IBM Accelerates Software Growth to 10%+ on AI Platform Momentum and Mainframe Strength
Q1 2026 Earnings Call, April 22, 2026
IBM delivered a robust first quarter that exceeded expectations across multiple dimensions, with management now projecting software revenue growth exceeding 10% for the full year, up from prior 10% guidance. The acceleration reflects strengthening demand for the company's AI-enabling infrastructure, successful M&A integration, and what CEO Arvind Krishna described as structural tailwinds from enterprises moving beyond AI experimentation to production-scale deployments.
The company reported 6% constant currency revenue growth in the quarter, 140 basis points of operating margin expansion, and $2.2 billion in free cash flow representing the highest first quarter performance in a decade. More significantly, the results validate IBM's strategic positioning as what Krishna termed "Switzerland" in the AI infrastructure layer, remaining neutral on frontier models while capturing value from the underlying platforms enterprises need regardless of which models they ultimately deploy.
Software Acceleration Driven by Data Portfolio and AI Tailwinds
The software segment delivered 8% growth in the quarter, with management raising full-year expectations to above 10% growth from 10% previously. CFO Jim Kavanaugh attributed the increase partly to Confluent's early close in mid-March, which occurred roughly two months ahead of the mid-May timeline originally contemplated. However, the raise reflects more than just acquisition timing.
The data portfolio emerged as the star performer, growing 16% in the quarter. Kavanaugh now expects data to grow in the "low 20-plus percent range" for the full year, contributing 5 points to total software growth. Of that growth, "a little bit north of 15 points" comes from M&A contribution, primarily Confluent and DataStax, leaving mid-to-high single digit organic growth in the underlying business.
Red Hat growth accelerated two points sequentially to 10%, driven by what Kavanaugh described as "stabilization of consumption-based services revenue growth that we expected." OpenShift has crossed $2 billion in annual recurring revenue with "high 20% growth," while virtualization has accumulated over $600 million in contracts signed since early 2024. The company did note RHEL deceleration, attributed to federal government closure impacts in Q4 that flowed through to Q1, and what Kavanaugh called "a very dislocated hardware supply chain market."
Automation grew 7%, with management highlighting that HashiCorp marked its one-year anniversary in February with record bookings and adjusted EBITDA accretion "ahead of expectations." Transaction Processing grew 2%, monetizing the z17 mainframe cycle that continues to outperform prior generations.
Mainframe Emerges as AI Inferencing Platform
Infrastructure delivered perhaps the quarter's most impressive performance, growing 12% overall with Hybrid Infrastructure up 25%. IBM Z grew 48% in what management characterized as another record quarter, with distributed infrastructure growing double digits driven by Power11 and new flash storage offerings incorporating "industry-leading agentic AI capabilities."
Krishna provided detailed commentary on how mainframe is evolving into an AI inferencing platform, representing what he described as a "third kind of compute capacity" beyond traditional MIPS and Linux workloads. He explained that financial services clients currently sample perhaps 10% of transactions for fraud detection because taking transactions off-platform introduces unacceptable latency. The Spyre Accelerator enables running 20-30 billion parameter models directly on the mainframe with "only milliseconds of latency," allowing fraud detection on 100% of transaction volume.
"If you can take your fraud rate down from 50 basis points to 40, you can now do the math on what that is," Krishna noted, adding that a fully populated system can handle "about 450 billion inferences a day." Financial services clients are "saving tens of millions of dollars" through real-time fraud detection capabilities.
Kavanaugh quantified the z17 cycle's performance, noting that in the first full year, IBM increased hardware placement value by "over $1 billion" compared to z16's first year, which was itself a record. With a 3-4x software and services multiplier on hardware placements, this translates to $3-4 billion in future monetization opportunity. Management emphasized that for four consecutive quarters, z17 has shipped over 100% year-over-year MIPS growth, with higher capacity translating directly to higher monetization.
Consulting Returns to Growth with GenAI Representing 30% of Backlog
Consulting grew 1% in the quarter with signings returning to growth at 6%, marking an inflection after several quarters of contraction. More importantly, generative AI now represents approximately 30% of consulting backlog, reflecting what Kavanaugh described as "how generative AI has become embedded in the work we do."
Management provided new disclosure that consulting GenAI annual recurring revenue "eclipsed $4 billion" in the first quarter, growing over 40% year-over-year. Approximately 80% of GenAI bookings come from net new clients, with the company adding 400 new clients in the quarter. Backlog quality metrics improved with erosion stable, duration declining, backlog realization accelerating, and yields up 4 points year-over-year.
The company highlighted its Consulting Advantage platform as driving both client value and internal productivity. Krishna noted that IBM's internal AI development system, Project Bob, is now generally available with "average productivity gains of 45%" across the developer workforce. The system "automates the full software life cycle from legacy modernization to security using specialized agents and multimodal optimization."
