Predictive Analytics Times
Predictive Analytics Times
EXCLUSIVE HIGHLIGHTS
Data Reliability and Validity, Redux: Do Your CIO and Data Curators Really Understand the Concepts?
 Here are two recent entries on...
On Variable Importance in Logistic Regression
 The model looks good. It’s parsimonious,...
Data-Driven Decisions for Law Enforcement in Toronto
 For today’s leading deep learning methods...
AI, Machine Learning, and the Basics of Predictive Analytics for Process Management
 APQC Chair Carla O’Dell interviews Predictive...
SHARE THIS:

3 months ago
Analytics Assessment: A Blueprint for Effective Analytics Programs

 

If you are like me, when you first heard of analytics and its ability to benefit businesses, non-profits, and government agencies, you felt invigorated and excited, having to hold back the urge to shout “Charge!”  As a data scientist, I found this excitement to be well-founded: analytics powerfully leverages data to address important, existential questions facing any organization:

  • Are we effectively meeting our customers’ needs?
  • Are there more efficient ways to reach our strategic goals?

For many organizations, this excitement is sufficient to obtain the initial investments needed to get started:  gathering data, assessing IT-resources and tools, and finding talented analysts and/or data scientists.   None of these tasks are trivial, and for some organizations, building the analytics team proves so difficult it tempers that initial excitement.  Once these difficult steps are moved to the “done” column, there can be a tendency to sit back and wait for the analytics team to start delivering the insights the organization is so hungry for.

As one of those data scientists responsible for delivering insights, let me report that reality is more complex. I have seen situations in many organizations, where there are conditions that erode organizational trust in even the most talented analytics teams.  Here are some key examples:

  • Return on investment is not clearly defined: This speaks directly to how the organization defines “success”.  Without clear metrics, organizations may shut down even successful analytics programs simply because the value the team brings to the organization is not made clear.
  • Ineffective process: Analytics works best when an iterative approach is taken that allows for feedback to be integrated into the analysis process.  If the organization imposes a strict, pre-planned “waterfall” processes on the analytics team, suboptimal results usually follow.
  • Ignored insights: If the recommendation of the analytics team – despite its wealth of evidence – is inconvenient or counter-intuitive to decision makers, its results may be thrown out and ignored.  A perfect example of this is illustrated early in the book or movie Moneyball.

Even if your organization has developed an appropriate Analytics Governance Strategy, to realize long-lasting success with analytics, the organization must truly embrace data-driven decision making.

Success Starts with a Data-Driven Decision-Making Culture 

With the explosion of data in recent years, it is possible that business leaders today are actually leveraging less of data available to them, on a relative basis, than they were 10 years ago!  To know where we are with respect to the possibilities around extracting information from data, it is important to understand data-driven decision-making (DDDM).  Here are three salient features:

  • Measurement-oriented: DDDM relies upon data to generate quantitative metrics, often quantitative in nature, that objectively evaluate progress toward defined goals.
  • Results-oriented: DDDM is obsessed with achieving sustainable results, which requires metrics to assess those results and an honest accounting of the cost of achieving them.
  • Evidence-based: All decisions have a shelf life dictated by their continued effectiveness, which must be re-assessed regularly based on the latest evidence. If different options demonstrate better or more efficiently obtained results, then the organization should embrace this evidence and pivot to the improved return on investment.

To truly embrace data-driven decision-making, an organization must be willing to:

  1. Listen to the results of data analysis — even when inconvenient — honestly considering how existing decisions and processes could be improved.
  2. Invest in collecting and analyzing data to create metrics designed to provide actionable insights.
  3. Continually monitor results and re-assess decisions on a regular basis, even if it means abandoning a recent data-driven decision that may not have turned out as expected.

While governance is a good place to start, it only speaks to the general prerequisites necessary for an effective analytics implementation: it doesn’t capture all the nuances of your organization’s needs and current status.  The reality is that every organization is different.

What is the Best Path Forward?

Every organization is unique.  Some organizations have used analytics for years while others are just starting to realize its potential.  Begin by creating an analytics strategy and defining metrics to measure return on any analytics investment being considered.  Start small and move forward deliberately, as the time spent planning will more than pay for itself over the long run. Creating a successful analytics strategy requires assessing your unique organizational challenges, matching those challenges with relevant data and resources, and establishing processes that grow your capabilities.

No matter how far along the path your organization is toward developing an analytics program, it makes sense to assess your progress to ensure the right conditions are in place to foster the long-term success of analytics programs.  These assessments should provide:

  • Strategies for using analytics to improve organizational decision-making.
  • Recommendations on how to grow analytics capabilities and foster a culture of data-driven decision making.
  • Plans for short-term and long-term analytics opportunities, prioritized based on feasibility and return on investment.

When assessing your analytics program consider the following five areas:

  1. Analytics: How sophisticated is your organization’s analytical capability, including its ability to evaluate solutions and provide governance over data analytics?
  2. People: Are process stakeholders within your organization capable of, and involved in, framing questions in a manner that enables data-driven decision-making?
  3. Processes: Is a framework in place to guide project execution that allows for assessment of multiple potential solutions using specific selection criteria aligned to the organization’s needs?
  4. Infrastructure: Is the appropriate IT and technical infrastructure in place to meet your analytical processing needs?
  5. Culture: Is the “tone from the top” fostering a culture where people feel safe to accept data that may mean admitting a past decision is no longer the right decision?

Use the results of the assessment to develop an analytics program roadmap that ranks each prospective project along three dimensions—Cost, Return on Investment, and Actionability— to help identify low-cost, high-return actions that will foster a Data-Driven Decision culture and build analytic momentum.

Summary

All organizations, from business to non-profits, to government agencies, can better accomplish their mission by leveraging analytics with a data-driven decision process.  Using analytics to achieve a sustainable competitive advantage and generate significant return on investment begins with a well-conceived analytics strategy and roadmap for success that is aligned with, and supports, the overall business strategy.  An analytics assessments that includes technical, process, and cultural components can address the nuances of your organization and ensure that the conditions are right for the long-term success of your analytics initiatives.

About the Author:

Robert Pitney, Senior Data Scientist, Elder Research, Inc.  Robert Pitney enjoys listening to the needs of clients and finding ways that data can be used to solve problems, increase efficiency, or prevent fraud. Previously, Robert worked six years in the private sector as an IT Systems Integrator followed by eight years in the government sector as an Information Security Manager and an IT Auditor. In these roles, he learned the importance of analyzing data to ensure the continued security and integrity of key information technology systems, and became a Certified Information Systems Security Professional (CISSP) and Certified Information Systems Auditor (CISA).

Robert joined Elder Research after earning a Master’s of Science in Analytics from North Carolina State University, from which he also had earned undergraduate degrees in Economics and Computer Science. He actively seeks to cross-pollinate ideas from these fields to maximize positive impacts for clients.

In his spare time, Robert enjoys reading a good book, running, snow skiing, and cheering on his son at multiple sporting events. He has a passion for teaching and has served as an instructor for CISSP exam review courses.

Leave a Reply