On Tuesday, February 3, 2026 A meteorite fell and many dinosaurs were seriously damaged. It was the day that $285 billion evaporated from the technology stock market following Anthropic’s Claude Cowork Legal announcement. Thomson Reuters fell 16%, RELX 14%, Adobe hit six-year lows. The enterprise software sector went into absolute panic and technology companies in general trembled. But some of the dinosaurs held up better than others, as happened with IBM. It is not that the technological history emerged unscathedits shares lost more than 6% that week. Investors feared that Anthropic’s new plugins, especially legal and data analytics, would directly threaten IBM’s premium software and consulting business.
But here is the really interesting thing about this collapse: while analysts downgraded dozens of companies in the sector to “sell”, Jim Cramer went on CNBC saying that long-term investors should Buy IBM right now. Their argument: IBM has exceeded profit expectations for 17 consecutive quarters, projects 5% growth in revenue and a $1 billion increase in free cash flow by 2026, and has seasonal patterns that historically close February and March in the green. Wall Street agrees: the consensus remains “moderate buy” with Price targets implying a 30% upside potential.
How is it possible that IBM, with 115 years behind it and a business apparently vulnerable to agentic AI, be considered a safer bet than technology startups of the last decade? I have followed IBM for a long time and I am fascinated by one thing that it has masterfully demonstrated many times: reinventing itself without losing its essence. While the tech sector was sinking, IBM presented results that confirm its transformation: 8% growth in revenue, record free cash flow of $14.7 billion, its software segment expanding 9%, and a stock price that has more than doubled in three years. But if IBM fails in something, especially in Spain, it is in the story, in its marketing strategy. And that’s a real shame.
Why IBM fell only 6% when others plummeted 16%
The difference between losing 6% and losing 16% is not trivial in a market of 270,000 million capitalization. It represents the distinction between a technical correction and a structural vote of no confidence. IBM fell, yes, but it fell less. And there are three key reasons for Big Blue’s relative resilience.
The first reason is that IBM’s business model does not depend on charging for software seats. While companies like Thomson Reuters or LexisNexis live on the “so many users, so many licenses” model, IBM has been pivoting towards consulting and implementation services for years. Of the 12.5 billion in its AI portfolio, 80% are services, not licenses. The company does not sell access to a platform, it sells systems running in production. If Anthropic automates contract review, the client still needs someone to integrate that into their critical workflows, to manage data governance, to ensure compliance. That someone is still IBM.
The second reason is the strategic commitment to small and specialized models versus giant generic LLMs. IBM’s Granite models don’t compete with GPT-4 or Claude on overall capability, they compete on specific enterprise use cases that can run on-premise or in private clouds. When Anthropic releases a plugin that works through its APIs, IBM can respond with implementations that keep data within the client’s security perimeter without sharing it with the provider. For regulated sectors such as banking, insurance or health, that It is not an advantage, it is an unbreakable regulatory requirement.
The third reason is the hybrid architecture. With Red Hat and now Confluent, IBM does not force its customers to choose between public cloud and on-premises. Orchestra both. When agentic AI threatens to commoditize traditional software, the ability to integrate legacy systems with new AI capabilities without migrating everything to the public cloud becomes a critical differentiator. IBM isn’t selling the cheapest cloud or the most powerful LLM. You’re selling the integration that makes everything work together, something that traditional companies with decades of technical debt desperately need.
These three reasons explain why, although IBM fell, analysts did not panic. In fact, they did the opposite: Jefferies raised their price target from $360 to $370, Oppenheimer set it at $380, and Bernstein and Stifel raised their recommendations following the Q4 results. The market understands that IBM plays a different game than the rest of the sector.
How to survive the meteorite
The first lesson offered by the IBM case is clear: Don’t compete where you can’t winorchestrate. IBM didn’t try to beat AWS, Azure, or GCP on their own turf. With the acquisition of Red Hat in 2019, it created a hybrid orchestration layer that allows companies to mix public hyperscalers with on-premises infrastructure and private clouds for sensitive workloads. This strategy has profound implications for IT departments who have been trying for years to justify 100% migrations to the public cloud for critical systems such as financial, compliance or customer data.
