Canva guided 2026 growth down to 20%. The worse detail is who never got asked.

Canva has guided 2026 growth down from 30% to 20%, according to SaaStr's Jason, who spent the week arguing that the number is the least interesting part of the story ^[1]. The mechanism he describes is that SaaStr stopped paying for Canva and for Notion without either product losing a comparison, because the need simply went away ^[2].

That is the part worth sitting with. SaaStr built its own ad server and creative generation network on top of its agents, and by his account the agent never once reached for Canva, a product he had paid for and liked ^[3]. His conclusion is the operationally useful line: you can win back a customer who churned to a competitor, but there is nothing to win back when you were never in the running ^[4]. The suggested diagnostic is cheap enough that there is no excuse for not running it, which is to hand an agent the job your product does, give it no instruction to use you, and watch what it picks up ^[5].

The arithmetic is worse than a ten-point haircut sounds. Cutting guidance from 30% to 20% removes a third of the guided growth rate ^[11]. If the lower rate simply held, three years of compounding lands revenue about 21% below where the old path would have put it, which is the gap that valuation marks have to absorb ^[12].

His structural claim is that a whole product generation is exposed. Airtable was a no-code database dressed as a spreadsheet, Notion is one dressed as a word processor, and Canva was a no-code way to design that removed the need for a designer or any HTML; the category of "what humans can do without engineering resources" is winding down because models now do that natively ^[6]. Being the workaround for a missing skill was a good fifteen-year business and is now a shrinking one ^[6].

Two honest caveats. First, this is one operator's read: the agent-substitution mechanism is presented as his own observation and inference, not as anything Canva disclosed about why guidance moved ^[13]. Second, the timing is uneven. He cites Gartner data suggesting fewer than 10% of enterprises have successfully deployed an agentic application, and says the hot enterprise AI companies are still serving early adopters, while the prosumer side has no lag at all because buyers are already ChatGPT-fluent ^[7]. His framing of that gap is that selling to the enterprise buys you a few quarters of cover, not a few years ^[8].

The narrative risk compounds the revenue risk. He points to the vibe-code-your-own-CRM meme as something that made little practical sense, took hold anyway, got picked up by shorts, and is now priced in, and says the equivalent hazard for Canva is "you can make a poster in ChatGPT for free" ^[9]. On marks, he is blunt: after Airtable and Canva he is writing valuations down harder, and "we raised at $42B so we're carrying $42B" no longer works as a defence ^[10].

What to watch: whether Canva or any comparable self-serve business starts reporting the composition of its churn rather than just the rate, since need-disappearance and competitive loss look identical in a net-retention number. Watch also whether the enterprise lag Gartner describes narrows, because that is the only cover the same category of product currently has ^[7].