IBM reportedly shopping Watson Health just as healthcare gets hot

IBM could be looking to sell the Watson Health division for a mere $1 billion, according to an Axios report. The question is why is IBM running away from the healthcare vertical just as it seems to be heating up, and for such a low price?

Just last month, Oracle spent $28 billion to buy digital health records company Cerner. Last spring, Microsoft spent close to $20 billion to buy Nuance, which is used heavily in the medical industry, boasting 10,000 healthcare customers. That’s huge money, suggesting that enterprise companies are looking to embrace the healthcare vertical and willing to spend big bucks to do it.

IBM launched Watson Health in April 2015 to much fanfare. It was supposed to take Watson, IBM’s artificial intelligence platform, and put it to work on healthcare problems. The argument went something like this. Even the best doctor can’t read all of the literature out there, but a computer can do it quickly, and could come up with suggested courses of action to augment the doctor’s expertise and produce better outcomes.

It then did what IBM does when it focuses on something. It opened up a fancy headquarters in Cambridge in September that year. It also began announcing partnerships. It checked all the boxes, partnering with the likes of CVS, Apple and Johnson & Johnson.

Then it started buying companies. The first acquisitions were medical data companies, Phytel and Explorys. That was part of a pattern. Next came $1 billion for Merge Healthcare, a company that would provide medical imaging data. Later it would make its most expensive purchase, Truven Health Analytics for $2.6 billion. In total it spent $4 billion, according to reports, which seems kind of modest now compared to what Oracle and Microsoft just spent, but it was a lot of money in 2015 and 2016 when it was gearing up.

All of this was about taking a data-centric approach to feed Watson Health’s machine learning models. For whatever reason, it didn’t really work as planned, but it was a big part of former CEO Ginni Rometty’s plan to modernize the company by focusing on areas like cloud and AI.

Rometti was optimistic when she spoke to the Harvard Business Review in 2017:

Our moon shot is bringing world-class health care to every corner of the world. Some of that is already happening. Watson is being trained by the best cancer centers in the world and then being rolled out across hundreds of hospitals in China and India. Some of those areas have just one oncologist for maybe 1,600 patients. People in those regions have had no chance of getting world-class health care. Now they can, with Watson as an oncology adviser assisting doctors with decision making. And this is just the start.

But Rometti left in 2019, and her replacement, Arvind Krishna, has different priorities. He told Axios that broad healthcare vision might have been too optimistic. That could explain why IBM is looking to get out, Holger Mueller, an analyst at Constellation Research tells me.

“IBM is really focussing on its hybrid cloud strategy. In the process it is trying to get rid of all assets that divert attention and capital, as well as carrying the risk of reputational damage. Watson Health certainly qualifies for all three, so it comes as no surprise that IBM may divest the unit,” Mueller said.

While IBM will likely continue to pursue healthcare business in other ways across the company, even if it ends up dumping Watson Health, it would have to be considered a failed strategy after pouring so much money into it and getting so little back. Of course, it still qualifies as a rumor, even if it wouldn’t be a huge surprise to see it come to pass.