Strategic Positioning as Platform Layer Switzerland
Krishna articulated IBM's strategic positioning with unusual clarity during the call, particularly in response to questions about application layer exposure and M&A appetite. He noted that only "4% of our portfolio, if I'm to be generous, could be called an application," specifically citing Maximo as the primary example, though even that functions more as a system of record than traditional application.
The portfolio consists primarily of what Krishna termed "enabling software" rather than applications, with Red Hat providing operating systems, containers and automation; the data portfolio providing databases, data movement through Confluent, and AI enablement through watsonx; and automation software addressing IT infrastructure management through Turbonomic, Apptio and HashiCorp.
"As agents replace people, for some fraction, we can debate how much of the interactions, then the interaction layer by itself is not sticky," Krishna explained. "The agents are going to be interacting much more with the underlying data and the business logic. And we sort of saw that coming 6, 7 years ago, and that is why we picked the portfolio we did."
On frontier models, Krishna emphasized that IBM made "the decision about 3 years ago that we were going to be neutral and Switzerland-like" on frontier model usage. "We actually do not want to even predict which of them will be the eventual winners. We want to work with all of them." The company positions itself where clients want hybrid deployments across multiple clouds or private infrastructure for sovereignty, brand, privacy or economic reasons.
Productivity Flywheel Drives Margin Expansion Ahead of Schedule
Operating margin expansion of 140 basis points exceeded expectations, with segment profit margins expanding 720 basis points in Infrastructure and 60 basis points in Software. Kavanaugh highlighted that since 2023, the company has driven $4.5 billion in productivity savings with an additional $1 billion expected in 2026, calling it "a proven repeatable AI-enabled transformation engine that is accelerating."
The productivity gains enable simultaneous margin expansion and accelerated innovation investment. Management absorbed incremental dilution from Confluent's early close, which was originally expected to contribute approximately $600 million in dilution for the full year "driven largely by stock-based compensation and interest expense." Despite the early close adding to that dilution, actions to accelerate cost synergies keep the company on track to expand operating margins by approximately 1 point for the full year.
Software portfolio economics continue improving with annual recurring revenue approaching $25 billion, up 10% year-over-year. Approximately 80% of the $30 billion trailing twelve-month software revenue base comes from high-value recurring revenue, with 20% transactional. The AI platform, agents, assistants and orchestration components now exceed $1.5 billion on a trailing twelve-month basis, representing 25% penetration of the software business and growing over 40%, contributing 2 points of growth on an annualized basis.
Macro Environment Prudence Despite Strong Underlying Trends
Management maintained full-year guidance for 5%+ revenue growth and approximately $1 billion free cash flow growth despite the strong start, citing prudence given only one quarter of results. Krishna noted this represents standard discipline, pointing out "I've been in this role now 9 years. Arvind has been in the role 6, 7 years. I don't think we've ever raised guidance in the first quarter."
On macro conditions, Krishna provided geographic color noting the Middle East "had the strongest growth we have seen in decades, not years, decades" with no impact from recent developments. Europe also performed strongly in the quarter. He noted "if the Strait stays closed for another few weeks, then we know that there could be energy impacts in Europe, but that is speculative. That is not what we are seeing."
The company emphasized diversification as a key strength, with Kavanaugh noting "arguably, yes, we're operating in a dynamic world, and there's more uncertainty than there was 90 days ago" but within IBM's lens "we're executing extremely well across our high-value innovation software, infrastructure and consulting."
M&A Appetite Opens in Dislocated Market
Krishna provided candid commentary on M&A appetite following Confluent's integration progress. "Yes, the values that are out there right now are very attractive. That does not always mean that the sellers are willing to accept these values. That may take a few months for them to acknowledge that this is a new baseline."
He noted that "second half, if things stay where they are and if the values are where they are, maybe we can do something in the second half as we build up our cash balances, and we are 100% sure that Confluent is off to a strong start." The company exited the quarter with $11.8 billion in cash and $66.4 billion in debt including $12.8 billion for the financing business, with the receivables portfolio 80% investment grade.
Management's willingness to consider additional M&A represents a shift from the immediate post-Confluent close period, though discipline remains paramount. The company invested $10.5 billion in acquisitions during the quarter primarily for Confluent while returning $1.6 billion to shareholders through dividends.
Client Wins Demonstrate Platform Value
Management highlighted several client examples illustrating platform positioning. ServiceNow leverages watsonx for automated data quality and observability to deliver AI-ready data and code generation. Visa continues software and data modernization supporting VisaNet scale and performance. Nestle uses NVIDIA-accelerated watsonx.data to embed AI into order-to-cash operations for real-time supply chain insights. NatWest and RBC modernize mainframe environments using watsonx Assistant and watsonx Code Assistant for Z to improve resiliency and developer productivity.
