Duolingo was founded in 2011 to make a dent in language education. Six years later, with 170 million users worldwide, it is safe to say that it has.
The free—and addictive—language learning program is the most downloaded app in the Education category on both iTunes and Google Play. Designed with game elements—levels to complete, skills to unlock, studying streaks to maintain, and virtual goods to buy with virtual currency that a learner has earned—the app brings fun to the often tedious pursuit of a new language, and naturally lends itself to word-of-mouth marketing.
Kickstarting that word-of-mouth expansion at scale in different global regions has required finesse and an effective strategy. At the Traction Conference in Vancouver this month, Duolingo’s VP of Growth and Marketing, Gina Gotthilf, described some of the tactics and principles the team used to attain its enormous user base.
Start with good intent and a compelling story
Naturally, having genuinely good intent behind a business provides the basis for an interesting story.
Duolingo’s raison d’être is to provide more accessible education. To achieve this, Duolingo founder Luis von Ahn took on a high leverage piece of education that he claims 1.2 billion people are currently pursuing: the learning of a foreign language. He has said that 800 million of these learners are people studying English to get out of poverty.
In a 2015 discussion with President Obama at the White House’s first demo day, von Ahn said, “Most people talk about education as something that brings equality to different social classes, but I always saw it as the opposite: something that brings inequality. The people who have a lot of money can buy themselves the best education in the world, and because of that remain having a lot of money, whereas people who don’t have very much money barely learn how to read and write, and because of that don’t end up making a lot of money.”
Learning a second language, especially English, can impact a person’s earning ability. However, mastering a language requires a long term commitment and repeated exposure. Effective formal training in languages outside of public school is expensive, setting up an ironic situation for those looking to lift themselves out poverty through language acquisition. This typical catch-22 is what makes Duolingo’s free service compelling.
Duolingo has also had an interesting effect on the popularity of certain languages. Take the Irish language for instance, for which the team launched lessons in 2014. Gotthilf explained that while there are only approximately 94,000 native Irish speakers, over a million are learning the language on Duolingo.
Beyond this, von Ahn us something of story himself. He helped to create CAPTCHA, the program on websites that requires users to input a set of characters to distinguish them from bots, and then went on to sell two companies to Google, and win a MacArthur Fellowship, also referred to as a “Genius Grant.”
PR as a Growth Hack
While public relations is not typically included in growth hacking, Duolingo’s team has been able to leverage the press to create a ground swell of buzz that has led to significant new user acquisition, especially when entering new markets.
Gotthilf says, “First I’d create relevance around Luis,” reaching out to top universities in new markets to see whether they would be interested in having him as a speaker. While very few people around the world know von Ahn and his accomplishments, universities computer science professors often do. To these academics, von Ahn’s accomplishments and the fact that he’s also a computer science professor at Carnegie Mellon University have relevance. Professors are often receptive to a pitch for a free in-person appearance by the founder.
These pitches would be timed to occur while prominent tech conferences were scheduled in the region. Gotthilf would use the credibility gained from the university appearances to pitch von Ahn as a guest speaker at the conferences.
Because of how tailored these offers were to their respective audiences, it was often easy to get a “yes,” and with those appearances confirmed, Gotthilf now had an angle to pitch to reporters she and her team would contact over LinkedIn and Twitter.
“I used that hook to capture reporter awareness,” Gotthilf says. It was then only a matter of “reaching out to journalists to say, ‘Hey, the most important university and the most important conference have both invited [von Ahn] to speak,’ before asking if they’d like to interview him.”
Gotthilf says, “We were able to impact our number of new users doing that.”
Relentless A/B Testing
“A/B testing is one of the key factors that led to our growth,” Gotthilf says. “We question everything and we look for answers in the data.”
According to Gotthilf, the necessity of A/B testing crystallized in their minds after researching literature on language acquisition only to encounter completely contradictory assertions about what worked. There was no consensus on best practices.
So, when in doubt, they experimented. Which group would show greater engagement? The 50,000 users learning plurals before food vocabulary, or the 50,000 doing the opposite? Putting conventional language acquisition theories aside, they let the data lead their way.
Applying this philosophy to the inclusion positive reinforcement yielded stunning results. The team programmed their digital mascot Duo to appear and say encouraging words, especially when a learner’s performance was poor. This positive reinforcement led to a 2% increase in daily active users, 6% increase in seven-day retention, and a 7% increase in fourteen-day retention.
Showcasing how far the team would go to optimize the experience, they even A/B tested the number of tears Duo would shed when a lesson’s questions were answered incorrectly.
The combined results of a strong mission, tailored public relations, and relentless A/B testing have led to a virtuous cycle: A/B testing leads to more engagement and user success which in turn leads to a greater number of compelling stories to pitch to the press which in turn leads to more users to include in A/B tests.
The result: a rapid approach to the 200 million user mark.