Big buzzwords can cause big confusion. Take artificial intelligence: In pharma, the technology has been touted as everything from the next big drug hunter to a customized communication tool.
Is AI all of those things to pharma? Possibly, say ZS consultants, but more research, data and education are needed.
ZS is aiming to provide some of that education with a new series called “AI &.” The first article, called “AI & Pharma: Start Small But Think Big,” debuted recently, and each article in the series will focus on a different AI-related challenge and opportunity for pharma. Upcoming topics include product launches, patient insights, forecasting and clinical trial operations.
ZS principal and analytics practice lead Arun Shastri plans to interview other ZS specialists through the course of the series to tap insights and examples for pharma readers. Pharma and data science lead Pratap Khedkar gave his take first, focusing on why pharma companies shouldn't just spend lots of money and hire a bunch of people to figure out AI.
“Start with clear, specific and even small-use cases. What are the three, four (or) 10 use cases I can think of to start with? Maybe two or three good ones in sales, two or three good ones in marketing for a given brand (and) maybe there are a couple of use cases for outpatient services,” he said.
Shastri backed up that thinking. Often money and time are spent looking at the big picture, he said, when pharma should instead focus on using AI in smaller ways.
“The AI mandate tends to be broad, sometimes pulled up into the C-level. It’s not that this is a mistake, but sometimes you don’t make progress with this approach. Because while you might create a muscle for AI, you might be floundering because you haven’t picked good problems to solve. In my experience and in general ZS experience, finding those problems and investing a little is almost always a better approach than building the capability and the muscle first for AI and then going looking for the problem,” he said.
Pharma is also exploring AI in areas such as extracting cognitive insights through patient-collected data.
As for the promise of AI for pharma, Shastri isn’t saying it's not there. But for pharma, rethinking strategies and getting the small stuff right should be the first order of business.
“Remember there was a lot of hype around IBM Watson and a lot of investment in AI for drug discovery, and by all accounts that’s still going on, but I think one of things they’re realizing is that they may not have sufficient training data for AI to be very successful,” he said.