A good friend sent me the article below, whose central claim is that “all workflow-driven enterprise SaaS companies will be either disrupted or dead in the next decade.”
https://www.msn.com/en-us/news/other/disrupted-or-dead-ai-is-crushing-a-generation-of-startups-built-before-chatgpt/ar-AA24xKZO?uxmode=ruby&ctsrc=dgst&ocid=edgdhpruby&pc=LCTS&cvid=6a1ea1476a9e40dbb5a9c77d02abe2c1&ei=29
It’s really good click bait, but the 70+ SaaS earnings calls we reviewed in Q4 and the 30+ Q1 earnings calls that we’ve reviewed thus far refute this.
Zeta Global’s CEO said “customers want to invest in applied AI, not road map AI,” then delivered the company’s 19th consecutive beat-and-raise quarter with revenue up 50% year-over-year thanks in large part to their AI positioning. Klaviyo’s CEO said “agents are only as good as the systems beneath them” then reported 28% revenue growth, 110% net dollar retention, and the company’s first positive GAAP operating margin since going public thanks to it’s AI positioning. Dynatrace’s CEO said customers require “answers, not guesses” then reported four consecutive quarters of 16% ARR growth and crossed $2 billion in ARR. Palantir’s CEO said it philosophically: “AI without an ontology is slop — you can go nowhere without it” then delivered 85% revenue growth and a Rule of 40 score of 145, the highest in enterprise software history (allegedly).
AI models are a commodity. What matters is the data AI runs on, the workflows it operates within, and the governance layer that makes its outputs trustworthy. These are the rails that software provides. MongoDB’s new CEO CJ Desai made the architectural version of this argument more precisely than anyone this quarter: “A rigid relational schema becomes a tax on every iteration. LLMs speak in unstructured document-shaped data, the exact form MongoDB was built around. We have been compounding both advantages for 15 years, well before the current AI wave gave them a tailwind.” Revenue from their Atlas product grew 29.4% year-over-year for the fourth consecutive quarter. Adobe, Zomato, and multiple Frontier Labs chose MongoDB as the memory layer for their most ambitious AI agent deployments. Braze’s CEO made the same historical argument from the customer engagement layer: “The same flaws that held back our competitors during the mobile and stream processing eras are hindering them again in the age of AI. The architectures that could not handle real-time data then cannot harness advanced AI today.” Braze delivered its fourth consecutive quarter of revenue growth acceleration, reaching 30% year-over-year growth with record free cash flow. The companies that made the right architectural choices before AI was the reason to make them, are now collecting a structural dividend that no new entrant can replicate in any reasonable timeframe.
Let’s be specific about what Q1 2026 enterprise software earnings actually showed, because the numbers are the argument:
Palantir: revenue +85% year-over-year, Rule of 40 score of 145, raised full-year guidance to $7.65 billion. Salesforce: $1 billion in Agentforce ARR growing 205% year-over-year, 3.8 billion Agentic Work Units delivered, 28.6 trillion tokens processed, record non-GAAP operating margin of 34.8%. Figma: revenue +46%, NDR 139%, paid customers +54%, AI credit monetization blasting off — the quarter after Anthropic launched a direct competitor and sent the stock down 16%. MongoDB: Atlas revenue +29% for the fourth consecutive quarter, $688 million total revenue beating guidance, customers with $100K+ ARR growing 16%, RPO up 88% year-over-year. CrowdStrike: $5.25 billion in ARR, first year exceeding $1 billion in net new ARR, AIDR growing 5x in its first weeks of availability. Cloudflare: revenue +34%, processing hundreds of billions of agentic requests per month, 97% of R&D employees using AI coding tools. Datadog: first $1 billion revenue quarter, new logo bookings at an all-time record, AI customers growing 20% of total but representing 80% of ARR. Braze: fourth consecutive quarter of revenue growth acceleration at 30% year-over-year, record free cash flow, NDR improving to 110%. Snowflake: product revenue +27%, RPO of $9.77 billion up 42%, largest deal in company history at $400 million, Snowflake Intelligence nearly doubling quarter-over-quarter to 2,500 accounts. JFrog: revenue +26%, cloud crossing 50% of revenue for the first time, stock +23% on earnings day.
In summary, these are not the numbers of an industry being disrupted. These are the numbers of an industry with significant tailwinds at its back and the sun shining on its face.
