The week the “AI is just a feature we’ll add” argument stopped working, and three pay bands replaced two.
This week on 20VC x SaaStr with Harry Stebbings, Jason Lemkin and Rory O’Driscoll, we started somewhere other than OpenAI and Anthropic for once: Canva cutting its 2026 growth forecast by a third as AI serving costs blow up. That led into Google losing Jeff Dean and Demis Hassabis, Elon’s $16.8B first installment on Terrafab, Revolut’s reported $50B CEO package, and Whatnot raising at $20B.
1. Canva going from 30% to 20% growth is an existential question, not a gross margin one
Canva did roughly $3B in GAAP revenue last year and entered 2026 growing 30%. Melanie Perkins disclosed mid-year that they’ll likely finish the year around 20%. Part of the stated reason: the AI features they’re shipping cost real money, and subsidizing frontier model calls across a prosumer base is expensive.
Rory’s read: the margin math is a second-order business model issue. What matters is whether 30 to 20 is on the way to 10. There are three large creative software companies right now: Adobe at $23B growing 12% and trading at three to four times revenue, Figma public at $1.4B growing 40%, and Canva private at ~$3.6B growing 20%. All three face the same question. Is AI a feature you incorporate, or a new thing that makes you obsolete?
2. Our agents never once suggested Canva
We churned Canva. We also churned Notion. Neither did anything wrong. We just no longer had a need for them in an agentic operating model.
Jason’s read: we built our own ad server and creative generation network at SaaStr, running entirely on our agents. It never occurred to the agent to use Canva for any of it. Not once. Users switching to a competitor is a fixable problem. The layer making the decision not having your product in its consideration set is a different kind of problem.
Harry connected it to what he heard from the President of Uber, whose single biggest fear is the disaggregation of UI: you say you want a car, and ChatGPT routes you to whichever provider is cheapest. Our agents already did that to Canva.
3. The era of no-code is winding down, and it covers more companies than people count
Amjad at Replit said it about Airtable, not Canva, but it applies to all of them. Airtable was a no-code database disguised as a spreadsheet. Notion is a no-code database disguised as a word processor. Canva was a no-code way to design things, and it was breathtakingly disruptive: you didn’t need a designer, you didn’t need HTML.
Jason’s read: the entire category of “things humans can do without engineering resources” is winding down, because AI does those things natively. If the job can be done inside ChatGPT or Claude, even accidentally, the standalone tool is in trouble.
4. Figma is more protected than Canva, and the reason is bureaucracy
Rory’s read: Figma is an enterprise product. Large groups of people coordinating on building software. Even if you fully automate the creative act, you still have corporate process and workflow to monetize. Canva is the prosumer motion: I want a flyer, I want a cheap website, I want a piece of content. That’s where AI is most accessible, because you type the request and get the output.
Figma took its hit publicly. Dylan was clear that gross margins will be meaningfully impaired as agentic products get used, and the stock traded down 20%. Taking that hit is more painful in public than in private, which is part of why Canva’s situation reads differently even with worse numbers.
5. Canva is probably worth around $12B today
Jason’s read: 20% growth at ~$4B ARR against current public comps, decelerating rather than accelerating. There’s some rule of 40 adjustment, but $12B is where it lands.
Rory’s pushback: you can find companies at 20-25% GAAP revenue growth, free cash flow positive, trading well above that. Datadog, Cloudflare, JFrog: mid-20s growth, 20%+ operating margins, 15 to 17 times NTM. The difference is that nobody is asking whether those companies exist in five years. If Canva carries the existential discount, $12B is right. If they transcend it, it’s $12B and up. Either way, liquidity only comes at the end of the journey, and that journey is at least 12 months long.
Harry’s note for LPs holding Canva: when the private window shuts and appetite dies, getting anything done at scale becomes very difficult.
6. After Airtable and Canva, be extra skeptical of marks
Jason’s read: these are two events that quietly hit old marks and are difficult to hide behind. You should probably have marked them down last year. “We raised at $42B, therefore we’re holding at $42B” is no longer defensible.
