The Salon on Crown Street
On a September morning in 2022, Melanie Perkins stood outside a hair salon in Surry Hills, Sydney, peering through the double doors at women in highlight foils and black capes. A half-mannequin in a feather boa sat on top of a cabinet near the back wall. She pointed to it. That's where the computers used to be. A decade earlier, this storefront — Sterling, it's called now — had been the cramped, chaotic headquarters of Fusion Books, the yearbook publishing company Perkins and her then-boyfriend Cliff Obrecht had bootstrapped out of what she later described as "a rummage pile of everything." They'd cleaned it out, installed a bank of monitors, and spent nearly every waking hour there, printing and shipping yearbooks to Australian high schools while nurturing a vision so disproportionate to their surroundings that it would have seemed delusional to anyone who happened to wander in.
The vision was this: build the world's most valuable company.
Not "build a successful Australian startup." Not "create a nice design tool." The world's most valuable company. Perkins was twenty-five years old, without a technical background, without Silicon Valley connections, without a completed university degree, operating from the back of what would become a hair salon in a country whose tech ecosystem barely registered on the global map. She had been rejected by more than one hundred venture capitalists. She would be rejected by many more. And the company she would eventually build — Canva, a visual communications platform now valued at $32 billion, used by more than 220 million people monthly across 190 countries, generating $3 billion in annual revenue — would not launch for another year.
But the dream was already fully formed. It had been since she was nineteen.
By the Numbers
The Canva Empire
$32BCurrent valuation (2025)
$3BAnnual revenue (2024)
220M+Monthly active users
30B+Designs created on the platform
190Countries with active users
100+Languages supported
8Consecutive years of profitability
Column B
There is a framework Perkins uses to explain how she thinks, and it is worth understanding because it illuminates nearly every decision she has ever made. She calls it "column B" thinking.
Column A, in her telling, is how most planning works: you look at the bricks you have — your resources, your experience, your existing constraints — and you try to stack them into something reasonable. Column B is the opposite. You start with the future you want to inhabit, the impossible thing, the outcome that sounds absurd when you say it aloud, and then you work backward, building a ladder to a destination that doesn't yet exist. "Most planning is often done by looking at the bricks and trying to stack them," she told Lenny Rachitsky in late 2025. "The column B thinking is thinking about what is that magical, wonderful future that you then want to invest years and decades of your life actually building."
This distinction matters because Canva's entire history is a story of column B thinking applied with a patience and granularity that most founders — most people — cannot sustain. The gap between the dream and the first step was not merely large. It was humiliating. "It feels a little embarrassing to be like, I want a future... and then to take such a microscopic step," Perkins admitted. The microscopic step was yearbooks. Australian high school yearbooks. The future was everything.
Born on May 13, 1987, in Perth — a city closer to Jakarta than to Sydney, separated from Australia's eastern seaboard by a vast stretch of desert and approximately 2,044 miles of geographic irrelevance — Perkins grew up in a household that was multicultural in the specific way that Perth sometimes is: her mother an Australian-born teacher, her father a Malaysian engineer of Filipino and Sri Lankan heritage. The family was supportive, middle-class, unremarkable in its circumstances but apparently extraordinary in its encouragement. Perkins woke at 4:30 a.m. as a child for figure skating training, a discipline that instilled the particular kind of goal-oriented stubbornness she would later deploy against venture capitalists. At fourteen, she started selling handmade scarves at Perth markets. "I never forgot the freedom and excitement of being able to build a business," she later said. It was, by any measure, a modest enterprise. But the psychological pattern was set.
She enrolled at the University of Western Australia to study a combined major in communications, psychology, marketing, and management. She would not graduate. What she would do, beginning around 2006, was tutor fellow students in graphic design programs — Adobe Photoshop, InDesign, the full suite of tools that the professional creative class took for granted — and discover that the experience of learning these programs was, for ordinary people, something close to torture.
"People would have to spend an entire semester learning where the buttons were, and that seemed completely ridiculous."
It was a simple observation. Almost banal. But Perkins held it in her mind and refused to let it dissolve into the background noise of accepted frustration. Facebook was taking off. People could jump in and use that. Why did design tools take years of training? The gap between the ease of consumer internet applications and the brutality of professional design software struck her not as an inconvenience but as an injustice — a failure of imagination by the companies that made these tools. Microsoft. Adobe. They had built for professionals, for people who already knew how to design. Nobody was building for everyone else.
I thought that in the future it was all going to be online and collaborative and much, much simpler than these really hard tools.
She was nineteen. She had no coding skills, no capital, no relevant industry experience. By her own account, she lacked "any relevant experience." But she had column B. And column B said: the entire design ecosystem, integrated into one platform, accessible to the whole world.
The bricks she had were her boyfriend, her mother's living room, and an AUD 50,000 loan from friends and family.
The Yearbook Gambit
The decision to start with yearbooks was not a compromise. It was strategy — though it looked, at the time, like a compromise to everyone who encountered it.
