Canva cuts 22% of staff and reorganizes product amid AI pressure

The announcement: 1,271 people affected

Canva announced on November 3 the first major layoff in the company's 13-year history. According to Fast Company, 1,271 people, about 22% of the total headcount, will be laid off, reassigned to new roles, or transferred to the company's headquarters in Sydney, Australia.

The announcement came in a statement signed by co-founder and CEO Melanie Perkins. In it, the executive does not cite a specific business reason for the cut, but describes the restructuring as part of a plan to make the company "more agile and less siloed".

I've come to the difficult conclusion that small changes won't be enough.

Melanie Perkins, CEO and co-founder of Canva

In the same letter, Perkins talks about "refounding Canva," returning to its roots of small teams with the autonomy to decide and take responsibility for what they do.

Why now: the cost of frontier AI

Fast Company notes that the cut comes at a moment of "growing pains" and doubts about the right timing for a possible initial public offering (IPO) by the company. In recent years, Canva has bet everything on becoming an "AI platform," building its own models and acquiring AI startups to expand its product.

This move comes at a price. Using frontier models, whether proprietary or third-party ones like Anthropic's, requires heavy computing power, which drives up operating costs at the same pace the user base grows and demands more AI features. That bill, according to the report, has been becoming a significant problem for the company.

In short: Canva is caught between two forces pulling in opposite directions: growing generative AI features is expensive, and competition in this market shows no sign of slowing down.

The era of "frenemies" in design

This tension isn't unique to Canva. In May of this year, Fast Company itself described the design sector as entering an "era of frenemies," when Anthropic launched Claude Design.

The product caught companies like Adobe and Figma by surprise, both of which had maintained a years-long collaboration with Anthropic to integrate Claude into their own products and didn't expect the company to launch a direct design competitor. Canva, for its part, revealed to the publication that it had helped develop Claude Design, even while having its own collaboration agreement with Anthropic.

This arrangement sums up the dilemma facing those who work with frontier models in design:

• Advantage: access to sophisticated models, faster feature releases than would be possible building everything in-house.
• Disadvantage: dependence on expensive infrastructure, which eats into margins as usage scales.
• Extra risk: today's partner can become tomorrow's competitor, as the Claude Design case showed for Adobe and Figma.

What changes in the product structure

Beyond the layoffs, the restructuring announced by Perkins directly changes how Canva organizes the people who design its product. The company will merge its product, growth, and platform teams into a single unit and reduce the number of internal "sub-specialties."

In practice, this means fewer isolated teams running specific pieces of the product and more small squads with broad mandates, from discovery to launch. It's a reversal of the deep-specialization model that mature SaaS companies tend to adopt as they grow, and a sign that Canva believes this model isn't delivering enough speed to compete in the AI race.

What this means for people building products in Brazil

Canva, Figma, and Adobe are among the most widely used design tools by product teams in Brazil, whether for prototyping, creating marketing assets, or generating images with AI. What happens inside these companies' internal structure directly affects the pace and direction of the features that reach Brazilian users.

Two points deserve attention from anyone leading product or design here. First, a massive bet on generative AI is no guarantee of margin: Canva's case shows that growing AI features can mean growing infrastructure costs at the same rate, something any team considering building on top of third-party models should factor in from the start.

Second, the decision to reduce sub-specialties and merge product, growth, and platform teams is an explicit bet that small, multidisciplinary squads respond faster than highly specialized structures, a discussion that also runs through product squads here as Brazilian tech companies mature their own AI teams.

What remains open, both in Perkins's letter and in Fast Company's coverage, is whether the reorganization solves the cost problem or merely reduces the payroll while the computational pressure of AI keeps rising. Canva has not publicly detailed financial figures or an IPO timeline tied to the cut.