Duolingo: Sheer Genius

I have wanted to learn Spanish for a while now, and that was why I started looking into Duolingo. I started the research thinking it’s just another language learning program. But boy was I wrong.Value creation – For students (users)

Duolingo is a convenient, gamified, engaging, and effective way to learn a language: A students creates an account, which keep tracks of his/her learning progress. One can access Duolingo via its website or mobile app, making learning a portable experience. The lessons come in very small, easy-to-digest modules.

As students progress, they collect Duolingo’s virtual currency (called Lingots), which enables students to purchase items such as skill level assessments. It is almost like the Candy Crush of language learning, except the only way to get lingots is through continued learning.

One key feature is its translation feature. As students advance, the program lets them practice through translating real life content from one language to another. In doing so, Duolingo made its lessons even more engaging and relevant (users have described the program as addicting).

Duolingo also actually works. One study has shown an average of 34 hours spent on Duolingo is equivalent to “a full university semester of language learning”.

Best of all, it is completely free. No more need to spend hundreds on Rosetta Stone.

Value creation – for businesses (complimentors)

Remember how Duolingo lets students translate content to practice their learnings? Turns out, the content is actually sourced from real companies with real translation needs! Duolingo has partnered with CNN and Buzzfeed                to help with their international sites. Native Spanish, Portugese, and French speakers learning English will practice their skills while helping CNN and Buzzfeed translate English content into different languages.

Why is this helpful to businesses? For one, traditional translators are expensive (~20 cents/word). Further, in the traditional model, there are not enough bilingual speakers to complete all the translation needed.

Duolingo attacks those challenges by “employing” (for free) eager language users, thereby drastically reduces the cost, effort, and time. It is estimated that 1 million active users can translate Wikipedia into Spanish in just 80 hours (Duolingo has ~100 million users).  In addition, for each sentence, Duolingo asks multiple students for translation and aggregate the results to reach a final output on par with professional level translators’.

Value Capture

Today Dualingo generates revenue in two ways: by partnering with companies (such as Buzzfeed and CNN) to translate content and collecting a fee. In addition, is it also noted that for users accessing web version of Duolingo, the company can monetize on people’s browsing data

Network effects

In case it was not clear, Duolingo is more than just a product. In fact, a learning program is probably not too hard to replicate. It is the company’s “platform” strategy (connecting users with business needs) that will enable it to build a robust network ecosystem and deter new entrants.

Multiple levels of network effect exist in its business model:

  1. Direct network effect between users: as more users enroll in Duolingo, the positive feedback will bring in more users (I was introduced to Duolingo by friends who had good experiences)
  2. Direct network effect between companies: Similarly, as Duolingo generates positive feedback, more businesses can sign up for Duolingo’s translation services
  3. Indirect network effects:
    1. As more users sign up, there will be more availability of translators. Thus, more businesses will be inclined to purchase the service
    2. More users also allow Duolingo to collect more web browser data for further revenue generation
    3. As more companies become interested in the translation services, their needs/business will allow Duolingo to offer more relevant content for practice in more languages, thereby attracting more users
  4. Product improvement:
    1. Based on data collected from people’s learning behavior, the company can continue to improve its curriculum, increase its teaching effectiveness and attract more users
    2. Based on translation feedback from businesses, Duolingo can enhance its translation offerings to better meet business needs
    3. Duolingo can also take data and feedback to develop new product offerings (e.g., partnering with schools for language education)
  5. Brand: To take a page from BSSE, today the language learning and translation industry are offering solutions that “not good enough”. Therefore, by building up a strong, integrated network, Duolingo is establishing a powerful brand that people will be willing to pay a premium for.


Okay, time for me to go back to the app and try to collect some Lingots now.


How AirBnB Might Ultimately Fail


Lending Club – Pioneering P2P Lending

Student comments on Duolingo: Sheer Genius

  1. Great post! Dueling is now of the few companies that have managed to crack the online education model effectively.

    Dulingo’s translation model reminds me of Captcha, which used crowdsourcing to improve image recognition softwares. It is certainly successful but it remains to be seen how scalable and substantial it can become.

    BTW do you know if Duolingo is profitable?

  2. For the most part I agree with you; however “positive feedback” attracting new users is not a direct network effect. If that were the case, essentially all products would be utilizing network effects. My buying a Fiat because my friend told me it has served him well (unlikely) is not a direct network effect unless my Fiat experience somehow improves because he owns one too. You are spot on when you say that more user translation data improves duolingo content for everyone though. 🙂

    1. Thank you for the feedback! You are totally right that the Fiat example/positive feedback is not a direct network effect. This also reminds me that I should’ve also noted Duolingo has enabled “discussions” (e.g., a forum like feature) for users to interact with each other, thereby building a community to retain its users.

  3. I have heard of (and used!) Duolingo, but had no idea they sold translation services to large companies — how cool and compelling. I feel like this topic could be an interesting lead-in to our crowdsourcing module…

  4. Thanks for the post! I did not know that Duolingo captures value by “employing” users for translation services – that is brilliant!

    I have not tried Duolingo myself but have used competitor Babbel instead. They are harnessing direct network effects by encouraging users to interact and help each other progress. For example, someone who is a native speaker in English and wants to learn Spanish is paired up with a native Spanish speaker who wants to learn English. Does Duolingo try to do something similar?

    1. Yes they do (and I probably should’ve noted that in the blog post too) – there is a discussion feature where users can post questions and topics. From a first glance, the topics cover a wide range (e.g., language questions, to motivational posts, to feedback for Duolingo). It seems Duolingo is also trying to build a community where users establish ties with each other.

  5. Interesting, thanks for sharing this example! I have used other language learning sites before (e.g., Busuu), but I have not tired Duolingo yet. What is fascinating to me is how certain companies found clever ways to monetize on their platforms by understanding the motivations and needs of their community of users. Duolingo’s users are effectively working for the company without realizing that and are recommending the product to others, brilliant!

    It would be interesting to know if language learners tend to multi-home, i.e., whether they use other sites in parallel, thus decreasing overall time spend on any one site. Are there any ways to decrease multi-homing? Also, is it possible to create stronger direct network effects among language learners to create a more sticky service?

Leave a comment