Perkins understood, with an intuition that belied her age, that she could not attack the full design market from Perth with no technology, no team, and no money. But she could prove the thesis. If she could build a simple, web-based tool that allowed people with no design training to create something beautiful and functional — and if she could demonstrate that people would pay for that — then the broader vision became defensible. Yearbooks were the proving ground because her mother, a teacher, had spent hundreds of hours wrestling with desktop publishing software to produce her school's yearbook each year. The pain was real, the market was tangible, and the stakes were low enough that failure wouldn't be fatal.
In 2007, Perkins and Obrecht launched Fusion Books. Cliff Obrecht — who had been studying at the same university, who would become her co-founder, business partner, husband, and the person with whom she would eventually sign the Giving Pledge — was, in those early days, simply the boyfriend who said yes. He is a quieter presence in the Canva narrative, less visible than Perkins but no less consequential: the operational counterweight to her visionary ambition, the person who would eventually serve as COO and manage the unglamorous machinery of scaling a company while Perkins sold the dream. They contracted a Perth software development firm called InDepth, run by a man named Greg Mitchell, to build their first platform.
Perkins's mother's living room became the office. The physical yearbooks were printed there. Perkins and Obrecht, wearing what they later described as "dorky suits," went around Perth selling the service to schools. Their first customer was a French school in Sydney. Fifteen schools in the first year. Thirty the second. Eighty the third. Within five years, Fusion Books was the largest yearbook company in Australia and had expanded into New Zealand and France. It remains active today.
But for Perkins, Fusion Books was always and only the first rung on the ladder. The yearbook business validated the core premise — that non-designers could create professional-quality work with the right tool — but it was, by design, a stepping stone. The "crazy, big dream" was Canva. And for Canva, she needed something Fusion Books could not provide: capital.
One Hundred Doors
The mythology of the startup pivot — the chance encounter, the lucky break, the single meeting that changes everything — is, in Perkins's case, both true and misleading. True because there was, in fact, a pivotal encounter. Misleading because the encounter only mattered because of what preceded it: years of failure, refinement, and the kind of relentless persistence that borders on the pathological.
In 2010, at a conference in Perth, Perkins met Bill Tai. Tai — a legendary Silicon Valley venture capitalist, prolific early-stage investor, and passionate kitesurfer whose investment retreats doubled as extreme sports vacations — was the first serious person from the world of startup capital to hear Perkins pitch. What happened next has been retold so often it has the quality of parable. Tai was on his phone during the pitch. Perkins thought he wasn't listening, that she'd blown it, that her brief window of access to Silicon Valley money had closed before it opened. But Tai wasn't ignoring her. He was texting her pitch to his contacts. Within hours, he was connecting her to people who could help.
Among them was Lars Rasmussen, the Danish-Australian technologist who had co-founded Google Maps — a man who understood, better than almost anyone, how to take a complex technological capability and make it accessible to ordinary users. Rasmussen became an advisor, then an investor, and his involvement lent Perkins the credibility she desperately needed.
But credibility was not capital. Over the next three years, Perkins pitched more than one hundred venture capitalists. The rejections were comprehensive and varied: she was too young, too Australian, too female, too non-technical, too far from Sand Hill Road, too ambitious, not ambitious enough in the right ways. She and Obrecht traveled repeatedly to Silicon Valley, two unknown founders from Perth — a place most American VCs couldn't locate on a map — pitching a design tool that would compete with Adobe and Microsoft. The idea sounded either naive or insane, depending on the investor's mood.
"When you don't have any connections, you don't have any network, you just kind of have to wedge your foot in the door and wiggle it all the way through," Perkins later told CNBC.
Tai, the kitesurfing VC, had one additional demand: if Perkins wanted to network at MaiTai, his exclusive investor retreat, she and Obrecht would need to learn to kitesurf. They did. It wasn't their thing. Perkins has been candid about this. But she did it anyway, because access is access, and when you're a nineteen-to-twenty-two-year-old from Perth trying to raise money in Silicon Valley, you take whatever path presents itself, even if that path is strapped to a board in high winds.
When you don't have any connections, you don't have any network, you just kind of have to wedge your foot in the door and wiggle it all the way through.
What the rejections did — and this is the part of the story that matters more than the eventual yes — was sharpen the pitch. "Every time we were rejected or had tricky questions we'd improve our pitch decks," Perkins has said. Each no was a data point. Each skeptical question revealed a gap in the argument. The pitch deck evolved, version by version, into something precise and devastating: a visual argument for why the entire design industry was broken and why a web-based, freemium platform could fix it. "Their feedback made us stronger and made our pitch deck stronger," she told Lenny Rachitsky. "It was like a really intense boot camp."
The breakthrough came in 2012. The missing piece had always been technology — Perkins and Obrecht could sell, could envision, could pitch, but they could not build the product themselves. They needed a technical co-founder. They found Cameron Adams.
The Third Founder
Cameron Adams had spent years at Google, where he'd worked as a designer and developer — a rare hybrid who understood both the aesthetics and the architecture of software. He was Australian, based in Melbourne, and he was exactly the kind of person Perkins needed: someone who could translate her vision into code, who understood the technical complexity that lay beneath the deceptive simplicity she was proposing.
