Achieving Breakthroughs One Step at a Time
Many organizations are looking for a sudden leap forward, but these case studies show that incremental change is more sustainable and practical.
- By David Stodder
- March 16, 2018
Organizations today need to consume and analyze data just to keep their operations running smoothly, but most of them are looking for a breakthrough. They want an insight that will push them past the status quo and into new dimensions of competitiveness, effectiveness, and profitability. It's critical for organizations to deploy the right technologies and have a culture and strategy that supports analytics for identifying breakthrough opportunities and acting on them.
In essence, this was the theme of the first TDWI Leadership Summit of the year, held at Caesar's Palace in Las Vegas last month. Amid the Augustan statues, burbling slot machines, lots of bling, and stands of Zach Galifianakis "Did Caesar Live Here?" shirts, professionals from business and IT gathered to talk about how organizations can bring together relevant data and develop and apply analytics insights that lead to breakthroughs.
I had the honor of co-chairing the Summit with Donald Farmer, principal of TreeHive Strategy and for decades an important industry leader in the development of self-service analytics, visualization, data mining, and other capabilities. Through case studies, expert sessions, panel discussions, and give-and-take with attendees, speakers shared their best practices and wisdom gained through experience.
Here are my top three takeaways from the Summit.
Takeaway #1: Pay attention to how bias and strategic context affect interpretation of data.
Humans are not objective, nor do they make decisions in a vacuum. In the opening session, Sheridan Hitchens, an independent consultant with long experience in analytics, examined how bias and behavior shape the interpretation of data and how humans make choices in reaction to how analytical results are presented to them.
Hitchens advised attendees to be aware of decision makers' natural overconfidence in making estimates and the common desire to look for how to avoid losses rather than make equivalent gains. Decision makers also often anchor estimates on numbers that may not be relevant to strategic objectives. Along with openly acknowledging and understanding these human tendencies -- and considering whether some decisions would be better made automatically by systems -- Hitchens suggested training focused specifically on the issue of bias in analytics.
Lyndsay Wise closed the Summit by offering practical advice about how organizations can clarify strategy. Formerly a consultant and industry analyst and now a solution director for Information Builders, Wise offered ideas about setting the right context for dashboards, reports, and analytics, including identifying what the cost/benefit of a decision would be and focusing on the top five key performance indicators, not hundreds of them. As for self-service BI and analytics, "know your audience," she recommended, including their ability to interpret visualizations. "Self-service means different things to different people."
Takeaway #2: Don't overlook the potential for incremental gains.
In an age of big data and data science, it's easy for organizations to get so focused on hitting the home run that they neglect opportunities for smaller improvements to user productivity and operational processes. Several speakers at the Summit advised attendees to observe where users are not getting the information they need at the right time and to use self-service BI and visual analytics to institute changes.
Melissa Pomeroy, VP and business information services manager at St. Mary's Bank, told the story of her organization's "analytics adventure" in pursuit of a single source of truth, which she considers the first step toward becoming a fully data-driven financial institution. With many users working with spreadsheets, Pomeroy's group first focused on helping these users gain speedier access to trusted data and showing them how better data presentation could improve their efficiency.
Her group employed visual analytics tools to improve data quality, assigned data stewards in each department to improve data quality and reporting standards, and set governance policies to hold business-side data owners more accountable for their data. She described how her organization is now able to move forward -- step by step -- toward more advanced types of analytics.
The Summit's Tuesday keynote featured Laura Rea Dickey, CEO of Dickey's Barbecue Restaurants, Inc. Dickey has a background in marketing and information technology, including serving as the company's CIO for eight years. Her understanding of the importance of good coordination between business and IT has been key to the company's advancement in using BI reporting and analytics to improve supply chain management, operations, and business performance. She described the challenges of providing the right information at the right time to franchisees, managers, and even barbecue pit masters.
She demonstrated how the company is moving to voice services; implementing technology from iOLAP, Dickey's connects BI and analytics with Amazon Alexa, allowing store personnel to ask queries and get immediate answers about supplies, customer patterns, and more. The improvement to information access is enabling stores to operate with great knowledge and efficiency, which is critical in a low-margin business, she said.
Takeaway #3: Be selective in choosing the right projects for advanced analytics
Staying focused on delivering gains -- especially to nontechnical users in business departments (or to franchisees, in the case of Dickey's) -- helps ensure a tight connection between advancements in visualization and analytics and the business purposes to which they could be applied. The same should hold as organizations move into more advanced analytics.
A good example came in a presentation by Dr. Todd Pawlicki, vice-chair for medical physics and director of the Division of Medical Physics and Technology at the University of California, San Diego Department of Radiation Medicine and Applied Sciences, and Andrew Cardno, chief technology officer at VizExplorer. Pawlicki described the complex, fast-moving processes in a radiation lab, noting that the faster they can move patients through procedures, the better for their mental and physical health as well as for the department's ability to take on more patients.
The UCSD lab has deployed VizExplorer's visualization and analytics technology to enable real-time spatial awareness based on data streams coming from sensors everywhere in the radiation oncology environment. UCSD is applying predictive modeling to understand how it can improve patient safety and better automate patient scheduling so that the lab can serve more patients with fewer delays, translating into less time for patients to sit in the waiting room. Pawlicki and Cardno showed how by more fully understanding workflow, the lab can understand why mistakes or accidents happened, predict how and when more could occur, and avoid future errors.
Joey Fitts, founder and CEO of outelligence, offered a useful framework for understanding the purpose of "intelligent" applications that use artificial intelligence techniques such as machine learning. Conveniently, they all begin with an "A": the applications should be aware, aligned, analytical, autonomous, and assimilative. We are already using applications that meet some of these conditions.
For example, Waze, the mobile GPS navigation system owned by Google, offers awareness; it can observe and orient to the environment to provide recommended routes before a person or system decides to act. Waze, Fitts said, offers a great example for organizations in identifying operational needs to create a focus and context for applying advanced analytics.
Breakthroughs with Feet on the Ground
The Summit's speakers stressed practical, real-world applications of technology and practices as a way to achieve breakthroughs. It was inspiring to see and hear how much they have accomplished -- from helping spreadsheet users move toward visual analytics to applying real-time analytics to dramatically improve processes and workflow. It was a lesson in how the most sustainable breakthroughs happen one step at a time, not through one big revolutionary change.
You can learn about upcoming TDWI Leadership Summits here.