IBM will never again be the company that defined the direction of computing for much of the corporate world. That is my judgment about its place in the industry, rather than a prediction that its business will disappear. IBM can remain useful, profitable and deeply embedded in its customers’ operations. Recovering the authority of the old IBM would require something much harder: rebuilding a market in which one supplier could set the terms for so much of enterprise technology.
The photograph captures the company’s enduring presence rather well. A tall black IBM Security panel dominates the center of an exhibition stand, carrying a blue shield made of points of light above the red words “Open. Simple. Connect.” Broad white frames stretch out on either side under spotlights, while blue strips cross the dark industrial ceiling. Below, a bright red counter and white pedestal stools give the stand a clean, carefully organized appearance. The crowd makes it human: a woman stands on her phone in the middle, visitors talk in small groups, and dark jackets partly obscure the view from both sides. It is a substantial corporate presence, busy and professional. What the scene cannot establish is whether the company behind the sign still determines where the industry goes next.

The old IBM’s strength extended beyond selling individual machines. Its hardware, software and service relationships helped organize how large institutions bought and operated computing. System/360 embodied that power: a compatible family of machines made a customer’s investment in programs and skills useful across different levels of computing capacity. The commercial advantage came from the surrounding commitment as much as from the equipment itself. Buying into the system shaped subsequent decisions, sometimes for a very long time.
That history matters because today’s discussion too often treats reputation as though it were an asset that automatically renews. It does not. A familiar name can get a vendor into the meeting. It cannot guarantee ownership of the next important platform. IBM’s own personal computer helped expand a market whose economics increasingly favored suppliers of operating systems and processors. Creating a market and controlling its profits turned out to be different achievements. That lesson still has teeth.
The industry now gives customers more ways to assemble their technology from different suppliers. Every additional layer creates a separate contest over pricing and influence. A company can have a strong relationship with the chief information officer and still lose the next spending decision to a specialist elsewhere in the system. An incumbent’s importance to yesterday’s installation offers no automatic claim on tomorrow’s budget.
Artificial intelligence sharpens that problem. Helping a customer connect AI to existing business systems can be valuable work. Yet the supplier doing the integration does not automatically capture the economics of the chips, the hosting infrastructure or the application that the customer ultimately uses. My concern about IBM is that the language of transformation can blur this distinction. Being involved in an important technology transition is a much lower bar than controlling a scarce part of it.
The investment test should therefore be fairly blunt. Does a new offering win additional customers on its own merits, or mainly attach to relationships IBM already has? Does growth survive after acquisitions and currency effects are separated out? Can the company retain more of the revenue as cash without weakening the product or service that customers are buying? Those questions would help distinguish a stronger competitive position from a better presentation of the existing business. Announcements alone cannot answer them.
There is a serious argument in IBM’s favor, and dismissing it would make the criticism lazy. Large institutions have complicated systems, sensitive data and expensive consequences when things fail. A supplier that understands those constraints can earn durable revenue. Replacing an established platform can cost more than continuing to run it. Those economics can support a good business for a long time, even when the business rarely excites people outside its customer base.
But durability has limits as a growth thesis. Revenue protected by the difficulty of leaving is different from revenue won because new customers actively prefer the product. Both can produce cash. They deserve different expectations. An investor who confuses them can pay for an expanding platform and receive a carefully managed installed base instead. Whether IBM shares are attractive at any particular price requires a separate assessment of valuation and cash generation; this argument alone cannot settle that question.
What would change my mind is evidence that IBM had become a supplier customers increasingly chose for new projects even when they had little existing dependence on it. Sustained growth in that kind of demand would carry more weight than another promise to reinvent the company. Even then, success would look different from the old IBM’s dominance. The market has changed too much for a simple restoration.
The exhibition stand still fills the frame. The IBM name still draws people to the counter. My view is that the company can build a worthwhile future, but the industry it once commanded is gone. Its next chapter has to earn its own value.