Oct 21, 2026 earnings call — AI and hybrid cloud drive software-led growth
IBM raised 2026 free cash flow growth to more than $1 billion after Q3 revenue grew 6% and software grew 9%.
Revenue $17.1B (est $16.8B) · EPS $2.95 (est $2.85)
Key quotes
"Organizations are recognizing that AI is only as effective as the system around it, the data that provides context, the workflows that orchestrate action, the infrastructure that secures and optimizes operations, and the hybrid cloud foundation that provides flexibility without lock-in."
"z17 continued the strongest innovation cycle in the platform's history, generating more than $1 billion of incremental revenue over six quarters, compared with prior programs."
"As I mentioned last quarter, we accelerated our productivity initiatives to help enhance our operating profit margin and free cash flow commitments while continuing to invest for growth."
Analyst Q&A 8
why customers chose infrastructure over software
Q (Amit Daryanani · Evercore): The analyst asked why customers prioritized infrastructure over software this quarter. He asked if transaction processing growth is recovery of deferred deals or new demand from infrastructure, and if Red Hat bookings signal software acceleration into 2027.
A: Management said AI projects drive customers to modernize apps and run them on Red Hat OpenShift, and that OpenShift and data tools like Confluent are long-term tailwinds. They said they are at the start of this cycle and are comfortable with double-digit Red Hat growth in the fourth quarter, while Jim added the software rebound came from both slip deals and net new business.
"I think we are at the very beginning of this cycle, so I expect that to carry on for a long time." — Arvind Krishna (CEO·CHAIR)
how many slip deals closed and TP outlook
Q (Wamsi Mohan · BofA): The analyst asked how much of the software upside came from slip deals versus net new business for TP. He also asked how many of the slip deals closed and whether TP should be thought of as mid-single digit cycle-to-cycle.
A: Management said the TP strength was timing, not demand destruction, and that 3Q combined deferred 2Q deals plus a normal 3Q buying cycle. Jim said the slip deals were tens of deals, about a $400 million miss from 2Q, they closed two-thirds (about $250+ million), which he said was roughly 25% of the quarter's software growth, and they remain prudent on the full-year outlook.
"That is the 25% of the contribution of the growth that we've had of 4 points." — Jim Kavanaugh (CFO)
confidence in long-term software growth
Q (Ben Reitzes · Melius): The analyst asked if management still expects to approach double-digit software growth long term and how M&A will help reach that. He noted they'd need to accelerate after this year and asked where M&A fits.
A: Management said they have a lot of confidence in the strategy and portfolio to deliver durable growth. Jim pointed to indicators: a $25 billion ARR high-value recurring book growing high single digits, new products like Lightwell and GenAI, and that 2027 will focus on growth synergies from acquisitions like Confluent.
"We've got a lot of confidence in our strategy, and our portfolio is well-positioned to deliver sustainable, durable growth on our model." — Jim Kavanaugh (CFO)
Confluent upsell and margin improvement
Q (Raimo Lenschow · Barclays): The analyst asked about upsell and cross-sell from Confluent contributing to Data growth. He also asked what drove the gross margin and productivity improvement.
A: Management said Confluent is important for moving data in real time and helps keep data in place while supporting AI, and they expect lake house and Streamhouse capabilities to add cross-sell and upsell. Jim said gross productivity improved about 500 basis points this quarter, contributing to operating leverage and free cash flow, and they have about $5.5 billion of productivity realized so far.
"There is Confluent, which is important for moving data in real time, allows data to remain in place but still be available for AI." — Arvind Krishna (CEO·CHAIR)
do AI promises match reality
Q (Brent Thill · Jefferies): The analyst asked whether some AI pilots failed to match claims, causing deals to be pulled back, and if IBM saw deals recoup when reality didn't match early claims.
A: Management said large enterprises are now more focused on true AI ROI and on redesigning workflows to get value; AI is not a simple bolt-on. He said companies must pick a few use cases and scale them, choose trusted partners for integration, and control token spend; that shift is a tailwind for IBM products like OpenShift, Confluent, and mainframe.
"If you do not spend the time and energy to redesign your workflow, you are not going to get the full benefit." — Arvind Krishna (CEO·CHAIR)
breakdown of MIPS growth and pricing
Q (Fatima Boolani · Citi): The analyst asked for more detail on the TP recovery: how much growth came from shipped MIPS in slip deals versus other factors, and whether IBM gained pricing power.
A: Management explained TP monetization: more capacity (MIPS) leads to more monetization and pricing opportunity. Jim said they run about 145 million MIPS worldwide, noted z17 shipped MIPS are up 18% through six quarters and shipped MIPS are up over 100%, and called that a good future monetization opportunity for 2027.
"TP is the core mission-critical software that enables the most secure, reliable, and scalable platform around." — Jim Kavanaugh (CFO)
are customers asking about sovereign AI
Q (Brian Essex · JPMorgan): The analyst asked how focused customers are on sovereign AI and how much it figures into their transformation plans and data/IP protection.
A: Management said sovereign AI conversations are growing and had little traction in 2024, some attention in 2025, and picked up in 2026. He explained enterprises want to prevent logs or traces from leaving, so customers choose on-prem infrastructure or open-weight models run privately; he cited banks wanting sovereign deployments for sensitive workloads.
"Sovereign becomes how do you make sure that none of that happens?" — Arvind Krishna (CEO·CHAIR)
balancing AI cost and control
Q (Matt Swanson · RBC): The analyst asked how enterprises trade off AI cost versus control and how they architect networks to get both cost efficiency and data control.
A: Management said the performance gap between frontier and alternate models has shrunk since 2024, making alternate models more attractive. He said many tasks can be done with cheaper alternate models that cost about 10% to 25% of frontier token cost and are now roughly 90% as effective, and that with modest infrastructure (two or three racks) and easier software, enterprises can run open-weight models privately.
"As those of us who work in this world produce much more easy button software that can give you that capability for the alternate models." — Arvind Krishna (CEO·CHAIR)