Adams didn't join immediately. He was skeptical. The idea was enormous. The team was tiny. The resources were thin. But Perkins, as she had done with investors and would continue to do for the rest of her career, simply would not stop. She pitched Adams the way she pitched VCs — with the column B future, the impossible dream, the conviction that bordered on unreasonable. Eventually, he said yes. He joined as co-founder and Chief Product Officer, and his arrival transformed Canva from a vision into a viable technology company.
With Adams on board, the team secured a $1.5 million seed round in early 2012, with contributions from Tai, Rasmussen, and others. An additional $1.6 million came from investors, supplemented by $1.4 million in matching funds from the Australian government. The total — approximately $3 million — was enough to begin building the product. A technical developer named Dave Hearnden joined the team. The company was incorporated in Sydney on January 1, 2013. They chose the name after a French engineer pointed out that "canvas" is pronounced "canva" in French — simpler, cleaner, more universal.
Canva launched later that year with a premise so straightforward it was almost disarming: drag and drop. Templates. A vast library of elements — photos, icons, fonts, illustrations — organized so that someone with no design training could produce professional-quality work in minutes. The thesis was not that Canva would replace Photoshop for professional designers. It was that Canva would serve the billions of people who were not professional designers but who increasingly needed to produce visual content — for social media, for school, for work, for life.
The timing was exact. The rise of social media had created an unprecedented demand for visual content. Every small business, every teacher, every student, every nonprofit, every person with an Instagram account suddenly needed to be a designer. And the existing tools — Adobe's Creative Cloud, Microsoft's PowerPoint — were either too expensive, too complex, or both.
Canva was neither.
The Geometry of Growth
The first million users arrived by April 2014, barely a year after launch. The speed was startling, even to Perkins. The product had struck something — not just a market opportunity but a latent frustration that millions of people hadn't been able to articulate until someone solved it. Word of mouth was the primary engine. Users made things, shared them, and other people asked: What did you make that with?
The freemium model was the second engine. Canva's basic version was free — genuinely, substantively free, not the kind of "free" that makes you feel poor. You could design a social media graphic, a poster, a presentation, a business card, without paying anything. The paid tiers — Canva Pro, Canva for Teams, eventually Canva Enterprise — offered more templates, more storage, more brand management tools, but the free version was good enough that people used it, loved it, and told their friends. This was not accidental. It was the deliberate architecture of a company that understood its core bet: if you make design accessible to everyone, a percentage of everyone will pay for more.
By 2017, Canva was profitable on a free-cash-flow basis. It has remained so every year since — eight consecutive years of profitability, a feat almost unheard of among venture-backed technology companies of its scale. The company had $700 million in cash on hand. When VC money was flush, Canva earned the backing of Sequoia Capital, Bessemer Venture Partners, and Founders Fund. When other startups were laying off workers, Canva was hiring. When valuations collapsed across the tech sector, Canva's revenue kept growing.
The numbers tell one story. The product decisions tell another.
Key milestones from founding to current scale
2007Fusion Books launched from Perkins's mother's living room in Perth
2010Chance meeting with investor Bill Tai at a Perth conference
2012Cameron Adams joins as technical co-founder; $3M seed round closed
2013Canva launches publicly on January 1 from Sydney
2014Reaches 1 million users by April
2017Achieves free-cash-flow profitability for the first time
2018Reaches unicorn status at $1B+ valuation
2019First enterprise products launched
The Two-Year Silence
There is a period in Canva's history that Perkins rarely discusses in detail, though she has acknowledged it exists. After the early growth, the company's technical infrastructure began to buckle. The original codebase — built fast, by a small team, to prove a concept — could not support the product Perkins wanted to build next. The decision was made to rewrite the entire codebase from scratch.
For two years, Canva could not ship a single new feature.
Consider what this means for a consumer technology company in a market moving at the speed of social media. Two years without new capabilities, without the kind of product launches that generate press coverage and sustain user excitement. Two years of watching competitors move while you are, externally, standing still. Internally, the team was working furiously — rebuilding the foundation of the product, making choices that would determine whether Canva could scale from millions of users to hundreds of millions. But externally, there was silence.
Perkins has described this as one of the most difficult periods in Canva's history. The discipline required — to invest in infrastructure rather than features, to accept short-term stagnation for long-term capability — is the kind of decision that destroys most startups. Growth-stage companies are addicted to momentum. Investors demand it. Users expect it. The market punishes anything that looks like stalling. Perkins made the call anyway.
When Canva emerged from the rewrite, it was a different product — faster, more stable, capable of supporting the ambitious expansion that would follow. The visual worksuite. Enterprise tools. AI integration. International expansion into more than 100 languages. None of it would have been possible on the old infrastructure. The two-year silence was, in retrospect, the most important strategic decision Perkins made after founding the company. It was also the most invisible.
The Worksuite Wars
"We had this in our pitch deck back in 2011," Perkins said in September 2022, standing onstage at Canva's first Create event. "This was the final pillar."
