Please rotate your device.

Our website uses cookies to ensure you get the best experience while you’re here.


SEI Cincinnati’s Eric Spaulding on the Importance of Explainability in Data & Analytics

By: SEI Team

Eric Spaulding, a consultant at SEI Cincinnati, is an expert in Agile delivery when it comes to data and analytics. Before joining SEI, Eric was involved in a healthcare startup that released new, data-driven solutions in two-week sprints. At SEI, he applies his expertise in Agile methodologies to improve the performance of corporate D&A teams.

Agile D&A solutions enable the data team’s members to focus on solving problems quickly by interacting directly with business sponsors. They can stop waiting on formal hierarchies and requirements to direct them, focusing instead on the analytic priorities that require their attention. “This level of autonomy is often sought for by knowledge workers,” Eric explains, “but it’s hard to come by in sprawling corporations.” SEI-ers like Eric are dedicated to empowering enterprises — and their employees — to exercise autonomy and push time-critical decisions out to front-line data teams.

To achieve this, SEI consultants work with teams to spot bottlenecks and eliminate unproductive hand-offs. This ensures there’s a functioning process in place for technical data implementations. Eric and his colleagues turn complex data pipelines into sustainable solutions that can foster greater productivity in a wide variety of analytical use cases. This enables team members to accelerate their work and improve their performance without compromising the usability of the final analytics solution.

The Need for Explainability in Data & Analytics

One of the tenets of Agile is “working software over comprehensive documentation.” Unfortunately, this can sometimes encourage short-sighted data teams to create black-box solutions to demonstrate value quickly. This creates downstream challenges for business users — the customers of the solution and the users who need the data and the analytical models to solve pressing problems. In Eric’s view, this can be solved in part by capturing “just enough” to make an analytical or machine-learned model understandable to a wide audience. Documentation should also include assumptions or caveats the team is aware of at the time of its release. This way, it can be revisited as the team learns more through day-to-day use.

Eric anticipates that this process of building broad-based understandability for complex analytics applications — often termed “explainability” — will become a point of focus for many data leaders over the coming decade. Explainability enables employees to properly utilize decision systems. It promotes trust across teams and organizational functions.

Organizations that develop strategies for explainability will empower talented individuals with up-to-date data and tool knowledge to leverage their skills in tackling problems at hand. But as Eric warns: “companies that can’t, or won’t, address this will face internal criticism and scrutiny from watchdog groups and the wider public.”

Why do so many enterprises struggle with explainability? Eric believes it has to do with a general problem of focus. “There are so many potential solutions, platforms, and partnerships to explore,” he explains. Every day, data teams face an endless queue of new requests to handle and legacy solutions to maintain. It can be incredibly difficult for leadership to convene, assess the merits of an idea, and stick to a disciplined strategy that accomplishes critical activities like this one.

According to Eric, it takes “a true data visionary to deliver explainable solutions and help chalk up wins, gain momentum, and build the organization’s trust.” In his view, SEI consultants are precisely this — data experts that know how to cultivate broad-based understanding and exchange while building and implementing D&A solutions, whether through increased data literacy, governance activities, or right-sized documentation. He and his colleagues are dedicated to breaking down the silos that separate technical data teams from the customers they serve.

Fostering Confidence in D&A Through Explainability

Eric tells us that unlike SEI, “most consulting firms have a very inflexible, waterfall-style approach to building analytics solutions.” This starts with gathering long lists of requirements and “marching through implementations” without regularly reevaluating organizational priorities, the productivity of the processes, or the people performing them. “I’ve seen companies chase a ‘big’ data solution that wastes resources and is difficult to understand what you get in the end,” Eric says. “At SEI, our goal is to distill solutions down to what will drive the most business value.”

By designing D&A implementations from the start to provide both actionable and explainable insights, SEI-ers create a solid foundation for Agile data teams and promote widespread trust in the data. All these elements encourage better stewardship and higher adoption, enabling virtuous cycles of business transformation that move from organizational silos to much more responsive, Agile-led innovation.

Recently during his ten-year tenure at SEI, Eric worked with two large insurance companies that wanted to embrace an Agile, “team of teams” model — in one the focus was on enabling Tableau adoption, while in the other it was on improving data stewardship. “With both of these engagements, the goal was to define and implement a strategy that trained front-line employees to take the torch and run with it,” Eric explains. “The point is to stop waiting on formal permission from the corporate hierarchy to connect the dots and enable change.” Given the right training and mindset, employees can rapidly and confidently spearhead value-driving initiatives on their own.

Working Together for Better D&A Solutions

SEI consultants know the value of autonomy first-hand. SEI is an employee-owned company, and Eric and his colleagues are personally invested in creating a culture that values everyone’s input while empowering individuals. “We all get excited about solving business problems together,” Eric explains. “Our clients win because SEI is a team that puts tackling their challenges front and center.”

These collaborations, combined with in-depth knowledge and expertise, make SEI consultants experts at fostering understandability and trust within data and analytics initiatives. By helping everyone feel confident in their capacity to make decisions and in the data they leverage, SEI-ers can position a team to succeed by equipping them with an Agile, sustainable, and data-driven process.