The hybrid model It is not a compromise solutionis intelligent architecture backed by strong financial results. The acquisition of Confluent for 11,000 million in December 2025 consolidated IBM’s position in data streaming, crucial for agentic AI. But the key is in the approach: the company did not sell migration to the cloud, sold orchestration where it makes strategic sense. The practical application of this lesson involves auditing which workloads are in the public cloud by strategic decision and which are there simply because they had to be in the cloud. Many organizations would find that they have workloads paying a premium in the cloud when they could be on hybrid infrastructure with better performance and lower total cost of ownership.
The second lesson is about small models versus giant LLMs. As the industry chased size, IBM launched Granite: compact, specialized models for businesses. This allows deploy autonomous AI agents that do not depend on an external “mothership LLM”since they can reason, use their own data and execute critical tasks operating entirely on the client’s local infrastructure.
The strategic importance of this approach lies in the control of sensitive data. Organizations that submit critical information to GPT-4 or Claude via public APIs are taking on unnecessary compliance and security risks. Open weight Small Language Models run on-premise or in dedicated VPCs are sufficient for 80% of enterprise use cases without exposing data outside the corporate perimeter. The WatsonX platform allows you to build AI agents trained on corporate data using open weight models, and financial results show that It’s not vaporware: IBM’s AI portfolio reached 12.5 billion in 2025, double the previous year.
The third lesson is perhaps the most important: Implementation Consulting Beats Software Licensing in the World of Enterprise AI. Of IBM’s $12.5 billion AI portfolio, 80% are consulting services, not software licenses. The company does not sell pilots, it sells systems in production running in banks, the secondary sector and governments. This distribution reveals something fundamental: the value in enterprise AI is not in having access to an LLM, it is in correctly integrating it into critical workflows with proper governance. Providers that only sell access to the platform without helping to implement is a model that can suffer, as has been seen. The practical implication is that when evaluating AI vendors, organizations should ask not just about the technology but about production use cases similar to their own. If the answer is “we have many pilots”, it is a warning sign. If the answer includes customers in production with measurable ROI, the supplier probably understands the real business beyond marketing.
A big communication problem
And here comes the big problem. As I said at the beginning, I am a fan of IBM as a specialized journalist and as a technology fan. When artificial intelligence hit the headlines these years, IBM had been using its Watson for 70 years, facing it and even beating chess grandmasters and yet unable to sell it at a business level. She has never been considered a protagonist of the AI revolution, despite undoubtedly being one of the greatest pioneers. And it is one of the most innovative companies in technology, with more than $8.3 billion annually in R&D and a portfolio with more than 100,000 patents, not to mention its advances in quantum computing. IBM, as we have mentioned, has also executed one of the most profound reinventions in history, pivoting to the hybrid cloud after the acquisition of Red Hat.
But in the Spanish market all this is almost a secret. The reason is frustrating: a marketing deficit that borders on invisibility. I know this first-hand as a spectator and as a communications professional in the sector. And it seems to me to be a big mistake, a huge wasted capital of investments, research and strategic successes that are forgotten. We do not know the strategic commitment, we do not know the technological commitment and we do not know the success stories. The IBM case is a symptom of a broader problem in the Spanish technology market: the systematic confusion between media presence and real technological leadership. Companies with mediocre solutions but aggressive marketing capture disproportionate mindshare. Companies with solid solutions but poor communication disappear from the map of options considered.
But it is not only an image problem but it is something much more serious. It is a lack of investment in marketing and adequate communication that can be fatal, especially in an era where agents and artificial intelligence will have the last word for investment in technology and not be present with relevant content (interviews, success stories…) in specialized media. It will be a great handicap for your relevance and your solutions will not be considered by agents.
We will closely follow IBM in a battle that appears exciting and that twists the script daily. We will see if these strategic successes and technological bets mature and if at once they manage to understand the importance of investment in marketing and that this dinosaur finally dares to leave the cave.
