The mission is unchanged—let the world converse

2016 was a time of great exploration and learning and yet I have the same driving desire now at the year’s conclusion that I had when it started—make language learning faster, easier and cheaper. Most of my friends have been surprised that language learning still holds my interest. They fall into one of two camps. Either they’ve had great success and lost interest or they’ve basically given up. I fall between the two groups, having ended up doing pretty well but only after a lot of suffering.

English matters

another city, another Airbnb

This isn’t really about me though. In all honesty, I don’t need to have an easy time learning languages. Learning another will have at best a small effect on my future comfort and standard of living. But it’s not like that for people learning English. My good friend in college wasted his parents’ savings in an ultimately unsuccessful English learning attempt and went back to a diminished future in Japan. My neighbors in Vietnam at the beginning of 2016 could triple their incomes if they spoke English well. That still haunts me.

I know something about teaching English. I did it for years and even built and ran a school in Taiwan with a couple of partners several years ago. To this day, I’m still very proud of the educational outcomes. Students went from nothing to handling level 3 and 4 OUP Bookworms (graded readers at the B1-B2 level) in three years of study as part of a 4 or 6 hour a week class. I’ve visited a hundred English schools around Asia and never since seen any similar schools come close to that level of result.

It wasn’t perfect, of course. It was an all-consuming job keeping that school running. I was probably putting in close to 70 hours a week during the time I was there and I still failed to put quite enough effort into training the staff or much of any effort into marketing.

Another issues is that we were charging a premium to our competitors. Most families couldn’t afford it. In truth, about 1/4 to 1/3 of my students were affluent enough that they’d probably learn English well eventually, regardless of how much they picked up. The people who would really benefit the most would never by served by my school, other than perhaps indirectly by learning about the merits of extensive reading from their friends who attended.

Sadly, I’ve seen this problem in an even more extreme form in tech start-ups who serve language learners. With a few notable exceptions, such as Duolingo and HelloTalk, they focus on 1-on-1 teaching. That’s even further outside the price range of my Vietnamese neighbors from a year ago.

An online platform is the solution, but those who most need it can’t afford teachers.

How I spent the year

My 2016 had three seasons—the first of being a digital nomad, the second of working for a win-win collaboration with whoever I could, and the third of going my own way.

Being a digital nomad

Being out of the expensive and yet incredibly sketchy SF bay area was great. It gave me new perspective, a chance to be a beginner foreign language learner and a way to keep expenses down while investing in my own human capital.

Programming

Most of the programming I’ve done had been in JavaScript (or JS-based) and I’d done more front-end work than server-side work. So I decided to try again to learn Ruby on Rails. It didn’t go so well when I made the attempt as a non-programmer back in 2008, but this time (late 2015 / early 2016), I found it pretty easy. In fact it was so easy that I could build things more quickly than I had in Node.js (my previous server of choice)! It was definitely pretty gratifying given how supposedly uncool and out of date rails is.

I made a rails site to host the interviews I’d done with language learners. I made a basic app for estimating the size of a language learner’s vocabulary. I also got a webRTC video chat built, kind of like what Speaky has. It was pretty awesome. Rails wasn’t the ideal back-end for something with all the traffic that a free tool can bring, but it was enough of a start for me to validate the demand and start working on building something that could scale cheaply enough that it might be possible to give away for free.

Growth hacking

Early in the spring, strangely, both Google’s and Facebook’s recruiting teams reached out to me about engineering roles almost simultaneously. I was getting a bit low on savings and was seriously considering going for it. Though I wasn’t really looking for a job, both of those companies have fantastic engineering cultures and I have friends at both. As an added plus on Google’s side, maybe if I actually worked for them, I could get access back to my old YouTube account that got clobbered when they merged all the accounts into Plus 😛

But I really didn’t want to go back to the Bay Area and basically spend all of my time working to pay for a tiny $3500 apartment I don’t have time to live in. I asked both if they had anything available in Japan, Taiwan or Hong Kong and they said no.

