Taking Analytics for a Test Drive

By Jane Griffin  |  Posted 2012-02-15

The goal of business analytics (BA) is to help organizations make better decisions. Whether management wants to know the next thousand customers the organization is likely to lose, or what location makes the most sense for setting up a new distribution center, business analytics can provide actual answers to complex questions, as well as a path to improved performance.

Still, many companies are reluctant to start a BA initiative. For some, the idea of unearthing the necessary data seems daunting. For others, the job of securing cross-functional buy-in and participation seems next to impossible. For most, the question, “Where do we start?” looms large.

A pilot program is often the most manageable first step on an analytics journey. Effective pilots share several characteristics, including minimal initial technology and infrastructure investment, simplicity in the approach and involvement by a business owner who “gets it.” Because a pilot program is a highly visible initiative, other parts of the organization will watch to see if the benefits were worth the effort. So it’s critical to have the proper focus from the start.

An analytics pilot program should:

• be important enough to matter, but focused enough to manage;

• have goals that are achievable without the need for new infrastructure or large technology investments;

• be simple in approach and refined in scope to prevent disruption to the business before, during or after the pilot;

• have support from a business owner with a vision for how improved hindsight, insight and foresight can improve decision-making and operations within that function; and

• address an area that is highly relevant and current, but which many enterprises may not have a “data” handle on yet (cloud computing, social media and mobile analytics, for example) to offer immediate business opportunities.

Choosing Your Route

Once these critical requirements are met, it is time to explore the functional areas and project types that make good candidates for a BA pilot, such as customer analytics, supply chain management, finance, human resources and risk assessment.

A customer analytics pilot can offer tangible benefits quickly. For example, a large, national investment bank gained rapid results from its customer analytics pilot program. Looking to enhance its customer segmentation capabilities, the bank undertook an effort to reclassify customers into segments that had the greatest potential for growth. The project team segmented customers by brand fit, revenue potential, potential growth rate, current accessibility and market fit.

As a result of the pilot, the bank was able to identify and target a new customer base and focus its marketing efforts on previously untapped segments. The new segments were single female investors older than 35, risk-averse immigrants and middle-income customers that were atypically risk-inclined.


Forming an Effective Team

A successful BA pilot leverages the skills of people in the appropriate areas and takes advantage of talents that may not be evident in an individual’s current role within the company. For example, a business may identify people who are passionate statisticians and analytical by nature, giving them an opportunity to participate in the pilot program. This approach builds analytics champions and makes efficient use of existing resources without hiring from outside. These individuals can then take the good news of analytics to their functional areas.

Here are other roles critical to a pilot program’s success:

• The executive champion is a senior-level executive who can establish the vision, demonstrate the value to other senior-level staff, analyze and measure results, and celebrate the benefits.

• The business liaison can translate what is happening at an analytical level into the language of business value.

• Analytics professionals are skilled architects, developers and visualization resources who work as troops “on the ground.”

• Business partners are external resources experienced in the business, process or technology aspects of analytics; they may be willing to put some skin in the game.

pagebreak title =Launching the Pilot} 

Launching the Pilot

A successful pilot achieves a balance between being narrow enough to manage and broad enough to demonstrate real business benefits quickly. One key to launching a pilot you can build on is to include the basic visualization components, so that future efforts can leverage the pilot’s underlying structure.

While the pilot has a defined start and end point, the goal is to deliver enough value to create a foundation for future initiatives. The life span of a BA pilot is usually three to four months and includes the following phases:

Gaining Consensus: Before embarking on the pilot, get buy-in from the functional area or business units involved. Communicate, listen and respond to questions and concerns.

Pilot Planning: Define a very specific hypotheses, objective and business problem for the pilot to address.

Resource Planning: Involve internal resources as much as possible, so that the pilot is viewed as a “family affair.” If you don’t possess the experience in-house, look externally for help.

Training: The pilot should focus on training the staff involved. However, for future phases, everyone in the business should be trained on the basics of analytics.

Infrastructure Assessment: Pilots should leverage your existing infrastructure and systems whenever possible. The goal is to build a more aware and nimble organization.

Kickoff: Make sure the parameters, scope and time frame of the pilot is communicated ahead of kickoff and that schedules are adhered to throughout.

Testing/Validation: Test, validate, document and publicize insights internally.

Measuring and Communicating Results: Measure results—savings, efficiency gains, risk mitigation, etc.—and communicate benefits companywide.


Measuring Success

An analytics pilot—focused on a specific functional area and staffed with a team of enthusiastic people who have well-defined roles—lets organizations test-drive analytics before investing for the long term. But to pave the road forward, it is important to establish metrics for success up front.

At the highest level, the pilot should align with the organization’s strategic objectives, addressing a well-defined and specific issue that is readily measurable and meaningful to executives. For example, it may demonstrate how a value-based pricing strategy can benefit customers and the business.

If the pilot shows sufficient results, it can gain momentum and can be expanded to become an enterprise solution and a valuable corporate asset.


Jane Griffin, a principal with Deloitte Consulting LLP, has more than 30 years of IT experience, with a focus on business analytics and business intelligence, data warehousing and master data management. She can be reached at janegriffin@deloitte.com.