The pitch deck she was referencing — the one she'd refined through more than a hundred rejections — had always contained a slide showing Canva's ultimate ambition: not just a design tool, but a complete visual communications platform that would compete with Google Docs, Microsoft PowerPoint, and Adobe's entire Creative Cloud. For eleven years, that slide had been aspirational. Now it was a product launch.
The visual worksuite was Canva's declaration of war against the productivity giants. Canva Docs. Canva Presentations. Canva Whiteboards. Canva Websites. Tools that looked and felt like Canva — simple, template-driven, visually rich — but occupied the functional territory of products that billions of people used every day. The bet was audacious: that in a world where communication was becoming increasingly visual, the distinction between "design software" and "productivity software" was artificial and would eventually collapse. Why should a presentation look different depending on whether you made it in Canva or PowerPoint? Why shouldn't a document be beautiful by default?
The enterprise push that followed was methodical. Canva had always grown bottom-up — individuals discovering the tool, using it, loving it, and then bringing it into their workplaces. By 2019, the company had launched its first enterprise products. By 2024, it was actively courting CIOs and Fortune 500 companies with features like brand template management, auto-provisioning with SCIM, and direct publishing to platforms like Meta and Amazon. Enterprise customers included Workday, Expedia, and FedEx. At Workday alone, Perkins claimed Canva had saved 33,000 hours.
Bob Iger — the former Disney CEO, Canva investor, and a man who knows something about the power of visual storytelling — appeared onstage at Canva Create in 2024 in Los Angeles. "I think she's very capable of running a company that is much larger and more complex than the company she runs," Iger told Fortune. Coming from a man who had run the
Walt Disney Company, this was not a casual compliment.
But the enterprise transition also revealed the tension at the heart of Canva's identity. The company had been built for individuals — for the teacher making a classroom poster, the small business owner designing a flyer, the student building a résumé. Enterprise clients want control, compliance, security, the kinds of features that make software feel less magical and more institutional. "You can be as locked down with a brand template as you want, or as unlocked as you want," Perkins told Fortune. "There's the protection side and the unlocking side." The question — still unresolved — is whether one platform can serve both masters without losing the simplicity that made it beloved in the first place.
Magic, Not AI
When generative AI erupted into public consciousness in late 2022 and early 2023, it posed an existential question for every design company: if anyone can generate an image with a text prompt, why do they need a design platform?
Perkins's answer was swift, confident, and characteristically reframed. She was not worried. "There wasn't a moment of concern that this might be an existential threat," she told WIRED. "No, not at all."
This was either genuine conviction or masterful performance. Probably both. Canva had been investing in AI for years before the generative explosion — its Background Remover tool, powered by the acquisition of AI company Kaleido in 2021, was one of the platform's most popular features. When the generative wave hit, Canva was already positioned to ride it. In March 2023, the company launched its first suite of AI-powered tools. In October, it introduced Magic Studio — a comprehensive AI toolkit including text-to-image generation, AI-powered layout suggestions, and automated content creation. In 2024, it acquired Leonardo.ai, an Australian text-to-image generator, deepening its AI capabilities further.
The branding choice was deliberate. Canva did not call its tools "AI." It called them "magic." "We believe what our customers really want is magic, versus the tech ecosystem that talks a lot about..." Perkins trailed off in a Fortune interview, declining to finish the sentence. The implication was clear: AI is a buzzword for technologists. Magic is what users actually experience. The distinction was not merely semantic — it reflected Canva's core philosophy that technology should be invisible, that the user should never have to understand how something works, only that it does.
By 2025, Canva had launched Canva Code — an AI-powered coding platform that could prototype apps from text prompts — and Canva Sheets, a spreadsheet tool competing with Google Sheets but infused with Canva's visual DNA. Perkins was using Code herself to build internal prototypes. "I had an idea on the way home the other day, and I was able to turn it into a prototype and then get it user-tested within an hour," she told Fortune. "Which is just completely crazy."
There was, however, a darker note. Just before the Canva Code launch, the Australian Financial Review reported that Canva had laid off ten of twelve technical writers. "Every single person is going to have to be upskilling into this new era," Perkins acknowledged. The tension between democratizing creation and eliminating the humans who previously did the creating was not one she could resolve with branding. It is the tension of the age.
The Two-Step Plan
In September 2021, Canva raised $200 million at a valuation of $40 billion, making it one of the most valuable private software companies on Earth and the most valuable female-founded and female-led company in history. Perkins and Obrecht's combined 30% stake was worth approximately $12 billion at that valuation. They were, by any metric, spectacularly wealthy.
Perkins used the occasion to announce what she called the Two-Step Plan.
Step One: Build one of the world's most valuable companies.
Step Two: Do the most good you can.
The steps, she argued, were not sequential but symbiotic — each fueling the other. She and Obrecht pledged the majority of their Canva shares to the Canva Foundation, a nonprofit focused on alleviating global poverty. They signed the Giving Pledge. Approximately 30% of Canva's total value was dedicated to philanthropic purposes.
"It was a really easy decision," Perkins told TIME. "How can you do the most good you can do with billions of dollars?"
