Realizing Value through Analytics
23.02.2017
Caroline Conway

Over the past few years, the sentiment that analytics and big data are not delivering business value has grown. In a 2015 Xerox Forrester survey of business stakeholders, 44% of respondents said analytics and big data are overrated or deliver insignificant benefits. Only 8% described these tools as business critical. More recently, the Harvard Business Review published “Why You’re Not Getting Value from Your Data Science“, calling out similar concerns from both business and technology stakeholders.

Why are so many companies failing to realize benefits even as the field has grown exponentially? A big part of the issue is the gap in world views between people with deep analytics and data science expertise and people with deep business expertise. The former tends to start with a particular set of techniques to which they try to fit various business problems. The latter lacks clarity on what scientific techniques are even available to them. This inevitably leads to only a partial or mismatched solution from which the business owner struggles to derive value.

This gap can be closed by disrupting and broadening the process of delivering analytics and big data. While working as consultants and insiders at some of the largest companies in the world, our team at Control Enter has closed this gap and delivered results multiple times in the high six-to-eight figures.  We have found that the following techniques are essential to break the pattern and deliver real results.

 

  1. Understand strategy, financials, and maturity to identify true opportunity
  2. Be rigorous in how you define the problem to be solved
  3. Blend techniques to deliver the simplest, fastest, and most relevant solution
  4. Be disciplined in testing and tuning for relevant results
  5. Partner and iterate with the customer for feedback and adoption

We start by understanding strategy, financials, and maturity to identify true opportunity

 Even when a business problem has been identified up front, we spend time to understand the context and purpose so we can deliver relevant solutions later in the process.  Our initial inquiry draws from strategic and competitive assessments, including evaluation of industry context and strategic priorities, the connection to and timing of financial objectives, and the company’s relative maturity in its data quality and availability, business processes, and analytics.  We ask these questions because both business problems and relevant solutions look very different based on the need in context.  Understanding context, business drivers, and maturity enables us to deliver relevant solutions for their biggest needs.

We are rigorous and comprehensive in how we define the problem to be solved

Many times, an analytics provider is brought in to solve a specifically defined problem for which a specific technology solution is already assumed.  By starting with context and moving into rigorous problem definition, we are able to take a step back and evaluate all performance drivers and gaps.  In many cases, related problems need to be solved or specific issues related to the core problem like process design need to be addressed in parallel.  This level of problem definition allows us to pinpoint everything that needs to be put in place to achieve a complete solution.

We blend techniques to deliver the simplest, fastest, and most relevant solution

Once the problem is defined, an analytical tool is often required to solve the problem.  However, when viewed only through the analytics lens, providers often miss that more than just a tool is needed to successfully solve the problem and deliver results.  In nearly every case, multiple techniques need to be applied in parallel to solve the total problem facing the business – whether it be reworking processes, systems, organizations, or strategies.  Our transdisciplinary approach, blending problem solving methods from design, analytics, technology, cognitive science, process engineering, and management science, enables us to take an objective view and incorporate interrelated methods to solve a business problem in full.

Be disciplined in testing, tuning, and iterating for relevant results

For a solution to be truly adopted by the business customer, the customer must have full confidence in the level of the solution’s accuracy.  We have frequently seen cases where a solution is delivered and assumed to achieve certain results based on incomplete up front testing and incomplete or incorrect field testing through pilots.  We view appropriate testing and use of test and control methods when piloting as critical to any analytics solution’s success.  These methods ensure the solution performs optimally in the real world, not just in a theoretical environment.

We use design thinking, iteration and partnership to gain feedback and adoption

From beginning to end, understanding and viewing the business customer as a partner is the most important element of success.  Too often, analytics solutions are conceived of and designed without really understanding the business customer, even when using iterative or agile development methods.  We use an iterative approach and leverage our up-front investment in learning the business context, but our most powerful tool to overcome this is design thinking.  This method is focused on deeply understanding our customers’ needs, accurately translating what those needs mean for solution design, and seeking qualitative as well as quantitative feedback as a core part of iteration.  When we invest in this level of relating to the business customer, we ensure adoption and results.

By taking this comprehensive approach to delivering solutions and especially breaking the habits of viewing analytics as a discipline that is separate from business and problem solving more broadly, we believe the value delivery challenges that many companies are facing today can be broken.  This takes additional effort, but the returns on the investment end up speaking for themselves.