What was changed
When Arvind Krishna became CEO in April 2020, IBM had ceded the PC market, watched its Watson AI disappoint, and carried barely half the revenue of its 2011 high-water mark — Fortune notes it was the only one of 17 US tech companies valued at $100 billion or more to have lost market value over the previous eight years. Krishna, IBM's first CEO with a research and engineering background after 35 years inside the company, did not want to preserve what he saw as a dusty legacy.
The bet came in November 2021 at IBM's Global Technology Outlook briefing, a full year before ChatGPT: AI research head Sriram Raghavan presented large language models pre-trained on massive datasets. Krishna, a PhD engineer, saw the significance immediately ('I'm not sure all of our business folks got it at that point,' he recalls) and at once gave his management team the green light to steer billions in R&D into AI foundation models and the infrastructure behind them, alongside updates to the hybrid cloud platform he had championed leading IBM's cloud division.
The strategy stood on three pillars — AI, cloud computing, and a multi-decade bet on quantum — and aimed where rivals did not: the Watsonx flagship is optimized for niche business functions like financial services, supply-chain optimization and IT operations, embedded into client infrastructure, rather than chasing consumer-facing general models of the kind OpenAI, Meta and Google pursued.
The company that funded the bet had been shrinking for a decade — about 100,000 fewer employees at the end of 2024 than in 2015, with non-core business lines abandoned or spun off. By June 2025 Wall Street had come around: the share price had doubled since Krishna took over, with a string of record highs since August 2024, generative-AI bookings had reached $6 billion since June 2023, and the streamlined software business was growing rapidly.
What it achieved
Share price doubled since Krishna took over, with record highs from August 2024; generative-AI bookings reached $6B since June 2023 on $63B of 2024 revenue.
Why it worked
Krishna's engineering background let IBM read the technical signal early — he grasped the LLM presentation's significance a year before ChatGPT made generative AI famous.
The bet was aimed where IBM already had enterprise distribution — financial services, supply chain, IT operations — rather than at a consumer race against OpenAI, Meta and Google.
A decade of pruning funded the pivot: roughly 100,000 fewer employees than 2015 and divested non-core lines freed resources for R&D.
Quantum got an explicit multi-decade horizon, signaling the patience needed for research bets instead of quarter-to-quarter payoff.
What can be applied
Research depth becomes strategy only when a CEO can read it and acts early: IBM saw the LLM wave a year before ChatGPT and pointed it at enterprise infrastructure, not consumer hype.
Aftermath
As of the June 2025 article the strategy had market validation — a doubled share price and $6B in generative-AI bookings — but Fortune notes IBM still had not fully embraced the Silicon Valley ethos of moving fast and breaking things, and its revenue remained far below the 2011 peak. The article's own title keeps the question open: whether Krishna can bring IBM back to its former glory is not yet settled.
FOLLOW THE EVIDENCE