Skip to content

Duolingo’s Journey: From a Bot-fighting Test to Most Iconic Edtech Unicorn

There is a statement, ‘If you are not paying for it, then you are the product,’ that everyone agrees up in today’s data-driven internet market. While there are slim chances of getting a free service without losing data to cloud-based Advertisement AI today, Duolingo has managed to do it (1).

With a valuation of over 2.4 billion USD, Duolingo (2) is a free language learning portal. It offers gamified courses to keep users interested in learning a new language, accessible from both an app and a browser on every major platform, including Android, IOS, and Windows.

As of March 2021, Duolingo offers over 106 different courses in 38 languages, including tiny cards, a flashcard-based learning system, Duolingo for school, a language learning portal for students and tutors, and a Duolingo English proficiency test, which gives you an English proficiency certificate for a minimal fee (3).

The platform has a whopping 500 million registered user-based with over 42 million monthly active users who use the app to learn more than 2k words and 8k sentences of different languages of their choice for free (4).

The learning platform is also known for adding fictional languages like Klingon and High Valyrian to its list, which has gathered many enthusiastic geeks.

The Concept behind Duolingo

Luis Von Ahn, a professor at Carnegie Mellon University, Pittsburgh, known for his work in CAPTCHA and Re-CAPTCHA, and his graduate student Severin Hacker founded Duolingo at the end of 2009.

They based the platform on the principle of Re-CAPTCHA. It is a major crowdsourcing information program where the system takes information from users to verify if they are human. It also converts to text scanned images that the machine can’t comprehend (5).

But, how are all these related to Duolingo?

Well, it operated on a similar model to that of Re-CAPTCHA. While Re-CAPTACHE worked on crowdsourcing words that a machine can’t read, Duolingo took human help to change a text from one language to another. Even though Translate Tools are making progress, they are not yet capable of making a precise translation of one language to another, especially a few years ago.

To tackle this issue, Ahn introduced the ingenious business model of Duolingo. With the gamification process, users are captivated to learn a new language. Since the whole app is free, it offers added incentives to the users. However, what they fail to realize is that the app offers text to translate as a part of the learning process. It means users are translating text which is going to be useful to some websites. When a set number of users present the same translation for a particular sentence, the app considers it legitimate translations and saves it as the translated text.

How Does Duolingo Monetize?

When Duolingo was invented, it offered the premium experience of an ad-free app for free. So how was the platform earning revenue?

The platform had the client base of content-heavy platforms like BuzzFeed and CNN, who wanted translation for their content. Duolingo did it for them with the help of crowdsourced translation at an effective and much lower price than a professional translator. Thus, the platform helped users to learn a language while also making revenue to sustain itself.

However, in the past few years, Duolingo has tabled the initial idea and has added tiny while not intrusive ads to generate revenue since the initial model was pretty cumbersome and didn’t generate enough profit. It has now moved to the freemium model as it offers a paid version that removes ads and allows users to get gems, an in-game currency that lets users unlock content beforehand and offers extra chances.

The platform is also known to run A/B tests to find what ads work regularly, the amount of intrusiveness on the advertisements, and the consequent response of users to each ad. Overall, Duolingo tries to maintain the system as user-friendly as possible while finding out ways to maximize its revenue.

It also offers the Duolingo English proficiency test, a certified test that offers users validation certification at a nominal charge. It also offers Duolingo for schools, which offers two separate accounts for students and educators to track and train new language in educational institutes.

What Made Duolingo So Successful?

When Rosetta Stone, a language learning software company, went public back in 2009, it secured over 112.5 million USD on its first trading day. It built a successful public company by foraying into a huge but fragmented market, worldwide language learners, and recorded more than 209 million USD in annual revenue (6).

However, the founders of Duolingo believed that Rosetta Stone was capturing a too small market and growing on a bad business model. Rosetta stone was only targeting a small language learning market while charging them a fortune (7).