The Canva Foundation has since partnered with GiveDirectly, the direct cash transfer organization, committing more than $150 million to provide unconditional cash to families in extreme poverty in Malawi. More than 85,000 people received transfers from the initial $50 million commitment; a subsequent $100 million pledge in 2025 will reach another 185,000. "Each story is different," Perkins wrote on LinkedIn, "but all prove a simple truth: when people are trusted to decide what they need most, the impact is extraordinary."
The company also donated more than $1 billion worth of product through Canva for Education and Canva for Nonprofits — free subscriptions to schools and charitable organizations worldwide.
Is this genuine altruism, strategic philanthropy, or a sophisticated form of brand-building? The honest answer is probably all three, inseparable and mutually reinforcing in ways that resist clean categorization. Perkins has been transparent about the interconnection: the company's mission to "empower the world to design" and its commitment to doing good are, in her framework, the same project. Canva for Education creates future Canva users. The Giving Pledge generates goodwill that attracts talent and customers who care about purpose. The Two-Step Plan is a business strategy dressed as a moral commitment — or a moral commitment that happens to be excellent business strategy. The distinction may not matter.
There's always been forks in the road where we could do what every other company is doing, or just do what feels natural. Every time we do what feels natural, it always ends up being the thing that we build on.
The Architecture of Privacy
Perkins is, for a person of her visibility, remarkably opaque. She and Obrecht welcomed their first child in 2022 and have declined to discuss the matter publicly. When talking to journalists about any topic — not just her personal life — she monitors her own words with a precision that suggests both discipline and anxiety. "Sorry, that's not a good quote for you," she sometimes tells reporters, mid-answer.
This self-editing instinct is unusual among founders of her generation. The dominant Silicon Valley archetype is the founder-as-personality: visible, outspoken, cultivating a personal brand that becomes inseparable from the company's. Perkins has done the opposite. She is present but restrained, visible but contained. She speaks at conferences and gives interviews, but each appearance feels carefully calibrated — the talking points polished, the vulnerabilities revealed only in doses that serve the narrative. She will tell you about being rejected by a hundred investors. She will not tell you what she thought about at 2 a.m. during the two-year code rewrite when nothing was shipping.
The privacy is partly cultural. Australia is not Silicon Valley. The ethos of Perth — quiet, understated, suspicious of self-promotion — runs deep in Perkins despite a decade of global success. She and Obrecht married privately. They live, by all accounts, modestly relative to their wealth. When a 2021 Forbes feature noted that they had not made any flashy purchases, Perkins deflected: "I always like to have enough money in the bank that if the lights turned off tomorrow and everything disappeared, we've got enough capital to keep us together for a long time."
Being headquartered in Sydney, thousands of miles from Sand Hill Road, has been — depending on the moment — Canva's greatest liability and its most powerful asset. The liability is obvious: distance from capital, from talent pools, from the self-reinforcing ecosystem of Silicon Valley dinners and deal flow. The asset is subtler. Operating from Sydney gave Perkins the freedom to ignore the fads, the herd mentality, the whatever-is-fashionable-this-quarter thinking that drives so much of the Valley's strategic conversation. "Being private enables us to continue to flesh out exactly who we are and what we're doing," she told Fortune in 2024, "and make that really apparent to the world."
The world has been waiting for Canva to go public for years. Perkins doesn't seem to be in any rush.
The Crazy Big Goal
Canva's internal culture is organized around what Perkins calls "crazy big goals" — annual priorities so ambitious they sound, by her own admission, slightly absurd. One year the goal is to rewrite the codebase. Another year it's to launch in every language. Another it's to build AI. The goals are not aspirational in the polite corporate sense — they are the kind of targets that would make a reasonable person's palms sweat.
The company operates in what Perkins describes as "small startups" — teams of three to six people, each with their own goals and plans, designed to move as fast as a startup within the structure of a company that now employs more than 3,500 people. Decisions, according to several accounts, are radically decentralized. Teams ship features without management sign-off. Product choices that take competitors weeks of meetings happen in a single conversation. The philosophy is that control creates drag; clarity eliminates it.
"If a team could answer three specific criteria, they moved forward," as one account of Canva's internal process described it. "No manager loop required."
To celebrate achieving crazy big goals, Canva has thrown tomatoes in the style of La Tomatina. They have smashed plates. The celebrations are as deliberate as the goals themselves — ritualistic, communal, designed to mark the distance traveled. "If you're just trying to plod to the top of the mountain always and you never take a moment to pat yourself on the back," Perkins has said, "it can feel a little arduous."
The goals themselves — achieve them they do, though the timelines shift. "The time frame that we achieve them on has not always been reliable," Perkins admitted. "But over five years, they've really been coming into reality." This is column B thinking at organizational scale: the destination is fixed, the path is flexible, and the patience required is extraordinary.
An Image That Resolves
In late 2025, Perkins sat for a BBC interview and was asked about the current moment — about AI, about the future of the internet, about what comes next. She was thirty-eight years old. She ran a company worth $32 billion. She had been doing this for nearly two decades, since she was a teenager selling scarves at Perth markets, since she was a university dropout teaching Photoshop in a multimedia lab, since she was a young woman on a kiteboard in high winds trying to impress a venture capitalist who wasn't even watching her ride.
