What are the most and least promising categories to invest in, within enterprise software?

The enterprise software landscape is being reshaped by artificial intelligence, creating a sharp divergence in investment prospects between incumbent leaders and new entrants [2, 17]. For established players like Salesforce and ServiceNow, AI is largely seen as a sustaining innovation that enhances productivity and expands their addressable market [2, 3]. Their primary moats are not just code, but deeply embedded workflow integration, customer trust, data governance, and liability management—barriers that are difficult for new AI-native startups to replicate [5, 8, 12, 29]. However, this incumbency is being challenged by startups leveraging AI to drastically reduce customer switching costs from legacy systems, cutting migration times from over a year to **less than a month** [19, 27]. This dynamic creates a dual opportunity: value may accrue to both well-positioned incumbents who effectively integrate AI and new entrants who build AI-native solutions for previously unserved markets .

The most promising investment categories are consequently bifurcated. On one hand, the current market is considered ideal for acquiring high-quality, market-leading enterprise software companies at discounted valuations, while mediocre ones should be avoided . The proliferation of AI agents is also expected to create massive new security vulnerabilities, triggering a significant boom in the cybersecurity industry . On the other hand, AI is enabling the creation of entirely new software categories in verticals like legal, healthcare, and education, which have complex, unstructured workflows and lack dominant software incumbents [2, 6, 7, 21]. Initial enterprise AI adoption is being driven top-down into high-ROI use cases such as customer service, supply chain, and HR self-service, indicating immediate areas of opportunity . Foundational model providers like Anthropic and OpenAI are expected to act as an infrastructure layer, partnering with rather than competing against these application-focused SaaS companies [3, 14, 20].

Conversely, the broader, undifferentiated SaaS sector faces significant headwinds and investor skepticism . A major capital rotation is underway, shifting funds away from traditional SaaS and towards AI infrastructure, data centers, and semiconductors [4, 15]. This trend has fueled a "SaaSpocalypse" narrative, with one January 2024 report indicating a **10-to-1 ratio** of selling versus buying volume in software stocks [3, 26]. Investors are concerned about the durability of terminal values and the viability of per-seat pricing models in an AI-driven world [17, 24]. Vulnerable categories include single-function "point solutions" that can be easily replicated by AI platforms . A key long-term risk for the entire sector is the possibility that large enterprises may leverage AI to develop bespoke internal solutions, reducing their reliance on commercial software vendors altogether .