Language Learning Platforms show the limits of the MOOC model
MOOC platforms face technological threat as skill based platforms leverage superior learning technology to capture consumer markets.
MOOC platforms like to boast of their numbers, 37m for Coursera, edX half that. Duolingo has hit 300m users this month, up from 200m a year ago — a simply astonishing achievement for any company. It’s revenue figures are also impressive, launched as a free service it began adding premium product options and subscription, yielding $1m in revenue in 2016 and $17m in 2017, it is on track for $40m in 2018.
Duolingo is not a complex product and it has a straightforward market. The platform grew out of the ambition of Louis von Ahn (born in Guatemala) and Severin Hacker, Louis saw how expensive it was for Guatemalans to learn English and wanted to create an app to democratise education. The app uses a gamified memorisation as the core of its pedagogy. Users learn words in relevant clusters e.g. numbers, sports etc and then repeat them in different contexts. The gamification layer focuses on retention — users have to keep repeating words to ensure they stay fresh to help with long term memory consolidation.
The pedagogy is enabled by the technology — Duolingo was one of the first apps to effectively use speech and listening (it also requires reading and writing) to drive the learning — that was critical for the app to gain its viral growth — it had to be usable via mobile phone. A singular focus on language comprehension and memorisation has also enabled Duolingo to focus their data efforts — including their widely shared post ‘How we learn how you learn’ an effective use of their data to show the optimal times to memorise to aid long term retention.
In previous articles I’ve argued that the future of much of Edtech might well lie in specialists over generalists because of their ability to be more agile operationally and develop superior learning platforms for their content area. Coding is one obvious example and I’d argue the main MOOC platforms already come off worst for the basics of learning how to code (not so for the more abstract computer science). Language is unquestionably another.
Both follow a similar pattern in that they are more skills than abstract learning (and to a point, both languages) and that the bulk of user demand sits at the skill end enabling them to better match the market needs. Language learning is also replete with technological development — spinouts are emerging that are able to test on pronunciation, chatbots are emerging to provide basic conversational partners all fueled by big tech investment in AI-assistants.
For the consumer this isn’t a bad thing but for MOOC platforms it’s another area where their generalist technology — absent further innovations — will curtail their content reach. FutureLearn made a lot of hay off the British Council courses, the world’s largest MOOC, much as coding drove enrolment growth at Coursera and edX but looking forward its not obvious that MOOC platforms are a good delivery source for the core skills part of language learning — or coding. MOOC platforms may not mind — arguing the growth was important at the start but they’ve moved onto higher value areas — true enough — they have degrees now. However, the rise of nimbler — superior technology enabled companies — demarcates content boundaries for MOOCs and they may not stop at entry level skills. Duolingo has already sought its own English language credential, more platforms will follow and if they can get buy-in from their industry — they could start eroding Universities’ key advantage — their recognised credential — here