"We have a crazy big dream," she had written on LinkedIn earlier that year, "that by the end of our lifetimes, everyone on this planet will have their basic human needs met. It's terribly sad that this is still considered a crazy big dream."
The interviewer asked about Canva's IPO. Perkins said it was "on the horizon" but there was "no news." The world was waiting. Perkins was building.
Somewhere in Surry Hills, the mannequin in the feather boa still sits on top of the cabinet where the computers used to be.
The following principles are extracted from Melanie Perkins's two-decade journey building Canva — from yearbook company to visual communications platform to aspiring enterprise colossus. They are not conventional startup wisdom. Several of them contradict the dominant orthodoxies of Silicon Valley. That is, in many ways, the point.
Table of Contents
- 1.Start with the impossible future, not the available bricks.
- 2.Use the niche as a proving ground, not a destination.
- 3.Let rejection refine the argument.
- 4.Find the missing skill, not the missing capital.
- 5.Build the infrastructure before the features.
- 6.Make profitability non-negotiable.
- 7.Use distance as a filter.
- 8.Brand the technology, not the technology.
- 9.Decentralize decisions, centralize clarity.
- 10.Tie purpose to the business model, not the press release.
- 11.Stay private until the story is undeniable.
- 12.Set goals that make you feel small.
Principle 1
Start with the impossible future, not the available bricks.
Perkins's "column B" framework is the foundational principle of her entire career. Where most founders begin with their existing resources and capabilities — what can I build with what I have? — Perkins begins with the end state. What should the future look like? Then she works backward, accepting that the first step will be embarrassingly small relative to the destination.
This is not the same as having a vision. Every founder has a vision. Column B thinking is a method — it requires writing down the impossible future in concrete detail, decomposing it into intermediate steps, and then accepting the gap between where you are and where you're going without allowing that gap to paralyze you. Perkins went from "the world's most valuable company" to "let me help Australian teenagers make better yearbooks." The distance between those two points would have stopped most people. It didn't stop her because the yearbook was never the goal — it was the first rung on a ladder she could see even if no one else could.
Tactic: Before building anything, write a detailed description of the future your company inhabits at full maturity — not what's achievable in the next round, but the end state ten to thirty years out — then identify the smallest possible first step that validates a single element of that future.
Principle 2
Use the niche as a proving ground, not a destination.
Fusion Books was never intended to be a permanent business. It was a hypothesis test. Perkins chose yearbooks because the market was small enough to win, the pain point was real enough to validate, and the product could demonstrate — at low cost and low risk — that non-designers could produce professional-quality work with the right tool.
The critical discipline was maintaining strategic clarity about the niche's role. Many founders fall in love with their initial market. Perkins didn't. She ran Fusion Books for five years, grew it to be the largest yearbook platform in Australia, and the entire time she was pitching VCs on the bigger idea. The niche was a credential, not a cage. It proved product-market fit, generated revenue, and gave her a track record — all without requiring the resources that the full vision would demand.
Tactic: If your ultimate vision is too large or capital-intensive to attack directly, identify a narrow, achievable market that validates the core thesis, and treat success there as evidence for the broader pitch — not as your company's identity.
Principle 3
Let rejection refine the argument.
More than one hundred investors said no to Canva. Perkins's response was not to change the idea — it was to change the pitch. Each rejection generated feedback. Each skeptical question revealed a weakness in the argument. The pitch deck evolved, version by version, through what Perkins has called "a really intense boot camp."
This is a fundamentally different relationship with rejection than the one most founders have. The standard narrative is that rejection builds character — you endure it, you persist, and eventually someone says yes. Perkins treated rejection as data. She was running an iterative process on her pitch the same way a product team runs iterations on software: test, gather feedback, improve, ship again. The investors who said no were, unknowingly, her product testers.
Tactic: After every failed pitch, fundraise, or sales conversation, identify the single strongest objection, build a concrete response into your materials, and track whether that objection recurs — treating your pitch as a product with its own development cycle.
Principle 4
Find the missing skill, not the missing capital.
The inflection point for Canva was not a funding round. It was Cameron Adams. Perkins and Obrecht could pitch, sell, design, and operate — but they could not build the technology. The addition of a technical co-founder who had worked at Google, who understood both design and engineering, transformed Canva from a compelling pitch into a viable company.
Adams's role illustrates a broader principle: capital follows capability, not the reverse. Investors who had rejected Perkins for years began to take the company seriously once it had a credible technical team. The money came because the skill gap was closed, not the other way around. Perkins understood this intuitively — she spent years recruiting Adams, pitching him repeatedly until he agreed.
Tactic: Identify the single capability gap that makes your company unbackable, and devote more energy to closing that gap — through co-founder recruitment, key hires, or partnerships — than to fundraising itself.
Principle 5
Build the infrastructure before the features.
Canva's two-year codebase rewrite — during which the company shipped no new features — is one of the most counterintuitive decisions in its history. Growth-stage companies are expected to ship constantly. Perkins chose to rebuild the foundation instead, accepting short-term stagnation for long-term scalability.
