The Marketing Strategist:

Why Marketing Automation and Lead Scoring Aren't Enough

September 20, 2011

by Chris Koch, ITSMA, and Kathy Macchi and Ned Cullen, Allegro Associates
The path to lead management automation seems simple enough. Choose a system and begin tracking lead behavior. Then assign values to the different behaviors—such as a trade show booth visit is worth one point and a white paper download is worth two points—and create a threshold score for when the prospect is ready to be turned over to sales. Yet lead management systems are not sufficient on their own. Lead management systems let you track how customers and prospects are interacting with you through Web tracking cookies and unique customer ID numbers (for tracking offline behavior such as interactions with the call center). However, these systems, while necessary to create an automated process and to provide the raw data for analytics, don’t have the kind of built-in analytical capability needed to predict behavior. They merely help marketers make more educated guesses. Defining a sales-ready lead is another example of a good educated guess. Current wisdom in lead management circles says that marketing and sales should work together to come up with a shared definition of a sales-ready lead (i.e., a prospect that is likely to be responsive to a call from sales). Like lead scoring, defining a sales-ready lead is better than doing nothing and it’s a good exercise in marketing and sales alignment, but it’s not fact based. Truly understanding customer and prospect behavior requires analyzing and interpreting the differences in behavior between leads that become customers and those that do not. Revealing the statistically significant differences between those two groups gives marketers the insights they need to improve performance. To generate statistically significant insights, it’s necessary to build an analytical model and gather enough accurate demographic data descriptions of the two groups. You’ll also need an internal analytics czar or group that is proficient in arcane software like SAS and SPSS. There are many roadblocks to developing predictive analytics in lead management, especially in B2B, where buying processes are complex and the pool of data is smaller than in B2C. But it is possible even with a relatively small set of data. By identifying the activities that truly separate customers from noncustomers, marketers can begin to predict whether prospects will become customers based on their activities. For example, using the data analysis, we can predict that, based on behavior, a prospect has a 90% chance of becoming a customer. That can make a big difference in helping salespeople determine which leads to contact first. However, the success and continued accuracy of predictive lead analytics depends on getting salespeople to offer regular feedback on whether the model is working. For example, getting sales leadership to require that salespeople call all contacts with a 70% or better probability ranking ensures that the validity of those rankings will constantly be tested. If the accuracy begins to slip, marketing knows that it’s time to invest in tweaking the analysis. Creating a predictive model for lead management shifts the basis of decision making from gut feel and educated guesses to fact-based insight about what customers and prospects do, not what they say. And that can have a dramatic impact on the bottom line. To discover the critical success factors for applying analytics to lead management, read How to Create a Predictive Lead Management Process.

Want the latest news from ITSMA?

Subscribe to our newsletters and alerts for ITSMA blog posts, research publications, events, and more.

Further reading