Caroline Clark

Arcade is the software that 14,000-plus companies, Salesforce and Zapier and Carta and Superhuman among them, use to turn a product into something you can click through before you ever sign up. No screenshots. No two-minute video that goes stale the day a feature ships. You move through the real interface, guided by the same little pulsing dots Clark borrowed straight from video games. That borrowing is not an accident. She wants B2B software to feel like play, and she has built a company on the bet that joy converts.

The origin is a grudge. At Atlassian, where Clark was one of the first product marketers for Jira Service Desk, the team once needed a single interactive visual for a blog post. It took two months. "Such a painful process that we were like, okay, let's not do this again." Most people file that frustration under the cost of doing business. Clark filed it under unfinished business, and four years later it became a company.

The discipline she carried out of Sequoia

Before Arcade there was a year at Sequoia Capital, where she served as chief of staff to partner Bryan Schreier, recruited into the firm by Alfred Lin after Stanford business school. She talks about that year the way some people talk about a graduate seminar that rewired them. What she took was less a rolodex than a creed: "rigor, discipline, and focus on quality," plus what she calls a writing culture. When COVID hit she left to start Arcade. Her Sequoia mentors, the ones who could have talked her out of it, backed the leap.

Lin handed her a line she still repeats: "Be competitor-aware but not competitor obsessed." You can watch it operate inside Arcade's product decisions, which tend to ignore what looks clever and chase what an ordinary person can actually do at nine in the morning without reading a manual.

The rule of seven, and other refusals

Clark coaches Arcade's customers with a piece of cognitive psychology she calls the rule of seven: short-term memory cannot hold more than about seven steps. So her advice to people building demos is blunt. "Don't try to make a documentary." The whole philosophy is a series of disciplined refusals, of complexity, of feature creep, of the temptation to show everything you built. "You must earn the customer's trust," she says. "You have to earn the right for them to want to be able to understand what a product does."

The same restraint shows up in how she reads data, which is to say, suspiciously. Arcade once deleted its Record Video button on the logic that only twenty percent of users touched it. The backlash was immediate, because the twenty percent who used it used it constantly. The lesson stuck: a number can be true and still lie about what matters. Another time the team worried that leaning on screenshots instead of live front-end components would feel too static. Then they checked: out of a thousand companies, exactly two wanted more interactivity. She listened to the other 998.

When Kleiner Perkins led Arcade's $14 million Series A in November 2024, partner Mamoon Hamid joined the board and reached for a comparison most founders would kill for: Arcade, he said, creates the kind of storytelling richness that Slack brought to workplace communication. Existing investors Upfront Ventures and Foundation Capital re-upped.

Clark is not using the money to build a better demo. She is using it to make the word obsolete. "Currently, customers use Arcade to demo a product they have already built," she says. "The bigger vision is to own the entire product life cycle." That means Arcade at the napkin stage, validating an idea before a line of code exists, and Arcade at the other end, helping teams decide what to build next. The new AI-powered Creator Studio is the first move: it generates on-brand videos, visuals, and interactive demos in minutes, and they never go out of date, the exact problem that ate two months of her life at Atlassian.

There is a through-line in how she leads, and it is not about software. On hard personnel calls she has said she weighs what is ethically right above what is merely legal, even when it costs the company. The hardest part of founding, she says, is "the gray area," the decisions with no clean answer. She designed Arcade's brand to feel light precisely because the work underneath is not. "We want to be something people feel excited to use, and it's accessible to anyone, no matter who you are."

Talk to Clark long enough and you notice she treats product-market fit less as a finish line and more as a habit. She has spoken openly about cracking the funnel by iterating relentlessly, testing, watching, adjusting, rather than betting the company on one grand launch. It is the same instinct that makes her suspicious of tidy metrics. She has been candid about the parts founders usually edit out: burnout, launches that flopped, the 4:30 AM stretches where the whole thing felt like a mistake. She names them on purpose, because pretending the climb is smooth helps no one who is also climbing.

One word she keeps coming back to is juicing it, the small flourishes that make a tool feel alive rather than merely functional. The pulsing dots, the load times tuned to feel satisfying instead of slow, the sense that you are moving through something built by people who wanted you to enjoy it. In a category that usually optimizes for looking enterprise-serious, Clark optimizes for delight, and then argues that delight is the serious choice, because it is what makes a person stay long enough to understand the product at all.

The AI work follows the same logic. Arcade is investing in synthetic voice, automated translation, and content recommendations, not as a checklist of buzzwords but as ways to keep a demo current without a human babysitting it. The promise is a piece of marketing that updates itself, the opposite of the two-month, instantly-stale visual that started the whole story. To keep all that video running smoothly across thousands of customers, Arcade leans on serious infrastructure under the hood, the unglamorous plumbing that lets the playful surface stay playful.

What ties it together is a writing culture she imported from Sequoia and a refusal to confuse motion with progress. Clark would rather ship a small, clear thing that an ordinary user can finish than a clever thing that impresses other builders. "Always prioritize the product experience to accomplish a result and focus on your end user." It reads like a platitude until you watch her delete a feature, ignore a flashy competitor, or build for the 998 people who never filed a feature request. Then it reads like a spine.