Canva: how two Perth founders built one of the world's most valuable private software companies

There is a particular kind of Australian business story that gets told badly. Either it's triumphalist — the battlers who showed the world — or it's faintly suspicious, as though success on that scale must involve some trick. Canva is neither. It is, when you sit with the actual numbers and the actual timeline, simply one of the most methodical ascents in the history of consumer software. And it started in Perth, with school yearbooks, which is the kind of detail that sounds invented.

I've been writing about Australian companies punching above their weight internationally for long enough to develop a decent scepticism reflex. Admire the achievement; interrogate the structure. With Canva, the achievement holds up.

The yearbook problem that launched everything

Melanie Perkins was a university student in Perth — studying at the University of Western Australia — when she started noticing how painful it was to teach people to use professional design software. Adobe's tools were powerful and opaque. Students spent more time learning the interface than actually designing anything. Her observation was specific: most people only needed to do a small slice of what professional design software offered, but that small slice was buried under years of accumulated complexity.

Her first attempt at solving this was Fusion Books, a school yearbook design platform she launched with Cliff Obrecht, her then-boyfriend and now husband, around 2007. The premise was simple: give schools a browser-based tool to design and print their own yearbooks, without needing a design degree or a licensing budget. It worked. Fusion Books grew into one of the larger yearbook companies in Australia. But Perkins was already thinking past it.

The yearbook business was, in her framing, proof of concept. If non-designers could produce publication-quality work through a constrained, well-designed interface, the same logic applied everywhere. The pitch that became Canva was essentially: build Fusion Books, but for every kind of visual content, for every kind of person, at internet scale.

The fundraising years, and why Silicon Valley took time to pay attention

Getting there took longer than the polished origin story usually admits. Perkins and Obrecht spent roughly three years pitching investors before securing meaningful backing. By various accounts, Perkins made over a hundred pitches before the company was properly capitalised. Silicon Valley, at the time, had a geography problem: Perth was not a name that appeared on the mental maps of American venture capitalists, and the consumer design tool space already had entrenched players with deep distribution.

The turning point came through a connection to Bill Tai, a venture capitalist who introduced Perkins to a network that eventually included Lars Rasmussen, one of the co-founders of Google Maps. Rasmussen became an early adviser. The backing that followed — including from Blackbird Ventures, the Australian VC firm that has since become one of the country's most important early-stage investors — allowed Canva to formally launch in 2013.

That three-year gap between the idea and the launch is worth sitting with. It is easy, retrospectively, to flatten the timeline and make the ascent look smooth. It wasn't. The company was built slowly and then, once the product found its audience, very fast indeed.

The product logic: why Canva worked when others hadn't

Canva's core insight was not original in isolation. Browser-based design tools existed. Templates existed. The difference was the execution of a specific principle: the tool should constrain choices in ways that make good design almost automatic, while still feeling like genuine creative freedom to the user.

This is harder than it sounds. Software that is too constrained feels like a toy. Software that is too open recreates the problem Perkins identified in the first place. Canva threaded this needle well enough that its early users — social media managers, small business owners, teachers, nonprofit comms officers — became evangelical about it. The freemium model amplified this. You could do a genuinely useful amount in Canva for free, which meant adoption spread through organisations organically before anyone signed a procurement contract.

The product also benefited from timing. The explosion of social media platforms in the early 2010s created an enormous, largely unmet demand for visual content from people with no design background and no budget for designers. Canva was, in a sense, the infrastructure for that demand. Every platform that required a custom image size — Facebook headers, Instagram posts, LinkedIn banners — was effectively a distribution channel.

The valuation question

By 2021, Canva had raised funding at a valuation of around US$40 billion, which placed it among the most valuable private technology companies in the world and, at that point, the most valuable technology startup in Australian history. The numbers reported in subsequent rounds were similarly striking, though by late 2022 and into 2023, as private market valuations across the technology sector broadly corrected, there was public reporting suggesting internal or secondary-market valuations had pulled back from that peak.

