Language-learning apps are a rarity in the VC-backed consumer tech space. That may have something to do with the sector’s historical struggles to make money. Rosetta Stone floundered on the stock market for a decade before selling to private equity in 2020. Open English fell off the radar after growing its valuation up to $350 million in 2013. Duolingo, the most successful of the pack, told Forbes it would be profitable in 2019, but filings it made at the time of IPO showed it ended up losing $14 million (last year, net losses increased to $60 million).
San Francisco-based startup Cambly thinks it has what it takes to buck the trend. Unlike its predecessors, Cambly focuses only on English, and unlike many others, it focuses on human interactions—rather than a structured curriculum, it gives users access to one-on-one conversations with English tutors. Founded nearly a decade ago, Cambly has quietly reached the Series B funding stage with $60 million raised in total from venture firms Bessemer Venture Partners and Benchmark. It was valued at $250 million after the latest funding round in 2020, according to CB Insights.
“If you look at the pattern of failed language-learning businesses, they overpromise how effective they are,” says Bessemer’s Jeremy Levine, who sits on Cambly’s board. “They grow really quickly by burning a lot of money on consumer acquisition. All of a sudden, the churn of consumers catches up with them and then no one wants to invest anymore.”
Cofounder and CEO Sameer Shariff thinks Cambly can avoid that pitfall because it has been cash flow positive for the last five years—prioritizing profitability long before the market downturn led VCs to recommend that startups be more frugal. The company is now generating “tens of millions” in revenue, he adds. Users pay a subscription fee that gives them access to tutors for a set amount of minutes per week—added up over a year, the cost can range from about $500 to $4,400 depending on the amount of usage.
Levine, a self-admitted skeptic of language apps (“they’re mostly hoaxes,” he says of their business viability), offered to lead the Series A funding round in 2018 after reviewing the financials and being surprised that the company was growing without burning major money on marketing campaigns. “You can fake revenues and you can fake growth rate through all sorts of accounting and metric manipulation, but you can’t fake cash flow positivity,” he says.
Former fellow software engineering colleagues at Google, Shariff and Kevin Law started Cambly in 2013 after identifying a market opportunity while traveling abroad. In Argentina, for example, Shariff recalls his picking up on conversational Spanish much faster than he ever did in a high school classroom setting. They decided to hone in on only English because of the demand among working professionals. “The world has kind of decided that English is the language of international business and trade,” Shariff says. “If you think about you or me learning Spanish or French, it doesn’t have the same economic unlock that English does.”
Benchmark’s Sarah Tavel, formerly an investor at Bessemer where she was mentored by Levine, led Cambly’s Series B round in 2020 and quickly pushed the company to expand its headcount. Cambly had no finance staff, and less than ten people working on engineering, product and design when she first met the founders, she says. “In order to capture the full possibility of the market, you want the team to match the opportunity.” Today, the startup employs 150 people (not counting tutors), but Shariff says he wants to share Cambly’s story now to attract better talent to the company’s San Francisco headquarters.
About 80% of Cambly’s revenue comes from individual users, and the rest from companies who want to train their employees in English. Cambly’s top markets are Brazil, Turkey, Saudi Arabia and Japan; barely any business comes from the U.S. “We basically built a product for everyone not like us,” Shariff says. The software matches students with English tutors—many use Cambly as a side hustle, though some have made it their primary source of income, Shariff says—based on professional needs. For example, an airplane pilot in Turkey can be matched with an English speaker versed in aviation terminology.
Cambly charges all users, so the business takes a different approach than the freemium model of Duolingo, which reported more than 40 million monthly active users who helped it generate $250 million in revenue last year. Cambly says it too has tens of millions of “students,” but a PR representative later clarified that it defines a student as “anyone who has created an account, including someone who takes a free trial lesson, but doesn’t pay for a subscription.” The company refused to share any details on its actual active user base, but Levine says the figure is significantly lower: “If users were in the millions, the company would be making billions.”
Still, the investor thinks Cambly doesn’t need nearly as many users to be successful. Doing some napkin math, Levine estimates there are 3.5 billion people in the world who could benefit from Cambly’s roughly $1,000-per-year subscription plan. If the company is able to capture even 0.1% (or 3.5 million users), it would make $3.5 billion in annual revenue. “That’s an order of magnitude larger than we are now, it’s not 100 times larger,” he says. “Could we get there in five years? I think we can, maybe slower, but maybe even faster.”