Find one strategic AI use case to ease pilot chaos, add business value
In this podcast, Jim Dwyer of Sutherland explains how to identify an AI application that has proven business value and use it as a model for future AI deployments.
Many organizations are struggling to develop a coherent AI plan that corrals small pilots and freelance employee experiments into a unified strategy that adds business value while minimizing AI's legal, operational and financial risks. Consulting firms and system integrators have stepped into the breach by aggressively marketing new platforms and services to address the challenge.
Jim Dwyer, chief transformation and innovation officer at Sutherland, a business process outsourcing and IT consulting firm, offers advice that stands out amid the chorus of pleas to first develop some grand strategy to govern AI pilot chaos. He recommends identifying a single, high-impact AI use case and executing it well before moving forward.
In this episode of Enterprise Apps Unpacked, Dwyer explains why this approach works, how to identify the use case, as well as how to execute it well and then apply the lessons learned to future AI deployments.
A case for measurable success
Jim Dwyer
Dwyer's "pragmatic transformation" philosophy springs from more than 20 years of experience in digital transformation initiatives. Before joining Sutherland in 2020, he held senior IT management, business development and consulting roles in the healthcare industry. Now he leads Sutherland's AI and digital transformation practice as well as its healthcare division.
Dwyer said the need for more focused AI strategies is driven by boardroom and C-suite fatigue from AI pilots that prove the technology works but do little to deliver business value. But one well-chosen AI use case can provide that value.
"By getting to that high-impact use case, you're bringing all of the focus into one thing," he said. "You're focusing the leadership, funding, data access and change management, and you're trying to say, 'Let's do this one, prove its value and then extrapolate to the others.'"
Dwyer said Sutherland's experience with clients shows the best use cases tend to be where AI can relieve operational friction, workflows are broken -- and people are therefore willing to change them -- and AI can be harnessed to redesign the process.
"When we're thinking about identifying that use case, it is not so much what we can do, but what can we do and measure that value downstream?" Dwyer said. "The best AI use cases are usually hiding in plain sight. It's that high-friction, multi-compartmentalized, multi-department problem, and whether I can redesign the process, measure the value and establish traceability to a metric."
Choosing from potential AI use cases often depends on the nature of the company and its industry, according to Dwyer. Some companies clearly have a need to prioritize inward-facing applications that improve efficiency -- finance and accounting are common problem areas -- to stay competitive in their markets. Others are focused on outward-facing issues involving their customers, perhaps because they need to defend their position in the marketplace, he said.
Other topics discussed in the podcast include the following:
- Ensuring that the AI use case is executed successfully.
- How Sutherland applies the strategy to its own business processes.
- What companies get wrong when trying to use AI to drive digital transformation.
- Keys to success.
David Essex is an industry editor who creates in-depth content on enterprise applications, emerging technology and market trends for several Informa TechTarget websites.
