Knowing when the time is right to invest in AI

How one tech veteran determined which “shiny objects” were part of an effective digital transformation and which ones would end up a case study of what not to do.
Jeff Rowe

So you’re pretty sure you need to invest in new IT, but when’s the right time and what size commitment should you make?

Sound familiar?

As veteran tech exec John Glaser describes in a recent commentary at HBR, during his time as a CIO there were numerous time when he thought the latest “shiny objects” brought to his attention “might be very important. They might advance the organization’s strategies beautifully. They might enable strategies that we hadn’t even thought of before. They might be the key to our survival.”

Then again, he’d remind himself, they might not.  The key, he realized, was “to develop a way to determine whether a shiny object had significant potential or was a looming train wreck.”

First, he advises, focus on the transformation, not the digital. “Digital technologies are only valuable to the extent that they can be effectively applied to achieve organizational goals. So of course, first you need to have organizational goals that are sturdy enough to both sustain a transformative technology and resist the temptation to adopt an irrelevant one.”

AI, for example, can help a health system achieve the goal of providing world-class service for patients, says Glaser, as it can “analyze EHR and insurance claims data to look for treatment patterns that both improve outcomes and lower costs and then recommend changes in treatment protocols,” while also being “used in patient-facing applications to tailor patients’ experiences to their clinical conditions, language skills, and care preferences.”

Second on Glaser’s list of criteria is being clear on why a particular technology may be an important tool.

For example, he would surmise, “applying AI to electronic-health-records data may enable us to quickly and efficiently identify differences in the effectiveness of various treatments,” while “AI-based bots that recognize emotions and cultural expressions may enable us to provide a richer call-center experience to our patients who have questions about their health or recent bills.”

The third step for Glaser is on the practical side of considerations and is certainly applicable beyond the category of AI: Choose suppliers wisely.

Specifically, “(i)gnore buzzwords: disruptive, solution, platform, ecosystem, cloud. Make the vendor tell you what the shiny object does and what it has done for organizations like yours.

Step 4 is similarly practical, although over a longer period of time: Engage in iterative learning.

Specifically, introducing new tools may end up highlighting the need for further investment to maximize efficiency.  “For example, as you apply your newly acquired AI capability to identify treatment effectiveness, you may have to finally face up to and deal with the uneven, often poor quality of your EHR data. Maybe your physicians have been relying on the “notes” field instead of checking the boxes. Maybe key pieces of data end up in the wrong fields or are not documented at all. Understanding and fixing these issues is crucial in order to reap the full benefits of your new AI — or possibly to get it to function at all.”

Finally, and again over a period of time, Glaser points to the need to “sustain the digital transformation.”

The bottom line, he notes, is that transformation never stops, and neither technological innovation.  But whether it’s investing in AI or another category of tools, says Glaser, it’s critical that you have a clear sense of what your organization wants from technology and why.