By Tatem Burns and Tom Hill
Many organizations tweak incentive compensation or quotas for their sales team when they fall short on revenue growth goals. This failure may be attributed to out-dated processes, team configuration, or decision-making issues that are holding performance back.
However, data science can help deliver effective solutions for the sales organization. Instead of simply reacting to poor performance, data science can allow the sales organization to adjust to changes and avoid failure.
Most sales processes within organizations follow a familiar pattern. The classic stages from lead generation to qualification, defining the problem, and negotiation may continue to work effectively in some product-oriented and transactional-selling environments.
However, in today’s business environment, many buying decisions have become more broad-based, with more stakeholders contributing unique points of view that must be specifically addressed. The classic stages were originally designed to be implemented by individuals with minimal support from back-office teams. In today’s new sales reality, teamwork is far more important.
According to Challenger, the prospect makes no decision 38 percent of the time. Perhaps even more importantly, the most common reason for this lack of decision is the inability of stakeholders to come to an agreement. To be successful in promoting buyer consensus, your sales team needs multiple areas of expertise and many points of contact with a prospect. Should the sale go through, the implementation, account management, and/or customer success teams also get involved to support the new client.
Team configuration and effectiveness are critical when you are fielding large teams to close on complex sales. This means shepherding the customer through their experience with your product, service, or company.
The same factors that are making teamwork more critical are also making the old way of determining how sales teams are performing together obsolete. The old way relies almost solely on sales results with some light-touch qualitative review. Sales leaders used to join prospect calls and pitch meetings to provide coaching, which could then be used down the line when they conducted pipeline reviews and evaluated conversion rates. Today, this may not be enough.
The new way of evaluating sales team performance combines the best of the old with three high-powered techniques using data-related, strategic imperatives to analyze and optimize effective sales team structure.
Sales team role modeling analyzes specific behaviors to develop and reinforce what “great” looks like; then creates team role profiles based on gaps and strengths in your sales team capabilities.
Data Required
To carry out this strategic initiative, you will need certain data on your team members, specifically educational and employment biodata and personality characteristics. To add objectivity and additional data to the process, it can be useful to administer a comprehensive personality test to team members and incorporate the information gathered.
Approach to Analysis
The next step is to analyze the composition of your team and other teams to compare composition to performance. The most important questions to answer are:
Two basic data analytic approaches can be applied to the data you’ve gathered: regression and clustering.
Regression is a machine learning function that predicts a number, such as a yearly sales figure. The regression task begins with a data set in which the target value is known. In the model build (training) process, a regression algorithm estimates the value of the target (sales performance) as a function of the predictors (personality traits, tenure, etc.) for each case in the build data. These relationships between predictors and targets are summarized in models. These models can help to identify which characteristics and experiences of team members and of a team as a whole, best predict sales performance.
Clustering techniques are a type of machine learning that uses the characteristics and experiences of your sales team members to identify common categories. For example, to perform an analysis that compares members’ number of training experiences with their time interacting with a customer, clustering might identify two groups of employees – one group that has more training and more time with the customer, and the other group that has less training and spends less time with the customer. These profiles can then be linked to KPIs (key performance indicators).
Example: Spotting Patterns Points Way to Improvement
Using data science-based techniques, one technology company was able to discover some common characteristics of their low performing sales teams. They were more geographically spread out, had more diffuse sales targeting lists, and had sales competencies that were less aligned with their roles than high performing teams. Having knowledge of otherwise difficult-to-perceive patterns can lead to successful remediation of subpar performance.
Decision making analysis can pinpoint specific times during the customer buying (and customer success) timeline when certain actions should be taken in order to maximize sales KPIs.
Data Required
The two types of data that are required for this technique are:
Approach to Analysis
This analysis specifically finds the best times for sales team members to engage in certain customer-facing activities by examining the trajectory of sales team actions, decisions, and communications across the customer buying journey. This is then mapped against your sales process by linking timing to the sales team’s performance.
Regression techniques, as described in the last section, can also be used here. In this case, you can first create “codes” in your data to represent the points in the sales team timeline when certain actions, decisions, and communications are occurring. These codes can then be used as predictors of sales performance in your models.
Example: Discovering What the Best Have to Teach
A law firm was trying to understand the behaviors of its top rainmakers so that it could distill the insights into effective training for other business developers in the firm. A time and motion survey plus analysis of communication patterns were critical in accomplishing this purpose.
Data Science Technique #3: Sales Team Collaboration
Analyzing sales team collaboration can help you determine how effectively your sales team members are working together, engaging with the customer, and how these team interactions drive sales.
Data Required
You will need data that helps to translate team collaboration into actual revenue and profit, including:
Approach to Analysis
This data can be used to illuminate who sales teams are communicating with, how frequently, and which communication patterns are associated with high sales performance.
Organizational network analysis (ONA) is a method for studying communication and socially-driven networks within a company. This technique creates graphical and statistical models that can help you fully understand your sales team’s communication patterns.
Three common networking metrics produced by ONA are:
These metrics, once developed, can be used in the predictive modeling of sales team KPIs.
Example: Highlighting Behaviors That Lead to Success
A field-based consumer organization sought a deeper, data-driven understanding of how their sales managers contributed to growth and profit. Through predictive analytics, ONA, and modeling, they were able to identify the task approaches and crucial behaviors that contribute to the success of high-performing sales managers. Personality data, network signatures, and sales manager behaviors predicted 18 percent of the variance in year-over-year sales.
Don’t make the mistake of assuming that quotas and compensation are the sole determinants of sales success. Process is critical. Make sure your sales operations team has data science-driven tools to optimize process and performance plus the knowledge they need to make the most of them.