Krishna emphasized that clients deploying watsonx Code Assistant for Z are "growing MIPS capacity 3x faster than those who have not," providing concrete evidence of AI driving consumption growth rather than displacement. This represents a critical proof point for the thesis that AI creates net new workload demand for IBM's infrastructure and enabling software layers.
The quarter demonstrates execution on the multi-year portfolio transformation while validating strategic positioning at the intersection of hybrid cloud, AI and mission-critical enterprise infrastructure. The software acceleration, mainframe strength, and consulting inflection provide multiple growth vectors with improving economics as the productivity flywheel continues accelerating.
International Business Machines Corporation Deep Dive
The Platform-First Architecture
The market's residual perception of International Business Machines Corporation as a legacy IT services provider is structurally misaligned with its current operational reality. Over the past five years, the company has executed a methodical replatforming into a software-led, hybrid cloud, and artificial intelligence enterprise. The core business model is now anchored by three symbiotic segments: Software, Consulting, and Infrastructure. Software is the economic engine of the business, currently representing approximately 45% of total revenues and serving as the primary vehicle for margin expansion. This segment captures high-margin, recurring revenues through Red Hat's open-source platforms, the watsonx artificial intelligence environment, and a rapidly expanding suite of automation and data management tools.
The Infrastructure segment remains a highly lucrative computing substrate, driven primarily by the cyclical but deeply entrenched Z series mainframes and distributed computing systems. Rather than viewing hardware as a declining legacy business, management has successfully repositioned it as a critical node for highly regulated compute workloads. The Consulting segment, which accounts for nearly a third of revenues, acts as the implementation layer. While its stand-alone margins are structurally lower, Consulting serves as a high-touch, enterprise sales channel that embeds the company's software and infrastructure solutions deep into the workflows of Fortune 500 clients, pulling through high-margin platform subscriptions.
The Competitive Landscape
The company operates within a complex matrix of competitors spanning public cloud, enterprise software, and global IT services. In the cloud infrastructure layer, it no longer competes directly against hyperscalers such as Amazon Web Services, Microsoft Azure, and Google Cloud Platform for raw public cloud dominance. Instead, it occupies a highly strategic middle ground by positioning Red Hat OpenShift as the agnostic hybrid cloud operating system that allows enterprises to bridge on-premises data centers with multiple public clouds. In the data and artificial intelligence domain, the company competes against nimble pure-plays like Databricks and Snowflake, as well as the proprietary artificial intelligence stacks of the hyperscalers.
In the Consulting segment, fierce competition comes from global system integrators, primarily Accenture, Deloitte, and Infosys. While competitors like Accenture boast larger standalone consulting volumes and faster top-line services growth, the differentiation lies in owning the underlying proprietary software infrastructure. This dual capability allows the company to architect and deploy end-to-end solutions, capturing a greater share of the artificial intelligence value chain rather than merely acting as an implementation vendor for third-party technologies.
Structural Moats and Sovereign Artificial Intelligence
The most formidable competitive advantage lies in the deep architectural stickiness of the transaction processing infrastructure. Mainframes handle an estimated 70% of the world's transactional workflows. The recent launch of the z17 platform has proven to be a blockbuster cycle, driving a 48% year-over-year revenue growth in the Z segment during the first quarter of 2026. Crucially, the z17 architecture incorporates the Telum II processor and the new Spyre artificial intelligence accelerator. This hardware evolution allows clients in financial services, telecommunications, and government to run massive artificial intelligence inference workloads directly on the mainframe, eliminating the latency and security risks of sending sensitive data to public clouds.
Beyond hardware stickiness, a unique economic moat has been cultivated in sovereign artificial intelligence. While competitors focus on massive, consumer-facing large language models, the watsonx platform and Granite models are tailored strictly for the enterprise. These models are designed to run securely within a client's private perimeter, ensuring regulatory compliance and data privacy. Furthermore, the watsonx.governance framework provides the auditability and hallucination-mitigation necessary for highly regulated industries. This enterprise-first focus is yielding tangible commercial traction, with the generative artificial intelligence book of business surpassing $12.5 billion by the end of 2025.
Industry Dynamics and Strategic Opportunities
The enterprise technology sector is currently transitioning from an exploratory phase of generative artificial intelligence into a phase of scaled, production-level deployment. This shift presents a massive secular tailwind for hybrid cloud orchestration tools. As data gravity increases, large organizations are realizing that migrating legacy, mission-critical data to public clouds is prohibitively expensive and fraught with compliance risks. The architectural philosophy of bringing the artificial intelligence models to the data, rather than moving the data to the models, is resonating strongly with chief information officers.