This decision required enormous discipline and internal credibility. It meant telling the team that the most important work they could do was invisible to users. It meant tolerating the anxiety of watching competitors move while you stood still. And it meant betting that the product roadmap that the new infrastructure would enable — the visual worksuite, AI tools, enterprise features — would more than compensate for the lost time.
It did. The rewrite enabled everything that followed. But the lesson is not "always rewrite your codebase." The lesson is that infrastructure investments are strategic weapons, and the willingness to accept pain now for capability later is a competitive advantage that most companies cannot match.
Tactic: Periodically audit whether your current infrastructure can support your three-year product roadmap — and if the answer is no, make the investment before the limitations become crises, even if it means a visible slowdown.
Principle 6
Make profitability non-negotiable.
Canva has been profitable on a free-cash-flow basis every year since 2017 — eight consecutive years. This is extraordinary for a venture-backed technology company of its scale and growth rate. Most comparable companies treat profitability as something that happens eventually, after growth has been maximized. Perkins treated it as a constraint from the beginning.
The result is optionality. With $700 million in cash on hand and self-sustaining economics, Canva has never been forced to raise capital at unfavorable terms, never been pushed to go public before it was ready, never had to make short-term decisions to satisfy investors' exit timelines. "I always like to have enough money in the bank that if the lights turned off tomorrow," Perkins has said, "we've got enough capital to keep us together for a long time."
💰
Canva vs. the VC Playbook
How Canva's financial discipline differs from typical growth-stage startups
| Conventional Wisdom | Canva's Approach |
|---|
| Burn cash to maximize growth; profitability comes later | Profitable since 2017; growth and profitability treated as simultaneous requirements |
| IPO within 7–10 years of founding | Stayed private for 12+ years; offered secondary sales for liquidity |
| Headquarter near investors and talent (San Francisco) | Headquartered in Sydney; built local talent pipelines |
| Raise as much as possible at the highest valuation | Raised strategically; maintained large cash reserves as a buffer |
Tactic: Set a profitability target as a hard constraint alongside your growth targets — not as an aspiration — and use the discipline it creates to force sharper decisions about what to build and what to cut.
Principle 7
Use distance as a filter.
Being headquartered in Sydney, 7,500 miles from Sand Hill Road, was — for years — presented as Canva's biggest disadvantage. Perkins has reframed it as an advantage. Distance from Silicon Valley meant freedom from its herd mentality, its fads, its relentless pressure to conform to whatever the current venture narrative demanded.
When the Valley was obsessed with crypto, Canva was building design tools. When the Valley was obsessed with the metaverse, Canva was building AI integrations. When the Valley's startups were laying off workers en masse, Canva was hiring. The geographic isolation forced Perkins to develop her own strategic framework rather than borrowing one from the prevailing Silicon Valley consensus. "Being private enables us to continue to flesh out exactly who we are," she said.
This is not an argument against proximity to capital. It is an argument for strategic independence — for building a decision-making framework that is robust enough to survive without the validation of being in the room where everyone else is making decisions.
Tactic: Identify the three most common strategic assumptions in your industry and deliberately evaluate whether following them serves your company's specific situation — or merely provides the comfort of consensus.
Principle 8
Brand the technology, not the technology.
When Canva launched its AI tools, it did not call them "AI." It called them "magic." Background Remover. Magic Resize. Magic Studio. Dream Lab. The branding choice was not casual — it reflected a deep understanding of Canva's user base. Canva's 220 million monthly users are overwhelmingly not technologists. They don't care whether a feature is powered by neural networks or hamsters. They care whether it works, whether it's easy, and whether it makes them feel capable.
By branding AI as "magic," Canva accomplished two things: it made cutting-edge technology feel approachable rather than intimidating, and it avoided tying its brand to a buzzword that might lose its luster. If "AI" becomes associated with hype, disappointment, or fear — all plausible outcomes — Canva's tools will still be "magic."
Tactic: When integrating new technology into your product, name it for the experience it creates rather than the mechanism behind it — users adopt outcomes, not architectures.
Principle 9
Decentralize decisions, centralize clarity.
Canva's internal structure — "small startups" of three to six people, each empowered to ship without management approval — only works because the company's goals and values are defined with extraordinary precision. Perkins has described a system where teams can move forward if they meet three specific criteria. No escalation required. No manager loop.
This is the inverse of how most growing companies operate. As organizations scale, they typically add approval layers, review processes, and management hierarchies to maintain control. Perkins did the opposite — she stripped away control mechanisms and replaced them with clarity. If everyone understands what "good" looks like, they don't need permission to pursue it.
The result is speed. Product teams ship updates in days that would take competitors weeks. But the prerequisite is investment in alignment — in values, in goals, in shared understanding of the mission — that most companies treat as soft work and Perkins treats as infrastructure.
Tactic: Replace approval processes with decision criteria — define exactly what "good" looks like for each type of decision, publish it widely, and give teams explicit permission to act when the criteria are met.
Principle 10
Tie purpose to the business model, not the press release.
