Duolingo: Working for your Education
Duolingo claims to provide a social service (language education) for free, all supported by your translation services. Will it become sustainable?
Digital transformations have created operational and market efficiencies that generate value for businesses and their consumers. However, many new tech companies, driven by the competitive nature of open platforms like the iPhone App Store, have also catered to the millennial generation’s tendency to want services for free. In a space where the mentality is market share first, many apps provide free services, hoping critical mass will drive future ad revenues. However, other software developers have used more creative methods to both generate and capture value.
Duolingo is an example of such a creative value creation and value capture play. As its founder Louis von Ahn explains, duolingo was conceived as a ‘twofer’, a business proposition that generates two separate kinds of value from the exact same event or transaction.1 For duolingo, this transaction is your language lesson, or more specifically, the translating component of your language lesson. Duolingo creates a “gamified” language lesson experience2 on its smartphone app, part of which is contingent on users practicing their newly acquired language skills by translating short sections of text. When the user partakes in these translations, they exercise their language skills and gain points for completing each lesson. Duolingo’s business model however, hinged on the realization that there is also inherent value created through the act of translation. In effect, they can get users to translate any short segment of text provided for free, and through aggregation of many users’ exercises could gather a full translation of any document1. Duolingo can therefore monetize its free app by selling a translation service on the back end.
While technology has drastically changed the scope and reach of education through online platforms, very few online course platforms are profitable based purely on individual users. Language course platforms like Rosetta Stone cost over 100USD for most basic packages3, and have significant institutional customers4. Large non-institutional free course platforms like Khan Academy are non-profit and supported by donations5. More recent information suggests that duolingo’s translations business may not so sustainable and profitable in the long-term. As a privately funded company, little information is available regarding its financial status, however a $45 M series D funding round in June of 20152 suggests that profitability has yet to be achieved. While duolingo seems to have successfully provided translation services to high-profile customers like CNN, the company seems to have stopped scaling up the translations business, and is no longer accepting new clients for translations2. Instead, it may be turning to new sources of revenue such as language certification tests6.
Duolingo demonstrates both creative value generation and capture, and the challenges of successfully executing such a model to achieve profitability. In order to succeed, Duolingo needed to provide highly compelling value to both of its two separate customers; duolingo users, and translation service users. As a language lesson platform for individual users, duolingo has captured a significant market as a smartphone app. As a translation service however, its position may be more fraught. Despite the potential benefits of verifying translations by compiling a large volume of overlapping user translations, the fact that translations are being provided by language learners rather than professionals may deter a large portion of customers. Duolingo’s value proposition would therefore likely be as a low-cost translator. In addition, the most valuable translations would come from users who had invested significant time to become advanced users. Given that user churn is often high on smartphone apps, duolingo may have challenges retaining users until that point. Duolingo’s public rationale for not pursuing the translations business was that it would divert too much attention away from the social impact aspect of the business. In other words, valuing education over revenues. Given the initial vision for the company to be self-sustaining, I would speculate as to whether the costs associated with the duolingo translation business turned out higher than anticipated.
In today’s global environment, language is a critical barrier to unlocking economic value. However, professional translation services can remain exorbitantly expensive, in part because automated processes like google translate have not yet attained a level that is acceptable for professional translation of full documents. Duolingo has a strong social impact value proposition, providing free language education to bridge language gaps. The data duolingo is generating is academically valuable as it can give significant insights into human learning behaviors and patterns. However, its business proposition may be challenged by its ability to generate high-quality translation outputs as well as its limited perceived value to buyers of translation services. As profitability is key to duolingo’s success as a for-profit education company, I sincerely hope they will successfully monetize their business, either by improving their translation business value proposition, or by partnering with large institutions to support their language certification program.
Word count: 788
1 Viktor Mayer-Schönberger and Kenneth Cukier, Learning with Big Data: The Future of Education (Houghton Mifflin Harcourt, 2014), PDF e-book, [9], [https://books.google.com/books?id=OXzeAgAAQBAJ&oi=fnd&pg=PP1#v=onepage&q&f=false], accessed November 2016.
2 Frederic Lardinois, “Duolingo Raises $45 Million Series D Round Led By Google Capital, Now Valued At $470M,” techcrunch, June 10, 2015, [https://techcrunch.com/2015/06/10/duolingo-raises-45-million-series-d-round-led-by-google-ventures-now-valued-at-470m/] accessed November 2016.
3 Rosetta Stone, “Rosetta Stone® Language Learning Pricing Overview,” http://www.rosettastone.com/pricing, accessed November 2016.
4 Rosetta Stone, 2015 Annual Report, p. 28 [http://investors.rosettastone.com/phoenix.zhtml?c=227935&p=irol-sec], accessed November 2016.