Another emerging opportunity lies in edge computing and mobile artificial intelligence ecosystems. Recent collaborations with telecom and semiconductor leaders, such as the optimization of Granite models for Qualcomm's Snapdragon architectures, signal an intent to push enterprise capabilities out of the centralized data center and onto edge devices. Furthermore, Red Hat's virtualization stack is experiencing heightened adoption as a strategic hedge, as enterprise buyers seek viable alternatives to VMware following licensing changes and architectural uncertainties introduced by Broadcom's acquisition.
Threats and Disruptive Entrants
Despite strong positioning in regulated sectors, pronounced structural threats remain. The most immediate headwind is the persistent drag on the Consulting segment, where macroeconomic caution has constrained discretionary IT spending. As enterprise budgets remain tight, consulting growth has stagnated to low-single digits, forcing intense margin battles with leaner competitors like Accenture and specialized offshore integrators.
In the broader ecosystem, the threat of disruption comes from highly agile pure-play data platforms and agentic orchestration startups. Companies like Databricks and Snowflake are rapidly evolving from data warehouses into full-fledged artificial intelligence operating systems, actively targeting the same enterprise data layer that the watsonx.data lakehouse seeks to control. Additionally, the hyperscalers are continuously refining their own hybrid cloud solutions, attempting to bypass Red Hat and directly capture on-premises workloads. If the landscape consolidates around a few dominant, closed-ecosystem models, the open-source, multi-model approach could face adoption headwinds outside of strictly regulated industries.
Catalysts for Future Growth
Software revenue acceleration is heavily supported by an aggressive, targeted acquisition strategy designed to assemble a comprehensive infrastructure and data automation stack. The $6.4 billion acquisition of HashiCorp, finalized in early 2025, secured Terraform and Vault, solidifying a leadership position in infrastructure-as-code and multi-cloud security. This was followed by the multi-billion-dollar acquisition of Confluent, closing in the first quarter of 2026, which addresses the data-in-motion bottleneck. By integrating Confluent's real-time streaming capabilities, live unstructured enterprise data can now be fed directly into artificial intelligence models, creating a highly differentiated end-to-end data pipeline.
Looking further out, quantum computing remains a highly credible, asymmetrical growth driver. The company has consistently hit the milestones on its quantum roadmap and is projecting to achieve quantum advantage for select commercial applications by the end of 2026, with a clear path toward fault tolerance by 2029. While quantum revenues are currently immaterial, establishing dominance in early commercial quantum applications within cryptography, materials science, and complex optimization would solidify a foundational computing leadership position for the next generation.
Management Track Record
Under the stewardship of Chief Executive Officer Arvind Krishna and Chief Financial Officer James Kavanaugh, the track record of clinical execution has been exemplary. Management has successfully stripped away dilutive legacy operations, most notably through the spin-off of Kyndryl, and decisively pivoted capital allocation toward high-growth software and artificial intelligence. The financial outcomes validate the strategy: in 2025, the company achieved its highest revenue growth in years, alongside robust gross and operating margin expansion.
Most impressively, management has restored the cash-generating power of the business. Free cash flow reached $14.7 billion in 2025, marking the highest level of cash generation in over a decade and a record free cash flow margin. This capital efficiency provides the flexibility to sustain an elite dividend profile while simultaneously absorbing multi-billion-dollar acquisitions without compromising the balance sheet. A conservative but highly credible guidance framework has restored institutional trust, marking a definitive end to the era of chronic underperformance that plagued the business prior to the current management regime.
The Scorecard
International Business Machines Corporation has executed one of the most successful, albeit quiet, structural turnarounds in modern enterprise technology. By anchoring its strategy on the unglamorous but highly lucrative reality of hybrid cloud infrastructure and sovereign, regulated artificial intelligence, the company has built a highly defensible economic moat. The strategic synthesis of Z series mainframes running native inference, Red Hat managing the hybrid orchestration, and watsonx governing the data layer provides a comprehensive platform that hyperscalers cannot easily replicate within regulated industries. This operational discipline is fully visible in the company's record cash flow generation and software margin expansion.
While the Consulting segment remains a cyclical anchor and the broader artificial intelligence landscape is prone to rapid technological disruption by agile pure-plays, the downside risks are significantly mitigated by the profound stickiness of the core platforms. The recent acquisitions of HashiCorp and Confluent serve as force multipliers for the software division, effectively locking in the data and automation layers of the enterprise stack. Ultimately, the business presents a highly resilient, cash-generative profile that is uniquely positioned to monetize the enterprise transition from experimental artificial intelligence to governed, production-scale deployments.