Rory’s method: ground yourself in the actual growth rate first. We’re doing a billion growing at 30%, that’s this multiple. We’re doing a billion growing at 10%, that’s a very different multiple. Only then do you grade up or down for existential risk versus AI lift.
7. Selling along the way is mathematically true and strategically dangerous for small funds
Freestyle’s Dave Samuels noted their blended exit price on Airtable was around $6B, making the case for trimming in good times.
Rory’s data: roughly 70% of the time you should have sold, 30% you should have held. But the holders compound forever, and the ones you sold compound at zero from that point on. It’s the Bessembinder result: sub-1% of companies deliver 90% of the capital gains in public markets, and it’s the same in private.
Jason’s read: for a smaller fund, the math gets tricky fast. I’m lucky to have three fund returners. If I start taking early exits on those and never get a 10x fund returner, the LPs don’t get the 5x, 6x, 8x fund they’re underwriting. I ran the analysis across my entire history, angel and venture, on who I should have sold and who I should have held, and it broke roughly 50/50.
8. Companies that hesitated in 2023 and 2024 are paying the bill in 2026 and 2027
Rory’s frame: there are moments when a crevice opens between the before and the after, and if you’re going to make the jump you have to make it fast. Eventually the gap is too wide.
The best example on the apps side is Palantir, which went from 18% growth to 98% growth. Two things made it work. First, real forward deployed engineers, people who’d spent a decade and a half deploying operational change in the field, so when customers needed AI they already had the humans to do it. Most companies have solution architects wearing an FDE t-shirt. Second, genuine outcome-based pricing. Give me $2B, I’ll save you $8B. Everyone talks about outcome-based pricing. Almost nobody puts a $2B contract on the line for it.
Jason’s read: Replit is the harder version of the same story. Six years in the wilderness as a nerdy web IDE, then the models arrived and they rethought everything. Palantir could have chugged along at 20% and been fine. They grabbed the moment instead.
Worth separating out: Datadog, Cloudflare, Palo Alto and Twilio didn’t have to reinvent anything. They sell infrastructure into the largest infrastructure boom in history. Showing up and selling more is a very different job than the one facing Canva’s CEO.
9. Jeff Dean left over compute allocation, not money
Jeff Dean left Google after 27 years, taking three researchers with him. Demis moved into a chairman role. The stock dropped a couple hundred billion in market cap, which as Rory noted is an unusually validating exit for a non-CEO.
Rory’s read: look at how compute gets allocated at Google. Compute given to Cloud converts into 30% operating margins immediately, because they can sell it to Anthropic. Compute given to Gemini might produce a competitive coding model and consumer revenue. Compute given to drug discovery or materials science is a five to seven year moonshot indulged at the 10% level. If your work is permanently third in line, you leave.
Jason’s read: I was an SVP at Adobe running the number three business unit. It sucked. You sit with 50 other VPs and never get to talk about what you’re working on. We were doing $800M and it didn’t matter. If you can take your whole team and raise a billion dollars to do exactly what you want, you go.
The venture appetite for science-based moonshot labs has never been higher. Three years ago that money did not exist at this scale.
10. There are three compensation bands now, not two
When we started this show there were two: normal people and AI people. Now there’s a third.
Jason’s read: regular employees, the AI team whose salary bands you already broke, and the one to five superstars who require a seven figure package and equity 10 times what an employee at that stage would normally get. At $200M in revenue you can carry four god-tier employees without breaking the model. It’s rough on the rest of the team and nobody does it happily, but you are not going to pull off a Palantir or an Intercom without a skunkworks and a god tier.
Rory’s caution: someone who took $1M in Anthropic stock in 2023 is reportedly sitting on $51M. That signal ripples through the entire hiring market and distorts what everyone thinks is possible. Nobody hired today is getting 50x.
11. If you’re an app company, stop trying to hire frontier model talent
Rory’s read: when you built software in the PC era, you had fourth-tier chip talent, because you weren’t building a chip. If you’re an AI application company that feels compelled to build a frontier model, you’ve put yourself in direct competition with Anthropic and OpenAI and you will lose. Make the model a complement. Have A-tier talent in UI, in implementation, in your data domain, in fine-tuning.