So I started putting my skills at the intersection of engineering and marketing to work. I offered to help pretty much anyone I knew who was working on anything related to education. I ended up doing a mix of free work for free products I admired and paid work for a scaling US start-up that worked out well. During this time, I went into a burst of creative activity, journaling, writing ebooks, making LINE store stickers. It was fun, but in the back of the mind I could feel time slipping by. I also suffered a pretty bad injury.

Injury and doubts

Months of heavy computer usage, all on my Macbook Air took a fearsome toll on my wrist. Since laptops have connected keyboards and monitors, it means the monitor is always too low and the keyboard is always too high. The raised keyboard put stress on my wrist. The track pad was even worse. At first I just felt a bit sore after using the computer for several hours, but then my wrist started to hurt when I was just typing. A month later, it hurt to brush my teeth or unlock a door. I needed a stable home with a desk with a proper keyboard tray.

I also needed to work harder towards my mission. Why was it that after months, I’d made such a small contribution? A big part of it was a lack of financial security. Another was that I still hadn’t fully recovered from a painful personal betrayal suffered earlier in SF. Was I crazy to turn down offers from two A-list tech companies? Was I doomed to never succeed again after my epic transition from English cram school boss to being an engineer in respected silicon valley companies? Thoughts such as these were difficult to keep at bay for a few months.

Working with brick and mortar schools

One thing that I could definitely win at was building another English school in Taiwan. My last attempt had grown more quickly than any I’ve heard of any direct competitors doing in the area then or since. And back then I didn’t have a curriculum already written. I hadn’t made supplementary exercises for hundreds of readers. I didn’t have CDs full of recordings I’d made. Even more importantly, I didn’t have any real engineering or marketing skills back then. The incumbents in the market had barely changed in a decade wouldn’t have a chance against 2016 me!

But… building another school wouldn’t solve anything. I’d just be back where I was, serving only a lucky few people in my area who could afford it. For that matter, growing a school “really quickly for Taiwan” is still a snails pace compared to the growth of online schools or even language schools in China. I could easily spend years doing it and make only limited progress towards the larger goal. There had to be a better path.

Collaboration Attempt 1

I could offer my services to an existing school. I could help them both improve their curriculum and drastically improve the ROI of their marketing. I could provide such an unusual value to them as an experienced entrepreneur in their niche who also had a silicon valley tech and marketing background that there would be plenty of upside for both of us. I’d work with them, improve the outcome of their students and still earn enough to build an online platform for those who can’t afford English teachers.

It failed. I had little autonomy and no power at the school. They wanted me to teach and in theory wanted me to help with the curriculum, but it just wasn’t going to happen within a time frame that made any sense to me. It shouldn’t have been a surprise, given that most language schools are a bit too old-school for me to have much of a chance. I can understand where they were coming from but it wasn’t a good fit.

Collaboration Attempt 2

I remembered two foreign-run schools in particular from when I was in Taiwan before been doing well when I’d left. Both had grown quite a bit, had branches in many cities and had reputations for pretty open management. I contacted them and tried a slightly different approach—I offered to be a business within their business.

Basically, in exchange for their administrative support, I’d offer some new classes with separate branding that would start from my previous curriculum and I’d incrementally build tech tools that both offered value to the classroom students and would be a path towards the online platform I’m building. I’d handle all of the sales and marketing. Any innovation that worked, they could use. I’d let them take a minority stake in the business and give them the right of first purchase so they could take it over when my online-only platform was self-sufficient.

Just as I was talking with these two schools, FB recruiting approached me again, and this time with an even better position on the growth engineering team. I was nearly broke after my recently failed collaboration and this team was working on very interesting things that I knew I’d love. I nearly abandoned the course and just moved back to California. But I didn’t. I just couldn’t.

One of the English schools got back to me and said no, they don’t do franchises or share equity in collaborations. The other school expressed interest, in principle, and was also interested in other tech and/or collaboration. I told Facebook no and went for it.

This collaboration also failed. I do believe there was some genuine interest in what I had to offer, but it was simply too far outside the scope of their normal operations and it probably wasn’t apparent just how large the potential upside was. Ultimately the business wanted to engage my services only as a teacher. While the rate was excellent for the Taiwanese EFL market, taking a job as a teacher in a school similar to the one I used to own and for a lower salary and less autonomy than I had even in 2005 would have been madness.

