IBM Q3 2026: Data Software Leads Growth as AI Reshapes Its Consulting Mix

IBM reported third-quarter results on October 21 that put software, and especially data software, at the center of its AI story. Revenue rose 5% to $17.1 billion, software grew 8% while consulting was flat, and the company raised its free cash flow outlook. For CIOs, the numbers say less about any single AI product than about where enterprise AI money is flowing inside a large vendor: toward the data plumbing and operational tooling that agents depend on, and toward AI work inside consulting engagements.

What did IBM report for the third quarter?

According to IBM's earnings release, also filed with the SEC, revenue grew 6% at constant currency. Software brought in $7.8 billion, up 8%, or 9% at constant currency. Red Hat, reported as Hybrid Cloud, grew 11%. Infrastructure revenue rose 4% to $3.7 billion, and consulting held flat at $5.3 billion, up 2% at constant currency.

Free cash flow reached $3.0 billion in the quarter, $0.7 billion more than a year earlier, and $7.8 billion for the year to date. IBM kept its full-year forecast of 4% to 5% revenue growth at constant currency and now expects free cash flow to rise by more than $1 billion for the year, up from its previous outlook. Shares rose more than 2% in after-hours trading, according to a KuCoin News market summary.

Chief Executive Arvind Krishna used the release to argue that "clients are now deploying Super Intelligence in ways that transform entire operating models." That is a marketing frame, and IBM did not attach a revenue figure to it.

Why is data software growing fastest?

The most useful detail for enterprise buyers came on the earnings call. Chief Financial Officer James Kavanaugh said IBM's data business grew 20%, with bookings for Confluent, the data streaming company IBM now owns, up about 20%, Constellation Research reported. Krishna said Confluent helps unlock enterprise data for AI applications. Kavanaugh also said HashiCorp, IBM's infrastructure automation unit, grew bookings across all regions.

That pattern matches what practitioners describe. Agents that act on business events need current data, not last night's batch. At COLLIDE 2026, IBM's Ann Leach, Executive Director of Data and AI Product Marketing, put it plainly: "The speed of my data defines the speed of my business." Leach was describing IBM's own product strategy, so her view is not neutral, but the point holds well beyond IBM. We covered her session in our piece on streamhouse architecture and real-time data for agents.

Krishna made a related argument on the call, saying that "Deploying more models or building more agents will not be enough," according to Constellation Research. In his framing, the edge goes to companies that can run AI securely and efficiently across the business, not to those with the most pilots.

How much of IBM consulting is now AI work?

AI is changing what IBM consultants sell even though the segment is not growing much. Work tied to generative AI made up about 30% of consulting revenue in the quarter and about 35% of the consulting backlog, according to call summaries from INDmoney and KuCoin News. IBM's earnings release itself did not include an AI book-of-business figure, so those shares are management comments relayed by third parties.

For buyers, a flat consulting line with a rising AI share suggests that AI projects are replacing older work rather than adding to total spending. That is consistent with a market where budgets are being redirected, not expanded, and where integrators are under pressure to show measurable outcomes.

What about mainframes and quantum?

IBM Z revenue fell 20% from a year earlier, Constellation Research reported. Krishna said some deals that slipped from the second quarter closed in the third, which he attributed to timing rather than weaker demand, and he described the z17 as the strongest innovation cycle in the platform's history.

On quantum, Krishna said IBM signed 16 enterprise engagements this year and will share more at a Quantum Day event planned for December 10. Those are early commercial agreements, not production workloads.

What it means for enterprise AI leaders

IBM's quarter reinforces a theme AI Connect speakers keep returning to: AI programs succeed or stall on data. At The AI Enterprise Conference 2026, IBM's Tim Richer named the second big obstacle to scaling AI directly. "It's around stale data and data without context," he said, as we reported in our story on lessons from IBM's client zero program. A year earlier at COLLIDE 2025, IBM's Ryan Zombo compared governance to the brakes on a car, which exist so the driver can go fast, in our coverage of the five questions enterprises must answer to scale agentic AI.

The practical takeaway is to budget for the layers under the agents. Streaming and integration, infrastructure automation and governance are where IBM is seeing growth, and they are where many enterprise AI projects run short. The next thing to watch is whether IBM begins disclosing AI revenue in its own filings, rather than leaving investors and customers to piece it together from call commentary.

Frequently asked questions

How did IBM perform in Q3 2026?

Revenue rose 5% to $17.1 billion, software grew 8% to $7.8 billion, and IBM raised its full-year free cash flow outlook while keeping 4% to 5% revenue growth guidance.

How much of IBM's consulting business is AI?

According to call summaries, generative AI work was about 30% of consulting revenue and about 35% of the consulting backlog. IBM's release did not give an AI figure.

Why does Confluent matter to IBM's AI strategy?

Confluent provides real-time data streaming, which IBM says helps unlock enterprise data for AI applications and agents that need current information.

Filed under: Data & AI Infrastructure, AI Strategy & ROI