Perspectives

Everyone loves a moon shot, but there’s a better (and quicker) way to win with AI

By August 18, 2020 No Comments

Moon shots are the stuff of dreams. Think Apollo 11, a 43-yard touchdown in the fourth quarter when you’re trailing by three points, ending world hunger (or COVID-19). Moon shots inspire and capture the imagination because they’re big, ambitious achievements of monumental—and seemingly unattainable—goals.

They’re also usually complicated and rarely achievable.

While some level of moon shot is at the heart of every innovation project, sometimes in pursuing a monumental goal, we miss the opportunity to solve problems that deliver immediate results and create real change.

In 2013, MD Anderson Cancer Center launched what Harvard Business Review called a “moon shot.” It tried recommending cancer treatment plans using Watson, IBM’s cognitive system powered by next-generation artificial intelligence (AI). By 2017, and after a $62 million tab, they put the project on hold, never using it on patients. 

At the same time, the Cancer Center’s internal IT team was using AI to make hotel and restaurant recommendations for patients’ families, identify patients who needed help paying bills, and address staff IT problems. Results included increased patient satisfaction, improved financial performance, and a decline in time spent on tedious data entry by care managers. Today, MD Anderson continues to use AI to solve practical business process problems that serve patients and improve efficiency, rather than focus solely on moon shots.    

AI needs a business use case.

Over time, AI has consistently improved operational efficiencies by automating manual tasks that clutter daily workflows. Embedding AI into the enterprise, though, can be a wasted effort if you plan poorly. MIT Sloan Management Review confirms that 70% of businesses have seen little or no impact from AI so far.

When you fail to get value from AI, you haven’t landed on what problem to solve. An O’Reilly survey revealed one of the primary bottlenecks to AI adoption is the difficulty “in identifying appropriate business use cases.”

In a recent webinar on the “State of Automation,” panelist Preeti Asthana, Director, Innovation & Partnerships, Data Analytic Services at Aon commented, “If you are automating a process, but the process is not right, then you’re going to get the same output. You have to fix the process before actually applying technology to it.”

Building a strategy is about knowing why.

Defining a use case is about determining the “why” and then building a strategy around it. McKinsey recommends business and technology staff work together: “Start and end the conversation on all technology platforms with the business problem that they will address. Focus on that relentlessly, and ensure that all the right stakeholders across business and technology have joint accountability for the platform’s delivery of customer value. When building your platform, focus on building use cases, and instead of spending time up front on putting enablers in place.”

Consider the people, process, and technology:

  • If you solve for this use case, who benefits? 
  • Who knows the most about the process, and what could work better?
  • Where could AI technology be applied, and where does it need a human-in-the-loop?

Line-of-business employees need to help you define the business case because they have the most precise picture of the inefficiencies. They know which problems to solve, what might be unsafe to automate fully, and where a human should be involved.

One Alkymi customer, TIIA, wanted to help its client onboarding team speed up pension rollovers—a very human-intensive, time-consuming, and sometimes error-prone process. With Alkymi Data Inbox, rollover documents were read and extracted in real-time. The information is assembled in an interface that team members can review and validate to complete the rollovers. The results were an immediate process acceleration, with no additional human intervention, better client service, and measurably less work for TIIA. 

When AI and machine-learning technologies address a specific task with narrow parameters, they easily solve problems. By automating account rollovers, processing client trading restrictions, and mitigating compliance risks, Alkymi solved several business problems for TIIA that made its operations more agile and efficient. 

This approach to AI—asking it to address practical, everyday use cases—is what delivers the right answers for businesses. It may not be a moon shot, but it’s solving and making a positive change right now. 

Schedule a demo with us, and we can help you identify the right questions to ask.