Feb 12, 2026 · 1h 24m · news
Anthropic's Super Bowl Ad: Who Won & Lost? | Sierra Hits $150M ARR: Is Customer Support Too Crowded? · 20VC with Harry Stebbings
Mike Cannon-Brookes · 35m spoken Rory O'Driscoll · 21m spoken Jason Lemkin · 12m spoken Harry Stebbings · 5m spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this episode of 20VC, host Harry Stebbings, Atlassian co-founder Mike Cannon-Brooks, and tech investors Rory O'Driscoll and Jason Lemkin dissect the future of software, the rise of agentic AI, and changing macroeconomic realities. The panel challenges the 'SaaS is dead' narrative, outlines the workforce displacement caused by AI agents, and explores the unique pressures of running public versus private companies.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 8.1% of the talking time here. How this is scored →
Harry as informed peer 3.9 Guest teaching 6.8 Guest disagreement 4.0 Harry pushing back 4.1
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 40:27 Jason forcefully rejects Rory's premise on software seat preservation
Jason flatly rejects Rory's argument that AI tools won't reduce headcount, shouting 'No, it's totally different' and asserting non-engineering seats are at existential risk.
Hardest push from Harry ▶ 1:13:24 Harry presses Mike on stock price drop versus operational narrativeHarry directly challenges guest Mike Cannon-Brooks by asking if he is failing to tell Atlassian's story given its severe stock price drop despite great quarterly numbers.
Biggest teaching moment ▶ 42:11 Mike introduces input-constrained vs output-constrained software domainsMike re-educates the entire panel on how to evaluate AI disruption across sectors by introducing a clear schema differentiating fixed-input domains (support, legal) from creative output-constrained domains (engineering).
Harry holds his own ▶ 1:06:48 Harry counters panel's Super Bowl ad cynicism with Wix CMO dataHarry uses his own domain knowledge from interviewing Wix's CMO to push back against Rory and Mike's dismissal of Super Bowl ads as pure founder ego.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Episode Hooks and Title Sequence | 4 | 6 | 3 | 2 | Harry opens the show probing whether Anthropic's $149B revenue target will swallow enterprise software budgets. Mike educates the panel on revenue stacking, showing how SaaS usage of models flows through AWS back to Anthropic rather than acting as a zero-sum cannibalization of existing software. | |
| The Future of IT Consulting and Systems Integration | 5 | 5 | 4 | 5 | Harry offers a counter-hypothesis that IT integration and consulting spend via firms like Accenture will expand rather than contract during early AI cycles. Rory and Jason debate whether system integration consulting gets automated or constrained by wizard-level talent requirements. | |
| Is SaaS Dead? Software's Real Survival Rate | 2 | 7 | 5 | 2 | Mike firmly rejects the narrative that SaaS as a category is dead, labeling the market panic ludicrous and citing historical tech churn. He demonstrates Atlassian's health by highlighting accelerating cloud revenue growth and 44% RPO growth driven by multi-year enterprise commitments. | |
| LLM Coding, Infrastructure Margins, and the Fear-Driven SaaS Sell-Off | 3 | 7 | 3 | 3 | Mike explains how Atlassian's 10,000-person R&D team leverages LLM coding tools while actively optimizing AI infrastructure costs, resulting in expanding gross margins over six quarters. Harry voices typical VC investor anxiety over public SaaS holdings, which Mike attributes to short-term fear. | |
| Public SaaS Deceleration, Private Equity, and the IPO Drought | 1 | 7 | 3 | 1 | Jason and Mike analyze why public SaaS indices show overall deceleration, attributing it to private equity buyouts removing public peers and top-growth private companies staying private longer. Harry stays in the background as the panel breaks down market structure dynamics. | |
| The Harvey Deal: Evaluating High-Valuation Law Tech and TAM | 5 | 7 | 4 | 6 | Harry questions the risk-reward ratio of pricing Harvey at $11B ($600M projected ARR) versus buying mature public SaaS like Atlassian. Mike schooled the room on TAM models by sharing Accel's 2010 investment model in Atlassian, showing how elite companies dramatically outperform VC terminal growth assumptions. | |
| The Shift to Agentic Software and non-Engineering Headcount Reduction | 2 | 8 | 6 | 2 | Jason forcefully argues that non-engineering seats across B2B are facing existential downsizing from agentic AI. Mike reframes the entire topic by introducing his framework of input-constrained domains (like legal and support) versus output-constrained creative domains (like software engineering). | |
| The Customer Support Battlefield: Sierra, Zendesk, and Legacy Disruption | 6 | 7 | 5 | 7 | Harry strongly pushes back against the trend of funding customer support AI startups, asserting that the space is overfunded and unsafe for venture dollars given incumbent competition. Mike clarifies that high-value support agents focus on automated action execution rather than simple Q&A. | |
| The Anthropic Super Bowl Ad: Ego, Signals, and tech Maximalism | 6 | 6 | 5 | 6 | Rory dismisses Anthropic and OpenAI Super Bowl ads as top-of-bubble ego signaling between tech executives. Harry pushes back on the panel's cynicism by citing concrete ROI and impression stats directly from Wix's CMO after his multi-hour interview. | |
| Public vs. Private Constraints, Capital Allocation, and founder Resilience | 5 | 8 | 2 | 7 | Harry directly challenges Mike on whether Atlassian's depressed stock price indicates a failure in communicating its AI transformation story. Mike responds with a masterclass on founder resilience, accepting reality, and building long-term enterprise value through product creation. |
Statements from this episode (28)
Cannon-Brookes: The idea that software is dead is ludicrous
“The idea that software as a category is dead is ludicrous to me.”
