Demystifying predictive analytics: analyze the status quo

Saken Kulkarni

Saken Kulkarni

A few weeks ago, I wrote about how retailers can leverage analytics to get a more complete picture of their ultra-connected customers. I also presented a five-step road map pointing the way there.

Over the next few weeks, I’ll outline these steps in greater detail. In the process, I hope to demystify advanced analytics and help C-suite executives successfully implement a data-driven culture within their organizations.

Roadmap to advanced analyticsAnalyzing the status quo: start with your core competencies

Whether it’s starting a new exercise plan or setting up an advanced analytics program, getting started is usually the hardest part. Many organizations see the value in analytics, but don’t know where to begin. Starting with this simple question can make the process less intimidating:

Am I doing everything that I can to create a culture of analytics within my organization?

It’s a simple question to ask, but not necessarily to answer. Taking a look at your organization’s capabilities across the following four core competencies is a good place to start:

Competencies_advanced analytics

Executive sponsorship

Creating a culture of analytics requires a dedicated executive sponsor. This sponsor doesn’t need to be someone within your immediate management team or even your business line; it simply has to be someone who’s an advocate for using data to drive business decisions.

Take this example: a strategy and insights manager of a consumer packaged goods company may enlist its CMO to help drive the use of customer relationship management (CRM) data to gain a multidimensional view of its customer. This cross-functional partnership mirrors the diverse skills needed to succeed in a data-driven organization.

Statistical tools

Advanced analytics is advanced for a reason—it requires powerful statistical tools that can handle a high volume of structured and unstructured data and quickly generate insights. Investing in statistical analysis packages and data mining tools that can handle this pressure will pay dividends in the future. There are several options for these statistical tools and packages, but a robust one should contain the following:

  • Intuitive interface for business users to access and build analyses
  • Availability to access and manage data (MS Access, MS Excel, databases, flat-files, etc.)
  • Variable binding
  • Sampling
  • Rules building
  • Clustering
  • Descriptive statistics
  • Regressions
  • Decision trees
  • Vector machines
  • Time series and survival modeling

These packages can sometimes be expensive, given their power and complexity. The lowest cost option may be appealing, but pay close attention to features and total cost of ownership before selecting the discount option.

Human capital

Investing in sleek and powerful statistical tools won’t be productive unless the right people manage them. Here are five roles you need on your analytics team for success:

  • Predictive analytics ninja: This person is an expert in designing and executing predictive analytics. Decision trees, customer affinity, sentiment analysis, and data mining should be part of this person’s everyday language. A true data scientist with an MS or PhD in mathematics or statistics would be ideal, but not required.
  • Database/ETL expert: This role ensures that all the plumbing and wiring for advanced analytics is set in place before the house is built. The database/ETL expert has advanced SQL skills, is able to perform data integration tasks, and can design data models and develop data marts. Ideally, this person will also have the ability to design and administer data servers.
  • Data visualizer: After building your databases and executing predictive analytics, you’ll need a data visualizer to bring your data to life. Organizations traditionally think of static, front-end reports when it comes to data discovery, but it’s much more effective to generate interactive and visually rich dashboards. Someone in this role should have experience building dashboards in Tableau, Qlikview, or similar visual analytics tools; understand UX and design principles and best practices; and have a strong understanding of database principles.
  • Business strategist: Visual dashboards will help you quickly spot trends and patterns, but these dashboards will collect dust unless you have a business strategist to extract customer insights. This person can be a member of your in-house corporate strategy team, a business consultant, or simply someone with deep industry or functional expertise.
  • Organizational development expert: After building the relevant data marts, executing predictive analytics, visualizing data, and obtaining customer insights, you will also need someone who can drive user training and adoption. This person has experience with organizational change management, creating educational and training materials, and has a special skill in communicating technical items to a non-technical audience.


Management and marketing wonks may recall the Pareto principle, which states that 80% of effects come from 20% of causes. This is especially true when it comes to loyalty and evangelism. 80% of the promotion of your new data-driven approach will come from 20% of your employees. Thus, it’s critical to develop employee evangelists who will be advocates for your cause. This can be the marketing manager who can be seen playing around in Tableau or Qlikview during his or her lunch hour, the divisional VP who uses advanced statistical models in his or her status meetings, or the new, tech-savvy analyst with a background in quantitative marketing. These evangelists will spread the advanced analytics gospel to your employees and ensure adoption.

The first step to a strong foundation

Analyzing the status quo is the first step in creating a culture of analytics within your organization. Ensuring that your organization has a strong technical, analytical, and organizational foundation will get you started on the right track.

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

5 Responses to Demystifying predictive analytics: analyze the status quo

  1. Pat Hennel says:

    Data is only as good as the humans that review and do something with it! The people you make responsible for that data will determine what it means and how your company will utilize it.

  2. Saken Kulkarni says:

    Hi Pat,

    You are absolutely right! Stay tuned for the final post in this series, which will discuss adoption and training. The creation of a data driven culture will go nowhere if people aren’t properly trained.


  3. Pingback: Demystifying predictive analytics: evaluate your infrastructure | The Slalom Blog

  4. Pingback: Demystifying predictive analytics: design your analysis | The Slalom Blog

  5. 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: