IBM execs say years of huge investments in Watson, which employs 1K people, are yielding profitable opportunities in markets like healthcare and manufacturing
Watson, can you grow into a multibillion-dollar business and become the engine of IBM's resurgence? — IBM is betting its future that the answer is yes.
Context & Ripple Effects
In 2016, IBM was presenting Watson as the engine of its resurgence, building on the 2015 deals with Apple, Johnson & Johnson and Medtronic that carried the system into healthcare. Executives told the New York Times that years of investment in a roughly 10,000-person effort were finally yielding profitable opportunities in healthcare and manufacturing.
The subsequent record complicates that framing: by 2018 IBM had bought Red Hat and pivoted back to core strengths, insiders later described commercialization missteps, and by 2021–2022 Watson Health — roughly $1B in revenue and unprofitable — was up for sale.
First-order effects
- IBM's own commercialization push concentrated on healthcare, where the Apple, J&J and Medtronic partnerships made Watson Health the flagship bet for turning research spend into revenue.
Second-order effects
- The shortfall forced a strategic redirect: IBM's Red Hat acquisition marked an explicit return to core infrastructure strengths rather than the Watson brand, and Watson Health went from growth engine to divestiture candidate, with IBM later restarting attempts to sell the division.
Third-order effects
- The arc from 2016 optimism to a 2022 fire-sale attempt became a template case for how enterprise AI hype cycles unwind — insiders' account of Watson's missteps pushed IBM toward narrower, less ambitious commercialization rather than moonshot vertical bets.
The trend: Enterprise AI is shifting from branded moonshot platforms pitched as company-saving bets to narrower, distribution-led deployments — with Watson the cautionary data point that set the pattern.