Canva's Two-Step Plan — build one of the world's most valuable companies, then do the most good possible — is not a corporate social responsibility initiative. It is the company's operating thesis. The Canva Foundation owns approximately 30% of the company's equity. Canva for Education and Canva for Nonprofits have donated more than $1 billion in product value. The partnership with GiveDirectly has transferred more than $50 million in direct cash to families in Malawi.
What makes this different from conventional corporate philanthropy is structural integration. The giving is not funded by discretionary profits — it is embedded in the ownership structure. Every subscription, every design, every dollar of revenue fuels the mission by increasing the value of the Foundation's equity stake. This alignment means that the philanthropic commitment scales automatically with the business. It also means that the purpose attracts talent, customers, and partners who share it, creating a flywheel that competitors without similar commitments cannot replicate.
Tactic: If purpose is part of your company's identity, embed it in the ownership or financial structure — equity allocation, revenue sharing, product donation — so that it scales with the business rather than depending on annual decisions about how much to give.
Principle 11
Stay private until the story is undeniable.
The world has been expecting Canva's IPO for years. Perkins has repeatedly declined to set a date. Instead, she offered a $1 billion secondary stock sale in 2024 to provide liquidity for employees and early investors — solving the most pressing problem of staying private (illiquid shares) without surrendering the freedom that private ownership provides.
The strategic logic is clear: going public forces a company into quarterly reporting cycles, analyst expectations, and short-term market pressures that constrain long-term thinking. Every major strategic decision Perkins has made — the two-year code rewrite, the enterprise pivot, the AI investments, the philanthropic commitments — would have been harder to execute under public market scrutiny. By staying private, Canva can continue to invest on its own timeline.
Tactic: If your business is self-sustaining, evaluate whether the benefits of going public — liquidity, currency for acquisitions, brand signal — outweigh the costs of shortened time horizons and external pressure, and consider alternative liquidity mechanisms that preserve strategic freedom.
Principle 12
Set goals that make you feel small.
"If I don't feel small and inadequate before it, it's not big enough," Perkins has said. "If it doesn't simultaneously move me, scare me, excite me, and humble me, the goal probably isn't big enough."
This is not motivational rhetoric. It is a design principle for organizational ambition. Canva's "crazy big goals" — annual targets that sound absurd when announced — serve multiple functions: they align disparate teams around a shared direction, they create urgency that prevents complacency, and they filter for the kind of person who thrives under pressure and leaves when things get comfortable. The celebrations — smashing plates, throwing tomatoes — are the reward mechanism that makes the intensity sustainable.
The deeper insight is that scale of ambition is itself a competitive advantage. Competitors who set reasonable goals will, by definition, achieve reasonable outcomes. Companies that set unreasonable goals and build cultures capable of pursuing them will, over time, accumulate capabilities and momentum that reasonable companies cannot match. The timelines may slip. The goals will be achieved.
Tactic: Set your primary annual goal at a level that makes you genuinely uncomfortable — not sandbagged-uncomfortable but sweating-uncomfortable — and build celebration rituals that mark the achievement, creating a cultural memory of what's possible.
In their words
Most planning is often done by looking at the bricks and trying to stack them. The column B thinking is thinking about what is that magical, wonderful future that you then want to invest years and decades of your life actually building.
My mum's living room became my office, and my boyfriend became my business partner, and we started enabling schools to create their yearbooks really, really simply.
In order to empower the world to design, we needed to simplify the whole design process and integrate everything into one place.
I think she's very capable of running a company that is much larger and more complex than the company she runs.
We have a crazy big dream: that by the end of our lifetimes, everyone on this planet will have their basic human needs met. It's terribly sad that this is still considered a crazy big dream.
Maxims
-
Begin at the end. The most important decision you will make is choosing your destination before you have any means of reaching it. Everything else is sequencing.
-
Smallness is not the same as timidity. Starting with yearbooks while aiming for the world's most valuable company is not a contradiction — it is the only rational path when resources are finite and ambition is not.
-
Rejection is iteration. Each "no" from an investor, customer, or partner contains diagnostic information. Extract it, integrate it, and treat your pitch as a product with its own development roadmap.
-
The missing co-founder matters more than the missing check. Capital follows capability. Close the skill gap first; the funding gap will close itself.
-
Invisible work creates visible results. The two-year codebase rewrite — silent, unglamorous, terrifying — enabled everything Canva built afterward. The willingness to invest in foundations is the mark of an organization that intends to last.
-
Profitability is not a phase; it is a discipline. Self-sustaining economics create the optionality to say no — to bad deals, to premature public offerings, to the tyranny of quarterly reporting.
-
Distance is a filter, not a handicap. Operating outside the consensus geography forces you to develop independent judgment. Use it.
-
Brand the experience, not the mechanism. Users adopt outcomes. Call it "magic," not "artificial intelligence."
-
Purpose scales when it's structural. Embedding philanthropy in the ownership model — not the discretionary budget — ensures that giving grows automatically with the business.
-
The goal should make you sweat. If your primary objective doesn't simultaneously excite you and make you feel inadequate, it isn't ambitious enough to attract the people or produce the outcomes your company needs.