Jason’s read: the old founder test was “did you get an offer from Google,” which told you someone was top 5%. The new test is “did you get an offer from Anthropic or OpenAI, and what was the actual job?” A lot of those seven-figure jobs are watermarking or tuning a UI animation for 18 months. Interview enough people and you will find the ones who’d rather build LLMs for accounting. You’ll just pay them a lot more than you would have 24 months ago.
12. Data center NIMBYism is real, and the answer is a package, not a pitch deck
Rep. Ro Khanna is introducing a data center bill of rights giving local communities the right to say no.
Jason’s read: this podcast is entitled enough as it is, complaining that an Anthropic engineer only made $35M. Go out to the panhandle where nobody makes $50 grand. The Terrafab site has already created 3,000 jobs at 10% capacity. Those are real jobs with real money lasting years. We have 50 states and thousands of counties, and enough of them with water and power will want this business. The people closest to it think it works itself out.
Rory’s read: the reporting from affected areas isn’t “AI is awful.” It’s “I don’t know what I’m getting here and it’s all opaque.” If you want local support, you need no increase in residential electricity costs and probably a direct distribution to residents. Jobs plus a 25% power bill increase doesn’t get a yes.
13. Terrafab is the all-in bet, and Intel just raised equity for the first time since 1979
$16.8B first installment, between 2,000 and 3,000 jobs, and Intel is part of the consortium.
Jason’s read: this isn’t ambition for its own sake. It’s a decade of supply chain limitation staring him in the face. You can’t get RAM, you can’t get chips, you can’t get TSMC on the phone because Jensen is in the lobby. I don’t think we’ve ever looked forward and said “for ten years I cannot get what I need at a tolerable price.”
Rory’s read: he has a real track record on vertical integration. Launch capability folded into Starlink worked beautifully. But this is the all-in bet, and if AI spend slows at all, the all-in bet slows the fastest.
Intel is the other half of that picture. It funded itself from cash flow through the 80s, 90s, 2000s and 2010s. Now it’s back in the equity markets.
14. Revolut’s reported package is a control ask, and the participation rate is very high
Reported terms take Nik Storonsky toward 30% at $200B and roughly 40% if Revolut reaches $500B.
Rory’s math: going from $200B to $500B creates $300B of value, and the incremental grant is worth about $50B. That’s a 16% participation rate, which is abnormally high. And it raises the follow-on: is $500B to $1T another 10% of the company? These packages work when they’re tied to operational metrics. Build the biggest bank in Europe operationally and you’ve earned it. Pure stock-price triggers unwound badly in 2023 and 2024, when CEOs executed well into a down market and came back asking for a redo.
Jason’s read: this isn’t 2% to 6%. It’s 40% ownership, which is a control ask. I’ve also changed my mind on founder control generally. Being public today is miserable, and giving founders more control over their life’s work is an acceptable price for keeping the IPO path viable. Otherwise everyone does what the Collisons did and stays private.
Rory’s counterpoint: even Zuckerberg, who has had absolute board control for 20 years, just said he doesn’t want personal control over the decision to release new models. That’s the first piece of uncontrol he’s done in two decades, and it’s a smart man noticing that total control means owning every problem.
15. Almost every company in the portfolio not run by a founder is going to zero in the Age of AI
Jason’s read: I’ll answer Harry’s question about what LPs should do with what I’m actually doing. Almost any investment I’ve made that isn’t run by a founder, whether it’s at $20M or $200M in ARR, is going to be a zero in this era. If the price of not having a zero is giving a founder 40%, I pay it in a heartbeat. I do not believe a Jamie Dimon lieutenant with the starched shirt and cufflinks runs Revolut successfully. The space isn’t stable enough or mature enough.
Rory’s qualifier: true at $1M, $10M, $100M, maybe a billion. Revolut is around $5B in revenue with a billion or two in profit. At some point it stops being true, and you still need a governance dynamic that protects the other shareholders.
16. Whatnot at $20B shows how much value sits outside the AI story
$545M raised at a $20B valuation, roughly $8B GMV last year going to $16B this year, on a 12% take rate.
