Skip to content
a16za16z

Atlassian CEO on the SaaS Apocalypse, AI Agents & What Comes Next

Alex Rampell and Erik Torenberg speak with Mike Cannon-Brookes, cofounder and CEO of Atlassian, about how to make sense of the SaaS selloff, why not all software companies face the same AI-driven risks, and how Atlassian is thinking about the shift from records to processes. They also examine the real design challenge of getting everyday users to trust and benefit from AI agents in enterprise workflows. Timestamps: 00:00 — Introduction 03:16 — Risk Level Has Gone Up 06:44 — Three Types of SaaS Companies 11:07 — Vibe Coding Is Preposterous 14:23 — Businesses Are a Set of Processes 31:25 — AI Credits as Casino Chips 35:24 — How Atlassian Is Adapting 43:26 — Why Customer Trust Is Hard Resources: Follow Alex Rampell on X: https://twitter.com/arampell Follow Erik Torenberg on X: https://twitter.com/eriktorenberg Follow Mike Cannon-Brookes on X: https://twitter.com/mcannonbrookes Stay Updated: If you enjoyed this episode, be sure to like, subscribe, and share with your friends! Find a16z on X: https://twitter.com/a16z Find a16z on LinkedIn: https://www.linkedin.com/company/a16z Listen to the a16z Show on Spotify: https://open.spotify.com/show/5bC65RDvs3oxnLyqqvkUYX Listen to the a16z Show on Apple Podcasts: https://podcasts.apple.com/us/podcast/a16z-podcast/id842818711 Follow our host: https://x.com/eriktorenberg Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see http://a16z.com/disclosures.

Mike Cannon-BrookesguestAlex RampellhostErik Torenberghost
Mar 5, 202654mWatch on YouTube ↗

At a glance

WHAT IT’S REALLY ABOUT

Atlassian CEO explains SaaS reset, pricing shifts, and agentic workflows

  1. Public markets are repricing SaaS because AI increases uncertainty, but strong operators can adapt and even benefit if they modernize workflows and prove durability over time.
  2. AI shifts software from “filing cabinets turned into databases” to systems that can perform work, which changes which SaaS categories are at risk and which become more valuable.
  3. SaaS vulnerability depends on whether per-seat pricing is truly tied to human labor output, merely a “fairness” heuristic, or somewhere in-between—driving different outcomes for companies like Zendesk, Workday, Adobe, and Salesforce.
  4. “Vibe coding” is unlikely to replace complex enterprise systems with embedded edge cases, but it can dramatically increase extensibility by enabling cheaper, faster custom apps built on top of existing platforms.
  5. The hardest problems in enterprise AI are not model capability but design, trust, governance, and iteration loops that let humans and agents collaborate without confusion or runaway costs.

IDEAS WORTH REMEMBERING

5 ideas

AI raises SaaS risk, but “static world” thinking overstates the doom.

Markets fear that AI will commoditize software in 2–3 years, but that assumes companies and customers won’t adapt; in reality, incumbents can rework products, pricing, and workflows while continuing to execute on current demand.

Not all SaaS revenue is equally exposed to AI-driven seat reduction.

If seats are tightly coupled to human work that AI can automate (e.g., parts of customer support), per-seat revenue can collapse unless the vendor shifts to outcome/value pricing; if seats are a fairness-based metric divorced from work (e.g., per-employee HR pricing), the model can remain resilient.

Complex “edge cases” are durable moat; they’re hard to vibe-code from scratch.

Enterprise software often encodes years of exceptions from real-world operations, regulation, and geography (the “Indiana maternity leave” type problems); that tacit process knowledge makes full replacement risky and expensive despite improved code generation.

Vibe coding’s near-term killer app is cheap customization on top of stable platforms.

Instead of replacing Workday/Salesforce/Jira, AI-assisted building can create niche internal tools (e.g., a Miami-specific conference room workflow) that rely on underlying data, permissions, and business logic—making the core platform stickier.

“System of record” is an incomplete frame; enterprises run as interlocking processes.

Cannon-Brookes argues value lives in coordinating processes—some constrained by inbound demand (legal, customer tickets) and others constrained by creativity/output (marketing, product, engineering)—and AI changes each differently.

WORDS WORTH SAVING

5 quotes

Look, I think, um, the world is trying to work out how to rate or value software businesses in a highly disruptive stage, right?

— Mike Cannon-Brookes

The whole history of software from 1960 until 2022 was you would take a filing cabinet and you'd turn it into a database. The cool thing about everything that's happening in AI land is that the filing cabinet can do work.

— Alex Rampell

The idea I would vibe code my own workday and then run it is terrifying.

— Mike Cannon-Brookes

Businesses are a set of processes. They're not a system of record.

— Mike Cannon-Brookes

The AI credit world is really, really difficult for customers because they're like, "I don't really understand what this casino token you've given, casino chip you've given me is," right?

— Mike Cannon-Brookes

SaaS apocalypse and market valuation uncertaintyThree buckets of SaaS business models under AIPricing psychology (fairness) vs economic value deliveredSystems of record vs businesses as process networksInput-constrained vs output-constrained workVibe coding as extensibility layer (not full replacement)AI credits/consumption pricing and customer controllabilityAtlassian’s AI platform stack: context, graph, gateway, complianceAgentic workflows, human-in-the-loop trust, and UX iterationDocument creation paradigms and user learning curves

High quality AI-generated summary created from speaker-labeled transcript.

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.

Add to Chrome