5 Khan Academy, “Donate,” https://www.khanacademy.org/donate, accessed November 2016.
6 Duolingo, “Get recognized for your English fluency,” https://englishtest.duolingo.com/, accessed November 2016.
Lisa – I’ve never seen a twofer business model like this before, so thanks for the interesting read. Much has been made of “the wisdom of the crowds,” and Duolingo is capitalizing on this idea. Even if one language beginner’s translation isn’t quite right, aggregating thousands of them and picking out the overlap should in theory give you a good answer. How does the theory match up to the reality, though? Is there any benchmarking for how Duolingo actually stacks up to professional translation services or other technological solutions like Google Translate in terms of accuracy? I would imagine the app is effective for simple word translation, but may miss idioms and other nuances that require a native or very well-trained speaker. Who is the typical customer for this low-grade translation service?
This company has a very unique model and also a very problematic one–I echo Dan’s fears that the translations might not be enough and wonder if they have released any metrics showcasing their accuracy. I also wonder if writing translations is the best way to learn a language, as it keeps you only partially immersed in the new language and mostly immersed in thinking of your own language (which would be what you’d be translating). Not all translations are straightforward vocabulary issues, too, and often involve a high degree of familiarity with idioms, etc. Also I wonder how it could improve speaking skills rather than just written. Maybe Duolingo should focus more on the educational aspect and maximize revenue by being a low-cost alternative to Rosetta Stone, but not free.
6pm and I just received my Duolinguo reminder for my Spanish lesson so that article is right on time! Being a loyal customer of Duolinguo, I will say that there is a limitation to the value proposition currently offered by Duolinguo. It is a good start if you want to try a new language or it is a good way to keep maintain a good level if you do not have the opportunity to speak a specific language on a day to day basis. However, learning only through Duolinguo will not allow you to become fluent in any language. I believe Duolinguo has now managed to create a strong brand in the language education and should capitalise on that to develop a more complete value proposition. Integrating even more technology (AR/VR) could be a way to offer a premium paid option for users that want to go a step further in their education. It would seem a more logical path as a way to monetise its brands rather than pursuing into translation where you are competing with big giants.
I appreciate the value created (and captured) through the gamification of learning foreign languages, but I’m curious about its efficacy. Academic consensus indicates that the ability to master new languages dramatically falls off in the early-teen years. However, new research has shown that there is no “speed bump” after which new languages cannot be learned, but it cautions that a more developed brain is not as sensitive to intuitively catching grammatical errors as less-developed (i.e. younger) brains [1]. I wonder if Duolingo could improve the efficacy of its product by customizing the way languages are taught on its platform based on the user’s brain development.
[1] http://www.independent.co.uk/news/education/the-best-age-to-learn-a-second-language-a6860886.html
Lisa – I love your post about education technology and Duolingo. I launched the ed-tech practice at my old VC firm and find the sector fascinating and underserved. My suggestion for Duolingo is to advance their B2B enterprise business and continue to launch professional development services for corporations. These contracts will be large and sustainable and will help fund the consumer facing side of the business (which should be kept free). Monetization is always tricky with consumers and language translation applications are largely becoming a commodity. It will be interesting to see how Duolingo navigates this environment.
Interesting post! The concept of the twofer business model appears to be a great idea but seems to be hard to execute, as expected. I share the same concern that selling translations provided by language learners is not sustainable or competitive. Has Duolingo explored more B2B opportunities to sell to corporations? I think particularly international corporations could be more interested in providing language lessons to its employees and this business would be a more steady, sticky business. I think Duolingo can also leverage more technology to build more interactive features to extend beyond translations of text, as others have pointed out that is not enough to master a language.
I also wonder if Duolingo can redesign its programs to pivot to target school children and partner with elementary schools or private schools, where language lessons are increasingly popular and parents realize that children learn languages best at a young age. Young adults and working professionals can access many resources, including Rosetta, for learning a language, and perhaps there may still be less competition in the elementary school markets.
Great post on an interesting business model. I think this post describes the challenge of aligning two values in one product/service, as well as its potential. One of its education business’s value proposition is its simplicity, while the demand for such a simple translation seems small. Also, since Duolingo’s target users are language beginners, it’s difficult to make translation buyers believe the quality of the translation. To enhance the quality of its translation while maintaining the simple learning materials, is it a potential solution to create a more interactive, community-based education service, so that users can grade each other’s translation and Duolingo can specify customers who are valuable translator?
Interesting! I want to learn more about their translation business. It’s not completely intuitive to me how learners would be able to translate a text in such a way that their translations can be used to fully translate any document. And it seems Duolingo is having trouble convincing their potential clients of the same. In any case, the language certification tests seem like a more intuitive route, one that aligns to their primary language learning business better.
I worry about the sustainability of Duolingo’s model, given the advancements in the same AI technologies that are supporting the company’s success. Ultimately, progression in AI will lessen the demand for language learning as apps are able to quickly and easily translate something. I wonder how Duolingo can leverage a portfolio of products to ensure investors are receiving their required return.