Demystifying predictive analytics: visualize your data

Saken Kulkarni

Saken Kulkarni

There is little doubt that designers of tablets or smartphones and personal applications, like Twitter,, or Instagram, have put a great deal of emphasis on creating an elegant and personal user experience. These companies have devoted entire departments to understanding human behavior, optimizing design, and learning how consumers interact with applications to maximize satisfaction.

However, it is also evident that an elegant user experience doesn’t always transfer to business applications. While we utilize mobile banking to pay bills and order an entire catered lunch with one swipe on the train ride to work, we are often stuck using clunky and linear enterprise-wide tools to drive business value. This is most often the case with business reporting tools. Selecting filters, drilling down on hierarchies, and obtaining a static report is not how we should be viewing and interacting with data.

One of the most effective ways to derive insights from data and glean more information from static reports is to create interactive dashboards that utilize data visualization. The best way to describe the difference between reporting and visualization comes from Anscombe’s Quartet, via my colleague Peter Gilks.

Anscombes Quartet

Anscombe’s Quartet tells a clear story. In Figure 1, you have a table with x and y coordinates for four different customers (I, II, III, and IV). At a quick glance, it is difficult to understand the trends or patterns in the different customers. Figure 2 takes it a step further, by showing variance, correlation coefficients, and linear regression lines. But as you can see, the statistical summary for each segment is the same and therefore adds very little value. How do we better understand the patterns in data?

Finally, in Figure 3, you are able to obtain a clearer picture. Despite identical statistical summaries, you can see that each customer purchase follows a unique trend.

Now that you understand the importance of visualizing your data, how do you select a visualization platform and create a culture of analytics that catalyzes enablement?

Goodbye canned reports, hello true self-service

One of the most important business questions that a CMO or VP of customer engagement may ask of his or her team is “who are my top 15 customers by geography?” Unfortunately, this question is not terribly easy to answer. Business users must navigate through canned reports, which may or may not have been written to answer this question. Moreover, even if the report is written, there may not be a unified definition of a customer, which creates a data integrity issue. Imagine a world where data is modeled in a unified customer warehouse, and users can simply plug in and design their own visualizations based on their requirements. This is true self-service.

Create a platform for collaboration

Designing a system for self-service for customer analytics is a tremendous step forward, but what do you do when hundreds of business users are designing reports? How do you create design standards? How do you ensure that the right users are accessing the right data? The most effective way to do this is to create a platform where users can collaborate. Users can post their visualizations to projects and sites moderated by an administrator; and users with proper access levels can view, interact, or even modify visualizations. This truly creates a platform for social business.

Laser focus on user experience

One of the major differences between traditional reporting and visualization is in the user experience. Take a look at the two reports below.

traditional reportingmultidimensional customer profile

Both views contain the same multidimensional view of your customer, however, they do so in different ways. Report #1 is a standard view that you would obtain from your linear reporting platform. Columns and rows and limited interactivity plague Report #1. Report #2, on the other hand, allows you to quickly identify your customers based on critical metrics, uses visual representations of data, and allows the user to drill down into specific segments. It is clear that that designers and developers of Report #2 have focused on creating a seamless user experience. Which tool would you rather use?

Patiently drive adoption

Creating hundreds of data-rich, customer-focused visualizations is a tremendous achievement, but there is no way to create a true culture of analytics without ensuring that new dashboards are rolled out effectively. Be patient with this roll out—business users who have been using Excel or static reports for years may initially be apprehensive about a completely different way of looking at data. Leverage social media, YouTube videos, and collaboration forums (like SharePoint) to encourage adoption and address any concerns.

Visualize your customer analytics data

Creating a platform for visualizing customer data will differentiate those who lead from those who fall behind. Answering business questions on customer churn, segmentation, and share of wallet in mere minutes through a self-service platform dramatically decreases time to insight. It will truly help your organization take a step forward in the path to customer-centric analytics.

About sakenkul
Saken is a Business Analytics consultant at Slalom Consulting. He focuses on the intersection of customer engagement, analytics, and data visualization

One Response to Demystifying predictive analytics: visualize your data

  1. Pingback: Demystifying predictive analytics: train and drive adoption | The Slalom Blog

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: