developed a data model that reflects the correct alignment between the IBM-internal representation of an account and the view provided by an external provider (e.g., Dun & Bradstreet) of firmographic data. From a business perspective, the greatest modeling challenge has been to map the respective underlying business objectives to solvable machine-learning problems. For OnTARGET, this led to a classification method to estimate probability of new purchase, whereas for MAP, it led to a novel regression-based method to estimate the unobservable “wallet” of any potential customer in a specific product group.
The Value
As of December 2008, OnTARGET has been made available to over 13,000 IBM sales professionals in 25 different countries. Based on user logging capabilities built into the OnTARGET website, IBM has been able to generate evidence that sales leads identified through the use of OnTARGET close at a higher rate than average. A conservative estimate of the IBM revenue impact of this higher close rate is $100 million in 2008. The MAP wallet models are displayed in a separate Web-based tool that is used to conduct extensive workshops with sales leadership teams to elicit objective views of forward-looking revenue opportunity at each IBM account. Based on these interviews, a subset of sales resources have been redeployed to accounts with higher expected future revenue. The MAP process has been integrated since 2006 within the annual IBM planning cycle. IBM estimates the 2008 revenue impact of the MAP initiative to be approximately $500 million.
