ReadLang bought by Duolingo

[zenmonkey]

I don't mind the market being split in two parts, a large toy part, and a smaller serious part. It has always been this way (all those "learn Chinese in a month and 100 pages" courses). What I do mind is the serious part being pushed away from its already gained position and damaged by the toys.

If "serious" language learners want a good selection of effective tools, then two things need to happen:

1. There need to be enough serious language learners out there.
2. Those language learners need to be willing to spend money on tools (or write those tools themselves).

Sadly, the vast majority of people buying language courses (at least in the US!) aren't very serious. If you asked them to put in 20 to 40 minutes a day for 6 months, they wouldn't be interested. To use the usual sports metaphors, it's like training to run a marathon or getting in really good shape—most people can do it, but the price is higher than they want to pay. And that's OK.

But (2) is an important point, too. The iOS version of Anki costs $25, and the money the author earns from that also pays for desktop Anki. (The marvelous AnkiDroid version was written by volunteers.) And yet, if you ask around on language-learning forums, you'd find plenty of people who say things like, "Anki is way too expensive at $25!" Many customers are offended by software that costs more than $3, no matter how niche the market.

OK, sure, maybe Anki isn't worth $25 to you. But it's a great product aimed a niche market of "serious" learners. And it's not a huge market, which means it's going to be more expensive per person.

Let's see if we can figure out how big the market for the iOS version of Anki might be. I can't get download numbers for iOS apps, but the most popular paid flashcard app on Android seems to be Flashcards Deluxe at $3.99 with 50,000 to 100,000 downloads. This gives us a maximum all-time revenue of $400,000. And it's been around since 2009, which would suggest a maximum of $50,000/year before advertising, servers and the costs of running a business. Not bad, but a competent programmer in the US could certainly do better at a day job. Now, Anki costs $25, it runs on iOS, and it has a different audience, so we don't know how much it makes—it might be peanuts, or it might be a surprisingly nice living. But Anki has never turned into a giant company with investors and a marketing budget, so it's probably not making giant sums of money.

If "serious" language learners want good tools, the two main options are (1) pay enough for them in great enough numbers for the author to make a nice living, or (2) write the tools as open source and share them. Any other tools are living on borrowed time.

Just a small clarification Anki happens to be the top 60 highest revenue earning apps and top 30 downloaded in the education area. So Damian is making a little bit of grisby. It's number 7 once you remove the apps with in-app purchase.

It would not be overly optimistic to think it is downloaded well more than 1000 times a month which generates a revenue of well over $170000 a year.

[emk]

Just a small clarification Anki happens to be the top 60 highest revenue earning apps and top 30 downloaded in the education area. So Damian is making a little bit of grisby. It's number 7 once you remove the apps with in-app purchase.

It would not be overly optimistic to think it is downloaded well more than 1000 times a month which generates a revenue of well over $170000 a year.

Nice! I'm glad to hear he's doing well. I've gotten a huge amount of mileage out of Anki, and I'm glad to hear it's actually making somebody a very nice living. I guess this means it's likely to stay around for a while. (It's worth noting that Anki isn't just a language-learning product, which makes it easier to find a big market.)

The Andorid version AnkiDroid is an open source app maintained by a group of volunteers and distributed for free. This is incredibly generous of them, and if anything, it's even nicer than the iOS version.

...

And as for various earlier discussions in this thread, I agree that there are lots of problems with language-learning software. Just in this thread, people have pointed out that the UIs are too challenging (and doubly so for elderly learners), that $25 is beyond the means of many students, that the marketing is too focused towards Americans and other westerners with Internet access, and so on.

But the practical reality is that neither solo entrepreneurs like Steve Ridout nor open source volunteers have much chance of solving these problems alone. The solo entrepreneurs need enough sales to hire more staff—or at least enough that they don't give up and look for a full-time job. The open source authors need more volunteers, because it's not sustainable to ask them to give up too many paid work hours.

(Just to be clear, I'm doing just fine with my open source stuff, because I know when to say "No", and I've established clear limits as to how much time and money I'm going to invest.)

[Adrianslont]

...but I feel like there's been a lot of, "why did he take the job?" when to me it's a pretty logical answer

Really? I haven’t noticed anyone express that sentiment in this thread. And I’ve seen lots of discussion around the unprofitsblity of Readlang and the great financial opportunity Steve Ridout would be getting at Duolingo.

In this thread I’ve seen concern that Duolingo will kill Readlang or make it less functional but that shouldn’t be mistaken for “why did he take the job?” I’m pretty sure everyone in this thread understands why and wishes him well. I can say that for myself and I see other people saying similar eg emk, Cavesa and no one saying anything to the contrary.

[Cainntear]

Moving to the general point about good software, I think that the university market is the only place that can incubate good tools.

You've got a concentration of serious learners, and you've got teachers that generally aren't try to sell an "easy way".

The problem with the university market is that teachers want finer control than any of the commercial sector companies want to give them. For example, there's no option to create your own path in Duolingo for Schools -- you just set targets against the existing progress trees, so it's next to imposible with other materials and syllabuses.

Related to that is branding -- most universities have their own "virtual learning environments" now, and integrating your tool with university systems makes it more anonymous.

Still, I think building a "white label" service for unis is the only way to accumulate the tech for a complete computer-based course, rather than all these one-trick ponies..,,