IBM's stock fell 25% after chief executive Arvind Krishna acknowledged the company had "faltered," as enterprise customers redirected their technology budgets toward AI servers and storage. The commercial logic behind that shift is straightforward. The damage to IBM is not.
What changed for IBM's customers
The case for IBM has long rested on software and services that sit above the hardware layer. What's changed is where enterprise buyers are choosing to put their money first. When clients race to build AI infrastructure, servers and storage absorb the budget before it reaches the next layer up. Krishna's word "faltered" is precise: IBM did not lose customers to a rival in the conventional sense. It lost the timing. Buyers prioritized the compute and storage stack, and IBM's revenues reflected that sequencing.
The read-through for the broader enterprise technology market is pointed. Companies that sell software and services face a version of this same problem whenever a capital-spending wave hits the infrastructure layer hard. The wave does not have to last forever to do significant damage in a given reporting period.
The counterargument
The counterargument, and it deserves its own paragraph, is that AI infrastructure spending is a precursor to AI services spending, and IBM is positioned to capture the latter. Once the servers are racked and the storage is provisioned, enterprises will need the software and integration work that turns infrastructure into output. IBM has historically competed in that space. If the infrastructure build-out is as large as the share price reaction implies investors feared, the follow-on services opportunity could be significant.
That argument is real. The risk is that the timeline from infrastructure spending to services contracts is long, and competition for that follow-on work is substantial. The 25% single-session drop in IBM's share price shows what the market currently assigns to the probability of IBM capturing enough of it.
On balance
On balance, this is less a story about IBM losing to a rival than about IBM being late in the sequence of a spending cycle it did not control. Krishna's use of "faltered" suggests the company understands the diagnosis. The line to watch is whether AI services contracts follow the infrastructure build-out. For now, the market has answered that question with a 25% decline.