Lemkin: Investors must abandon traditional TAM calculations for revenue trajectory
“I just think we have to give up on TAM. We just have to let the revenue show us the path to TAM.”
Lemkin: Non-engineering software categories face existential seat shrinking
“Every category that I know of outside of engineering and product is at existential risk of shrinking seats.”
O'Driscoll: Safe consensus bets in AI signal a market bubble top
“It's the venture capital equivalent of you can't get fired for buying IBM. Right? You can't get fired for sticking money in with a one X preference on the consensus winner. This is classic top of the bubble stuff.”
Cannon-Brookes: Anthropic is one of Atlassian's largest AI models
“We use Anthropic a lot. It's one of our biggest models. We use, you know, multiple models which I think is what most good SaaS vendors are doing within their customers' choices, right? Like we use a lot of Gemini, a lot of Anthropic. We have a whole bunch of L…”
Stebbings: IT consultants will out-earn foundation models in short-term revenue
“I think they will be more revenue generative than the foundation model companies in the next two to three years, as they often are in early stages of kind of cycles.”
Lemkin: Customer success is dead as SaaS companies fail to convert teams
“This is why customer success is dead because all these B to B companies are trying to convert their CS teams to FDs. And I don't think even 10% of them can do it.”
O'Driscoll: Mass AI adoption requires product simplification, not consulting wizards
“If the only people who can roll out your shit are wizards, then you're going to run out of wizards, and then you're going to run out of revenue. So have mass adoption, you're going to need mass simplification.”
Lemkin: Replit hits $300M revenue with 300 employees and 11 GTM staff
“I was at Replit the other day at three hundred million in revenue, 300 people. It's, there's like 11 people in go to market.”
Cannon-Brookes: Atlassian cloud revenue grew 26% past $5B with 44% RPO growth
“We're accelerating our cloud revenue, which did a 26% quarter growth rate at well north of five, six billion at the moment. And we grew our RPO by 44%.”
Cannon-Brookes: Some Atlassian AI features are 1,000x cheaper to run today
“Some of our features are a thousand times cheaper to run than when we introduced them today.”
Cannon-Brookes: Atlassian gross margins increased over the past six quarters
“Our cogs has gone up, or rather cogs has gone down, right? Our gross margin has gone up. Last six or seven quarters.”
O'Driscoll: Median public SaaS growth held at 30% via IPO replacement
“It was 30% for about a decade and a half, but when you went into it, that's because anyone who goes, to your point, Jason, when you go below 10, you get the red card, you're sent off, you're told, you know, you go off. P.E. Land, but they were dropping another…”
Cannon-Brookes: Only four public CEOs remain from 2020 SaaS peer group
“In COVID in 2020, we formed a group of SaaS CEOs. I think there were 16 of us that got together, we'd meet on Zoom every single week, right? At one stage, 10, 11 of us were public, maybe more, we all had earnings calls, and you know, it was a really good thera…”
Cannon-Brookes: Technology's fundamental role is to change and expand TAM
“Our whole job in technology is to change TAM, actually.”
Cannon-Brookes: Accel's $60M Atlassian investment in 2010 was 100% secondary
“When the Excel guys invested in Atlassian, right? It was all secondary. They put sixty million bucks in 2010. And they bid highest of the five VCs.”
Lemkin: GTM AI agents charge $50K-$100K by eliminating human labor
“They're all 50 to a hundred grand because you get rid of the people.”
Lemkin: No software companies are reducing their engineering teams
“We are building more software than ever, and we need the humans and no one is reducing their engineering team.”
Cannon-Brookes: Software engineering is output-constrained as roadmaps never end
“The difference with something like software or engineering is it's creation, right? It's how much can you make, right? The roadmap is never finished. You can keep creating more.”
Lemkin: Customer support has seen the highest hiring reductions in tech
“No category has been more decimated in hiring than support. No one's hiring. Like it's radically down.”
Lemkin: Customer support AI is a Trojan horse for horizontal workflows
“The super interesting thing about support is it becomes a Trojan horse to do everything else. Because when that's the main agent, if that is the main agent you're working with, you'll start using it for sales help or for research or for marketing or for other …”
O'Driscoll: Scale passed on direct competitors to Sierra and Decagon
“We looked at a number of companies that were behind Sierra and Decagon and came to the conclusion that a direct head-to-head Against those companies, when you're a smart, it's just a little bit too hard.”
Cannon-Brookes: VCs waste capital following tech trends for five to six years
“That means a lot of capital is raised and deployed and a whole bunch of it turns out to be torched because VCs are very individual thinkers that they pile into the two winners and then the next 10 companies and probably the next 15 and a whole bunch of those d…”
Cannon-Brookes: VCs back OpenAI at huge valuations for logo marketing
“One thing that they don't say is a lot of VCs want the loose side on the wall, so when you walk into the office, on the wall is a nice logo that says, oh, you invested in OpenAI. You invested in Anthropic. Well, if you put money at three hundred fifty billion,…”
Lemkin: Anthropic is a generational company whose only weakness is consumer
“Anthropic is, I mean, it's a generational, generational company. It changed coding. It changed everything. But the only place Anthropic is weak is consumer.”
Cannon-Brookes: Public software firms neglecting AI R&D are in trouble
“I would argue if you're a public company today and you're not spending a lot of money on fundamental kind of new R&D around AI and your product category, you're probably in trouble.”