Honestly, the valuation story is the least interesting part of what Canva built. Venture-era technology valuations are, as the last few years demonstrated with some force, not always a reliable guide to underlying business quality. What matters more is the revenue trajectory and the depth of adoption. Canva has reported passing a hundred million monthly active users, and has disclosed annual recurring revenue figures in the billions of dollars — though, as a private company, it has no obligation to file detailed accounts, and precise figures at any given moment should be treated accordingly.

What those numbers do suggest is a business with genuine scale and genuine retention. That is unusual. Many consumer software products acquire users easily and lose them just as fast. Canva's user data, as far as it has been made public, points to a product people return to regularly. That stickiness is the real moat.

The enterprise pivot that most people missed

For its first several years, Canva was understood primarily as a tool for individuals and small businesses. The enterprise play — selling large organisations on a platform that brings design governance, brand consistency and collaboration tools under one roof — was a quieter story, but it has become increasingly central to the company's commercial logic.

Large organisations have a genuine and expensive problem with brand consistency. Design assets get duplicated, modified incorrectly, shared informally across teams. Canva for Teams and subsequent enterprise offerings positioned the platform as a solution to this, rather than just a design tool. The pitch to a large corporate is meaningfully different from the pitch to a freelancer, and it implies a very different revenue profile: longer contracts, higher average contract values, more defensible retention.

This is the strategic logic that makes the valuation conversation more interesting than it first appears. A consumer freemium product with a large user base is one kind of business. A consumer freemium product that converts a meaningful portion of those users into enterprise relationships is a different, structurally better kind. Whether Canva has fully executed that transition is a question the company's private status makes genuinely hard to answer from the outside.

Perth, and what the geography actually meant

It would be easy to frame the Perth origin as colour — local-girl-makes-good texture. I think it was more substantive than that. Perth's distance from the Australian east coast, let alone from San Francisco, meant Perkins and Obrecht built without a safety net of warm introductions and casual coffee meetings with potential investors. The fundraising difficulty was partly a function of geography. So was the discipline it produced.

Companies that bootstrap through lean early years because capital is hard to access sometimes build better cost habits than those that raise easily. That's a generalisation, and I'd be careful about overstating it — Canva raised substantial capital once it found its footing, and it spent it. But the early years of resourcefulness left a mark on the culture that Perkins has spoken about in various interviews over the years.

There is a comparison worth making here to other Australian companies that built slow, quiet global positions — the kind of story the Outliers & Global Empires section keeps returning to. Cochlear is the example that always comes to mind: decades of patient accumulation, a product most people outside the medical sector barely knew existed, an extraordinary global market position built on a narrow but deep specialisation. Canva is a different shape — consumer-facing, fast-growing, visible — but the underlying patience of the early years rhymes. The competitive moat Cochlear built took thirty years to become undeniable; Canva's took closer to a decade, but neither happened quickly by the standards of the stories that get told about overnight success.

What comes next, and what could still go wrong

Canva has moved aggressively into artificial intelligence-assisted design features, which is both the obvious thing to do and a genuinely uncertain bet. AI image generation and AI-assisted layout tools threaten to commoditise exactly the kind of template-based design work that built Canva's user base. The company's response has been to integrate those tools into the platform rather than compete against them — to be the interface through which users access AI capabilities, rather than the product those capabilities displace.

Whether that strategy holds depends on questions that nobody outside the company, and possibly nobody inside it, can fully answer yet. The AI design space is moving fast, the competitive set has expanded significantly, and the enterprise sales cycle is long enough that the next few years will be telling.

The Perkins and Obrecht story is, so far, genuinely remarkable — not because of the valuation peak, but because they identified a real problem, built a product that solved it well enough to retain a hundred million users, and did it from Perth, over years of rejection, without the geography and the difficulty becoming an excuse. That is harder than it looks.

The numbers, for now, back them up. I'll keep watching.

— Priya Naidu, Outliers & Global Empires