Duolingo believed that they had a better idea, and based on Google Trends, we can see that a lot of the globe agrees with them.


The founders found a way to go after an even bigger market than Rosetta Stone. They aimed to tap into the hundreds of millions of people who wished to learn a new language but could not afford expensive software (8).

Two primary things that helped Duolingo build a huge and successful company while sticking close to their initial vision of tapping a big market. Additionally, the platform meticulously experimented with and used its data to make its decisions.

The Factors That Have Shaped Duolingo’s Growth

  • Duolingo’s early business model of testing and crowdsourcing helped it create a helpful app for free.
  • The monetization requirement prompted Duolingo to double down its business translation services and find more ways to enhance user engagement to generate their translations.
  • Duolingo focused on user use-case and expanded its services around education, leading it to new monetization strategies.

Notably, Duolingo has constantly experimented with monetization while also experiencing struggles and failures. Even at present, they don’t seem to have a clear path. However, Duolingo’s growth is a lesson in setting up the right experimentation conditions. It built the audience and process required for testing and saw what worked best. Along the way, Duolingo learned a lot about their business and market. And most notably, the platform managed to create several business expansion options itself for the future.

Future Directions for Duolingo

As Duolingo was nailing down its plans to monetize the platform, it created many potential paths it can take in the future. Here are some ways it can grow its usage, expand its service, and add additional revenue streams:

  • Bifurcate into a B2B Translation Service and a Consumer-facing App: Although Duolingo may not want to do it initially, it is common for businesses to grow by targeting businesses and consumers at the same time with multiple offers. The strategy can be very successful for Duolingo as long as it allocates resources to the consumer site. With Duolingo’s funding and its plans to significantly grow the platform, it has room to expand its B2B side without taking additional time, resources, and people away from working on their core consumer use cases.
  • Continue Growing into Education: TinyCards was an excellent first move for the education space. However, there are several opportunities for other educational tools with similar monetization models. TinyCards allow users to study user-generated content around a wide range of topics such as geography, anatomy, etc. If Duolingo plunges deeper into some of these verticals, it can create entirely new education apps in a similar, gamified way. There are potential spaces for Duolingo to grow (9), enabling it to expand to language classrooms globally and classrooms for several different disciplines.
  • Branch Out to Other Businesses: Apart from translation, other adjacent B2B collaborations make sense for Duolingo. For instance, there are potential partnerships in segments like international travel, where Duolingo partners with hotels or travel agencies to offer two-week crash course lessons for international travelers. It only has to be sure to collaborate its tangential business services closer to its primary use case. Notably, Duolingo has already collaborated with Tinder last month to offer users a chance to find a partner in another language (10).

Entrepreneurial Lessons from Duolingo

Duolingo achieved something that one in the language and translation space would have ever thought was possible. It built a free language learning platform with millions of users hooked globally while growing a successful business along the way.

No matter what space you are in, Duolingo’s commitment to its vision and how it transformed an industry serves as an inspiration. There are three primary lessons entrepreneurs can learn from Duolingo’s journey. It can help any business stay adaptable and set itself up for long-term success.

Go After a Large Market, Yes, Even Larger Than Your Competitors’

If Duolingo wanted something simple, it could have built a gamified Rosetta Stone competitor. It would have been engaging and fun, and it could have sold the end product for hundreds of dollars. After all, Rosetta Stone was successful and doing well in a decent-sized market.

However, Duolingo looked beyond what others were doing. The platform considered real problems in language education and those most affected by these issues. It led the company to find a group of people who were not getting already available options in the market, leading it to an even bigger market than its competitors.

It is effortless to look at the current market size of your and your competitors’ products and take it. However, if you look beyond it, you can find untapped market potential that your competitors are completely oblivious to.

Hence, instead of thinking about where your company and your market are right now, start working backward. Commit to a goal for what you want your business to be doing in the next five, ten, or even twenty years, and think about what that means about the market you will have to go after and the type of business you will have to create (11).

