Inspiration
I was in my first year of university when I began to suspect that the way languages were being taught in traditional institutions was lacking. It was a university writing course (English) and there were two exchange students who both got 100% on our first grammar and sentence structure test. They could identify every grammar rule and the exceptions to each rule well enough to ace a multiple-choice test. However, while they surpassed everyone in their ability to recognize a dangling participle or a gerund, things I still have to look up to explain, they could not engage in much verbal communication. They could pass a test but could not describe what they did on the weekend. Years later, I found myself in the same situation while traveling in Japan. I had memorized the two basic alphabets and hundreds of Chinese characters. I could write them with the proper stroke order, understand the multiple meanings and sounds each character could make, and I could use them to construct a grammatically correct sentence. I could conjugate verbs and adjectives and have received perfect scores on both a listening comprehension and pronunciation tests! I had taken four years of Japanese in university and yet, the second someone in Japan started talking to me I froze. They spoke fast and used slang, and I found myself in situations that were not covered in my textbooks. I could pass tests, but my language abilities were far from functional.
|
A Growing Market |
Duolingo Won't Teach You Spanish! |
Right now, more than 2 billion people worldwide are studying a foreign language, with more than 29 million actively using the three largest digital language-learning apps, Duolingo, Rosetta Stone, and Babbel. Currently, digital language-learning generates $6 billion per year in revenue, but that number is projected to rise to $8.7 billion by 2025 (Adams, 2019).
The Competition Duolingo is by far the most popular language-learning app, hooking users with gamification, using reward points, treasure chests and “streaks” for continuous use. However, Duolingo has two major problems. First, with more than 28 million active users, Duolingo only realized $40 million in revenue last year, a mere $1.4 per user. The free version of the app comes with advertising, but an $84 USD subscription removes these but only 1.75% of users opt for the ad-free version (Adams, 2019). Secondly, despite the addictiveness and high retention rates among users, when it comes to actually teach a language, Duolingo does not work! Word Monkey overcomes both of these problems.
|
When using Duolingo to learn another language, a beginner is drilled on sentences like “I am the child”, or “I have a bowl”. This is audio-lingual drilling: there is no communication happening. Instead of basic communication, the users are drilled again and again in decontextualized, effectively meaningless sentences (Groves, Hopkins, and Reid, 2015). Joey J. Lee, the director of the Games Research Lab at Columbia University, suggests the addictiveness of tools like Duolingo has more to do with business models than with language learning (Freedman, 2018). Lee states, "Where most apps really fall short, he said, is in language “pragmatics.” “That’s the learning that’s based on real-world settings—you’re in a restaurant, in an interview, waiting for a bus,” he explained. “It’s usually lost in apps” (Freedman, 2018).
A Language App That WorksThis is where my app, Word Monkey, succeeds where others fail. Rather than simply drill grammar, sentence structure, and vocabulary, my app prioritizes the function of language as communication, not structure. While gamification will be used to introduce basic grammar and vocabulary, AI chatbots are used to engage the user in meaningful, pragmatic communication. Being able to conjugate a Japanese verb or saying "Watashi no oba no pen wa tēburu no ue ni arimasu" ("My aunt's pen is on the table") is not particularly helpful when trying to order dinner in a restaurant.
|
Language apps are incredibly popular,
highly addictive,
but they won't teach you how to
speak another language.
Word Monkey Works
People learn languages for different reasons and using Word Monkey, users will be able to select relevant scenarios in which to engage the chatbots in conversation. Apps like Duolingo and Rosetta Stone have attempted to address the limitations of app-based language-learning by introducing video conferencing options where users can practice with each other, but research shows that three-quarters of users hate the idea of talking to a stranger (Captain, 2016). Companies like Duolingo understand this and have experimented with chatbots, but voice recognition technology has struggled with accuracy and technical issues. Word Monkey combines the latest in voice recognition but also pairs this with facial recognition software that is not only capable of reading lips to support voice recognition but can also read user emotions. Liopa developed a lip-reading app that works with any device with a camera.
|
Liopa’s technology analyses the speaker’s lip movements and uses AI to decipher what the person is saying. The technology is not affected by background noise, which represents a significant problem for audio speech recognition, but when combined with it, leads to improved accuracy of the overall system. (McDaid, 2018). Users benefit from an immersive experience by engaging in realistic conversations that are situationally specific without fear of embarrassment. Companies like Affectiva, Amazon, Microsoft, and IBM now include “emotion analysis” as one of their facial recognition products, as well as a number of smaller firms, such as Kairos and Eyeris (Schwartz, 2019). Incorporating emotional analysis allows the AI to determine when a user is getting frustrated or bored, which allows for the app to adjust on the fly and modify the lesson to keep the user motivated. In order to incorporate the latest facial and voice recognition software with intelligent chatbots, any language-learning app is going to need substantial revenue.
|
Revenue Potential
The ability to outperform Duolingo in terms of revenue represents the second huge advantage of Word Monkey Duolingo makes most of its revenue by selling ad-free subscriptions. While only 1.75% of users are willing to pay, they still generated, $36 million in revenue, while Berlin-based Babbel managed $115 million in revenue from subscription fees it charges its million-plus subscribers (Adams, 2019). Despite Duolingo's huge base and roughly $40 million in revenue generated from advertising and subscriptions, it has yet to turn a profit (Adams, 2019). Word Monkey includes both an ad-supported free version, and an ad-free subscription option, but will generate additional revenue through embedded "product placement" ads. For example, a common scenario language learners engage in is ordering food and eating at restaurants. Instead of learning vocab, grammar, and engaging with our chatbot about ordering food at a generic or unnamed restaurant, advertisers can pay for product placement. Why learn to order a hamburger at a hypothetical restaurant when one can learn how to order a delicious Big Mac at McDonald's in Osaka or Paris? If one chooses to go on a coffee date with a chatbot, the AI could casually mention how delicious the coffee is at Starbucks. Even those paying for ad-free subscriptions will still be targeted for advertising, they just won't be aware of it. Emotional response data and demographic data from the facial recognition software can be sold back to the advertisers and used to develop more effective hidden advertising. User data can lead to targeted ads that can further generate revenue.
Educational And Profitable
Word Monkey combines the proven gamification and clean interface employed by Duolingo to teach basic grammar and vocabulary, but surpasses it by incorporating AI chatbots that simulate relevant real-world, immersive experiences. Users can select their language learning path based on their individual motivations and needs. Whether they want to learn another language for an upcoming vacation, business opportunity, education, romance, or just general interest, Word Monkey can provide them the opportunity to learn the language as intended, through meaningful communication but without the fear of embarrassment. Answering multiple-choice questions, or translating sentences is fine for learning vocabulary or grammar, but it will never teach someone how to speak another language.
With over 2 billion people learning a foreign language, digital language apps represent tremendous growth potential and a global market that is insulated against shifting demographics that can impact educational technology geared for a narrow market. The ability for language apps to generate revenue through both advertising and premium subscription services bodes well considering that potential revenue streams have yet to be fully explored.
With over 2 billion people learning a foreign language, digital language apps represent tremendous growth potential and a global market that is insulated against shifting demographics that can impact educational technology geared for a narrow market. The ability for language apps to generate revenue through both advertising and premium subscription services bodes well considering that potential revenue streams have yet to be fully explored.