Rory’s read: QVC and the Home Shopping Network worked. eBay worked and still carries a $40 to $50B market cap. People enjoy live selling. The minute you hear the model you know it works.
Jason’s read: there’s gold in studying what AI is not destroying, not just what it’s decimating. Commerce, restaurants, financial services. Revolut is the same category of answer: a large business built on a universal human need.
17. Atlassian beat the quarter and killed most free Loom seats
Atlassian posted its biggest jump since 2015. It also cut most of the free Loom seats, and Canva pushed a lot of features up into higher-paid editions.
Jason’s read: whenever the free base gets harvested, that’s a stress signal. No founder wants to do that. Loom isn’t Atlassian’s breakout product and it isn’t the end of the world, but collaborative free seats are how these products spread. Even the beats are hard right now. Shopify and Atlassian are leaving nothing on the table.
Rory’s confirmation: talk to a director of sales one level in at these companies during a 7% or 8% quarter and you find out they’re jamming customers on price and on overages. It isn’t sustainable.
18. The threat to HubSpot is dozens of very good new SMB competitors, not DIY CRM
Jason’s read: the dumbest AI meme of the last two years is that everyone will vibe code their own CRM. That makes no sense for 99.9% of the world. You can’t maintain it, you can’t build the integrations, and most people saying it have never used a CRM.
The real pressure comes from below. HubSpot spent five years beating Salesforce at the low end and became a CRM company. Now the low end is extraordinarily good, and the new AI-native entrants are putting up revenue growth we haven’t seen before. Monaco, Lightfield and Aurasell are all growing at rates that didn’t exist in this category. My first venture investment was Pipedrive, and it would have taken 40 years to get competitive with Salesforce. Now you walk into a board meeting and the competitive slide is full of companies that didn’t exist 24 months ago, with good agents and good models. That’s the bear case on the entire SMB layer: too many good competitors, and far more founders willing to attack niche categories than there used to be.
On the “buy the YC companies” fix: that strategy stopped working. Rippling has absorbed 30, Owner has absorbed 20, Revolut has absorbed 10. Every hot startup is running the same playbook, and a talented team will pick the hot startup over the PE-owned asset every time.
Quotable Moments
Jason Lemkin
“We built our own ad server and ad generating network that serves the B2B community. It’s all built on our agents. It never occurred to the agent to use Canva for this. It never once occurred to it.”
“Any investment I’ve made that is not run by a founder is going to be a zero in this age. Whether they’re at $20 million or $200 million, they’re all going to be zeros.”
“The era of no-code, of things humans can do without engineering resources, is slowly winding down. And if it’s in ChatGPT, I’m just worried.”
Rory O’Driscoll
“There’s going to be a lot of people paying the bill in 26 and 27 for a certain amount of hesitancy in 23 and 24.”
“Most of the time you won’t regret trimming. But on the few that you regret trimming, it turns out to be most of the value.”
“If you’re going to give someone $50 billion, you ought to spend more time thinking about what you’re getting for your $50 billion than ‘I’ll give you 50 if it’s valued at 500 in a few years.'”
Harry Stebbings
“What is the President of Uber’s single biggest fear? The disaggregation of UI, where you say ‘I want a car’ and ChatGPT routes you to Lyft or Uber based on price.”
“I’ve interviewed a thousand founders. I’ve never interviewed anyone like Nik.”
“How significant is it, honestly, to Google, losing this many high-caliber people this quickly?”
Two Different Movies, One Market
Datadog, Cloudflare and Twilio sell into the boom and mostly had to show up. Palantir, Intercom and Replit rebuilt the company, and they did it in 2023 when it was expensive and unpopular. Canva, Airtable and the rest of the no-code generation are now finding out what the deferred decisions cost.
Growth is the only defense against the existential narrative. Atlassian proved it this quarter: get the fundamentals back and the multiple follows. Canva has to prove it without the benefit of a public print, and with a base that already routes around the product.
This post is part of the ongoing 20VC x SaaStr collaboration with Harry Stebbings and Rory O’Driscoll. The Jason’s Takes companion posts Sunday.