For instance, if you are working towards becoming an industry authority on CRMs, you may aim to eventually have more than 15 million visits annually to a blog. To get there, you will have to figure out how to grow your blog from less than 150k a year visits up to 15 million and what you will have to do at each of these stages (12). It includes creating an initial blog, finding product-market fit, and creating an efficient content pipeline.

If you are like Duolingo, looking for a huge audience may mean you have to offer something for free. Moreover, Duolingo has demonstrated that there are always creative ways to monetize, and one can make it work if one keeps his focus on the end-user.

A/B Testings, Tracking and Adjusting Based on Results

One of Duolingo’s winning strategies was that it was data-informed. It made micro-changes, enabling it to grow usage and engagement little by little. Such incremental improvements may not seem like a lot at the moment. However, over time, it adds up. And thanks to their consistent tiny optimization, Duolingo climbed from 100k users to 500 million users in 2021 since its inception in 2009.

One of Duolingo’s most significant optimizations was counterintuitive and subtle. The team noticed many users were downloading and even opening the app but were not signing up. And to encourage it, the team experimented with moving the signup page back a few screens. The team had no idea if it would work, but they hoped that if they gave users a taste of the app before asking for sign up, they might not see several drop-offs.

They were right. However, the team would have never known it if they had not A/B tested the placement of the signup screen. With such small adjustments, the team was able to increase DAUs by over 20% (13).

Here are some ideas about running experiments and making data-driven decisions from Gina Gotthilf at Duolingo (14):

  • Constantly brainstorm test ideas. Don’t throw out any ideas until the implementation stage. Then, one should evaluate how many users it would affect and priorities getting statistically significant results for the sample size.
  • Use an analytics tool that enables tying data and actions to specific users. If one can not parse via the data with his tools, it is worth investing in one that offers high-quality user analytics.
  • Never assume that data is only quantitative. Qualitative data can be useful in market research, and it helps one get a different understanding of what is effective and what is not from the users’ perspective.

One can not set an endpoint in mind when setting up a successful system for testing and experimentation. One has to be open to the idea that one test will lead to another.

Don’t Be Afraid to Evolve The Business Model

Suppose Duolingo had stuck to its original plan to monetize based on B2B service. In that case, it may not have created all of the interesting and genuinely useful consumer use cases that it has now. Being experimenting and flexible with its business model allowed it to make crucial changes that pushed the company forward while staying true to its original mission.

Duolingo had the freedom to change its business model so much because it was in a really big market, and it had a lot of data on users that could back up its hypotheses. It was in an ideal situation to evolve.

Besides Duolingo, there are several other names in SaaS that have changed their business model over time, including HubSpot, Dropbox, and New Relic (15). In fact, for several companies, evolving a business model can be a catalyst for growth since it opens up new markets and opportunities.

Our advice for founders who are trying to find the right business model fit for their firm is to think of these points:

  • What’s your product’s value proposition, and what’s its time to value?
  • What is the competition?
  • What is the worth of each target customer and potential market value?
  • What is the quickest way to help your customers win?

In Duolingo’s case, moving from B2B translation service could have been terrifying for them, especially considering its promises to investors (16). For several companies, making a big business model change is scary. However, if one has a strong vision for his product and a mission to fulfill for the target market, change in the business model is needed for innovation.

Final Words

Duolingo has grown so much over the past years and has proved itself how useful it is to hundreds of millions of users learning a new language globally while growing a business.

Even though Duolingo has tabled its initial idea of offering a free premium experience and creating a win-win scenario, it is still considered one of the most ingenious business models. With the next generation of crowdsourcing data models coming in, Duolingo is set to be a pioneer in this incredible business model.

We expect more surprises from Duolingo as it continues to experiment and learn what works best for its business and its users.

The platform has so much potential, and we are excited to see what’s next!