Imagine you’ve just been put in charge of managing your company’s large sales force. Not just one salesperson, or even a hundred salespeople, but perhaps many thousands of salespeople.
Since each of these salespeople all have individual strengths and weaknesses, career histories, and characteristics, it becomes far too difficult to stay up to date with how each seller is doing. And yet, how well your sellers perform is crucial in determining whether or not your company will meet its goals for revenue and profit.
What if you could quickly answer the following questions:
*How does the strength of my sales force vary across different products and geographies?
*Which of my sellers are at risk of voluntarily resigning?
*Which of my sellers are at risk of missing their quota?
If you knew the answers to these types of questions, you could take proactive steps to keep your salesforce healthy, such as arranging for additional training, making strategic hiring decisions, or offering extra productivity incentives where they are needed most. Our team in the IBM Research Business Analytics and Mathematical Sciences group has created a prototype application called SellerScope that combines predictive analytics with interactive data visualization to give sales force executives exactly this type of insight.
We start by gathering useful data from across organizational silos. For example, human resources systems might record a seller’s job role, business unit, years of service, performance ratings, and promotion history. Finance systems might hold the seller’s base salary and information on commissions, while CRM systems track the seller’s assigned quota, progress and opportunities that are in the pipeline. After all this siloed data has been aggregated and annotated with historical outcomes, we use analytics to detect hidden patterns and allow us to make predictions on future outcomes.
While the output of this predictive analysis could be as simple as lists of names and predicted outcomes, without some context and explanations you wouldn’t have much reason to believe the results.
To help executives understand and trust the analytic results, the SellerScope application allows users to interact with a dynamic visualization that shows the interplay between the raw data and the factors that make up each prediction. We designed the visualization so that you can see the entire sales force at once, while still easily accessing the details for individual sellers.
In the main SellerScope visualization panel, each seller is represented with a small hexagon. The first image above shows the default visualization mode, which arranges all of the hexagons into a honeycomb-like grid. In the application status bar you can see that there are approximately 10,000 sellers currently being displayed. Note that because real sales force data is confidential, in this article, we have replaced it with example data that was synthetically generated.
SellerScope allows us to map the information we have about our sellers to different visual properties. In the first image, we are coloring the seller hexagons by the product line each seller represents. The visualization is interactive, and by clicking on individual sellers you can see either individual or aggregate information. Those details are displayed in the lower left-hand panel.
Other visualization modes in SellerScope include the bar chart mode (shown below) and table mode. In every visualization mode, users can choose to highlight particular subgroups of sellers, overlay our predictions of future outcomes, or filter down to a set of users that is particularly interesting.
As the user changes settings, SellerScope uses animated transitions to help the user maintain context. Examples of these transitions can be seen in this short video, which also demonstrates some other features of the system.
Why stop the insight with this article? Visit IBM’s visualization hub, IBM Many Eyes and join more than 100,000 like-minded visualization enthusiasts, academia and professionals, including additional insights from Jamie Rasmussen and other IBM visualization luminaries.
This blog, formerly known as the BA Blog, includes postings by subject matter experts across all disciplines of business analytics including visualizations, analytics, performance management, BI, social media analytics, risk analytics & much more!
A tag is a keyword you assign to make a blog or blog content easier to find. Click a tag to find content that has been assigned that keyword. Click another tag to refine the search further. Click Find a tag to search for a tag that is not displayed in the collection.