Here is the finding that surprised us most across 70+ transcripts: the same AI wave that allegedly threatens enterprise software is empirically driving demand for it across every segment we analyzed. AI-generated code ships faster and with less human review than human-written code which means more bugs, more vulnerabilities, and more system complexity per unit of time than ever before. Every observability and security company cited AI-generated code velocity as a demand accelerator for their core products. Datadog’s Pomel described the dynamic most precisely: AI training workloads have gone from “artisanal, one-off research projects to production workloads that have to be on all the time, reliable, and every minute you lose or every failure you have in your training round is a week you give away to the competition.”
Datadog’s Bits AI Security Analyst reduced security investigations from hours to 30 seconds. JFrog’s Ben Haim called artifact governance for AI-generated code “existential, not optional” and delivered 26% revenue growth and its first quarter of cloud revenue exceeding 50%. AI agents proliferate with elevated permissions which means more identity governance demand for SailPoint, where non-human identities accounted for 25% of Q4 SaaS identity growth, and more zero-trust enforcement demand for Zscaler, whose AI Protect product crossed $100 million in annual bookings. AI content floods the internet which means more brand safety demand for DoubleVerify, whose Slop Stopper product applied to 40% of all impressions, while AI-powered bot variants increased 140% year-over-year.
CrowdStrike’s Kurtz said it most directly from Q1: “AIDR grew more than 5x versus the prior quarter in its first weeks. Our sensors detected more than 1,800 distinct AI applications and approximately 160 million unique application instances across our customer base. The question is not whether AI creates new attack surface. It does. The question is who governs it.” The answer, quarter after quarter, is the incumbent software platforms that have spent years building the governance infrastructure AI requires.
AI is expanding the TAM of nearly every incumbent enterprise software company we’ve looked at.
Waystar’s CEO Matt Hawkins described this most directly: “the approximately $100 billion in annual revenue cycle labor services performed across the industry today, we believe we are well-positioned to automate a meaningful portion of this labor pool.” AI-powered capabilities already drove roughly 40% of new bookings in Q1. Procore’s CEO made the same argument for construction: “Our path forward is defined by a powerful economic duality, upside opportunity through AI monetization and downside protection through our volume-based model. Procore AI can access construction labor budgets well beyond the industry’s software spend.” Backblaze’s CEO described a $14 billion Neocloud data lake opportunity by 2030 from a base that didn’t exist before AI training workloads required cost-efficient cold storage at scale. PubMatic’s CEO compared AgenticOS to the introduction of real-time bidding in 2010 calling it “the third transformation” of digital advertising, with 1,000+ AI-powered deals transacting in the first months of availability. MongoDB introduced the newest and potentially largest TAM expansion: the memory layer for AI agents themselves. Adobe’s Journey Agent, Zomato’s Nugget platform orchestrating 15 million conversations per month, and multiple Frontier Labs are all choosing MongoDB as the cognitive infrastructure their AI agents reason from. Services TAMs are becoming software TAMs.
The agentic era is assembling a new infrastructure stack and it is being assembled by incumbent software companies, not AI startups. MongoDB owns the memory layer. UiPath owns the automation execution layer; its CEO described the closed loop with unusual precision: “AI creates automation. You run those automations, it is very cheap to run, very deterministic, reliable, auditable and only when these scripts break do you invoke again AI to fix the scripts.” PagerDuty owns the operations control plane for AI workloads, with its CEO describing a decade of proprietary incident resolution data that no generic AI agent can replicate. JFrog owns the artifact governance layer for every binary, package, and AI model that moves through an enterprise software pipeline. Palantir owns the ontology layer that grounds AI agents in the actual structure of each enterprise. Cloudflare owns the network edge through which hundreds of billions of agentic requests flow every month.
None of these companies invented these positions in the last 18 months. All of them built the underlying infrastructure over years or decades, for reasons that had nothing to do with AI agents. The agentic era arrived and found the infrastructure already in place. Salesforce’s Benioff named the disruption thesis directly and then refuted it with numbers: “You might have heard about the SaaSpocalypse. It is not the end of software. It is the end of software that makes humans do all the work. Human being, agent doing. This is the Agentic Enterprise.” Betting against software in the AI era is betting against the people who built the infrastructure the AI era runs on. Good luck with that.
Thank you for your readership. See more blogs and SaaS data at blossomstreetventures.com. Email the author at sammy@blossomstreetventures.com.