All Insights
Moving the Conversation Beyond the AI Bubble
There’s been plenty of talk about an AI bubble, reflecting skepticism about whether the massive investment in AI infrastructure will yield economic value. At Vista, we believe the conversation needs to shift. The scale of investment signals a once-in-a-generation transformation. As this foundation matures, we believe its value is likely to be realized through AI software and applications.
So far, most attention has centered on hardware and infrastructure —chips, data centers and large language models. These investments are essential to building the base on which AI runs. Bain & Company estimates roughly $9 trillion in cumulative AI value has been created since 2023, with more than 80% captured by hardware and infrastructure. NVIDIA alone now accounts for more than $5 trillion in market capitalization as of October 2025.1
Bain’s forecast shows that the next $5–7 trillion of AI-driven value by 2030 will come from software and applications, as enterprises integrate AI into workflows, products and decision systems. This segment is projected to grow 4.5 times between 2025 and 2030, becoming a dominant source of durable value creation alongside continued infrastructure expansion.1
What’s Driving This?
We believe the next era of AI products will be able to take action and perform real work. Agentic AI marks a structural shift in how tasks are executed, automating complex processes across technology systems. While Bain’s projection includes both Generative and Agentic AI applications, Vista sees agentic as a powerful tailwind, extending AI’s impact beyond content creation to full workflow automation.
We are already seeing tangible results across the Vista portfolio. More than half of Vista companies are successfully monetizing AI products and features, and 30 companies are building and launching agents in-market. This is supported by Vista’s Agentic Factory, a collaboration between our Value Creation Team and Microsoft, Google, AWS and Anthropic.2 One example: Gainsight’s Atlas Agents, developed on Microsoft Azure, automate customer renewal processes, cutting cycle times from seven days to one while reducing risk of customer churn by 90.3
Where Value Will Accrue
Markets have yet to fully recognize where value will accrue. Software valuations remain near 10-year average even as profitability and margins improve. Meanwhile, leading AI application companies now exceed $5–30 billion in valuations, signaling growing investor confidence in the software layer.4 Emerging performance metrics —revenue per employee, compute efficiency and the “Rule of 60”—will define how AI-enabled businesses are evaluated.
As with past innovation cycles, infrastructure leads but applications endure by embedding intelligence into daily operations. The companies best positioned to win will maintain sovereignty and dominion over data and workflows, a defining advantage as only about 1% of enterprise data is used to train public large language models.5 We believe proprietary data and workflow expertise will be the scarce resources around which next-generation systems are built.
Continued infrastructure investment is expected to continue to reduce the marginal cost of compute and accelerate development cycles, creating favorable conditions for AI-driven software growth.
Software leaders—both incumbents and start-ups—must stay focused on delivering customer outcomes quickly and efficiently. Success will depend on embedding AI deeply into products, workflows and data. We believe those with proprietary enterprise data will hold a competitive edge.
At Vista, we’ve navigated transformative technology shifts before. Over the past decade, our disciplined approach helped modernize companies from on-premise software to SaaS and from SaaS to cloud-native architectures. That operational intelligence now guides our approach to AI.
Agentic AI Is Here – What Investors Need to Know
Driving Innovation Across the Software Ecosystem