Building it myself

The managers of both schools I worked with were great people and I have respect for both their business sensibilities and their people skills. Given that there wasn’t a viable collaboration with either, I reluctantly set aside my plans of leveraging my domain expertise in brick and mortar language schools to help launch online efforts. I still feel that there was an potential path there, but I think the only way to have truly been credible to local entrepreneurs would have been to actually build another language school so they could see the results.

The next plan was to work part time doing whatever paid reasonably well and self-fund. I soon found an opportunity as an engineer at an Israeli-run tech start-up focused on language learning that seemed pretty cool. It would have paid more than even great teaching jobs in Taiwan and the founder was okay with me working four days a week, remote. I ran the idea past a couple of good friends. I’d been just about to go for it but they convinced me not to. Having seen my torturous path over the past half year, the made a compelling case to just throw all my effort into building the thing I believe in. I’d never be happy working full-time for someone else’s language learning company because it’s just too close to my heart.

Recent progress

  • raised a small amount of investment
  • opened a bank account in America (wow was that hard!)
  • gotten Amazon to give me a bunch of free AWS credits
  • rebuilt Lingsprout on a 50x more efficient tech stack
  • added basic social features
  • had many chats with the target market
  • reverse-engineered Twilio’s helpers to use their video product on this stack
  • abandoned Twilio because they raised their rates more than 100x
  • started testing market interest for a paid product to ensure survival without further investment
  • built to an email list of about 500

Serious problems remaining

  • lack of content
  • lack of users
  • a language learning site without the above isn’t useful
  • no credit/debit card from the back account causes enormous hassles
  • long-term visa situation

I’ve been aware from the beginning that there would be a network problem and that many of the things I could offer via software would have no value until getting a certain number of users. Where I’ve unexpectedly fallen short of my goals has been getting content.

I know a ton of businesses and people who have invested gathering or creating language learning content. I had expected it would be a fairly easy process to work out a licensing deal a new startup could afford, such as a revenue sharing or affiliate model. It wasn’t. After spending way too much time pursuing content deals, the one language learning company that agreed to license content didn’t actually deliver any. So, as soon as I’ve either finished building the paid product mentioned above or raised enough investment to ensure continued operation, I’ll start making content for English learners myself.

It’s rough working alone and having to do all the admin, coding and marketing, and there are a zillion different directions the product could go. There are a lot of paths this could take, but it will happen. That’s because I haven’t given up and I won’t give up until it’s built.

Plans for the next few weeks

  • add text chatting to the site (nearly ready)
  • launch sales page for paid product
  • figure out where I’m going to live

I’ve been incredibly frugal the past few months and for that I’m glad. However, revenue can still simplify a number of issues, including powering the development of more free features. The chat features should both help from a standpoint of helping me ask users their needs in real time and with selling.

The harder problem will be figuring out living arrangements. I don’t yet qualify for a start-up visa in Taiwan. There are a number of countries to consider, but nowhere else standing out as a clear winner. In terms of visa hassles, Vietnam is probably the best bet. Korea might make sense, too.

Lessons learned

Over the past year, I put so much effort into seeking mutually beneficial collaborations with everyone I encountered in the language space… and it was a waste. I tried to work with solo “language guru” entrepreneurs, I tried to work with huge tech companies, I tried to work with physical schools and I tried everything in between. Many times, there was a great potential for a win-win, but each time it went nowhere. This is despite the fact that, being driven more by mission than profit, I regularly offered the lions share of the gains from cooperation to the other parties.

I’ve always had a bias towards cooperation and seeking the benefits it can provide and perhaps that’s what made it take so long for me to realize that people don’t generally want to collaborate in business. Almost tautologically, people don’t know the value of the things they don’t know. They also don’t know how trust-worthy a potential collaborator is. More importantly, people place more value on potential loss than on potential gain. Finally, there’s a bit of plain old conservatism and laziness. It’s always easier to stick with the status quo and change nothing.

So here’s what I’ve learned: Friends mean the world and networking has value too, but actively trying to collaborate with a business isn’t nearly as worthwhile as I had thought.


Want to help? Comment below and tell me what’s hard about language learning and what resources you wish you had.