“And you may find yourself”…selling an Ed Tech app in Asia
Part II — The Pivot
A small part of our grand website scheme provided video chat between students. We believed that real life communication was the key to language learning, but were wary of beginners using video chat without sufficient guidance and content. They’d be under a mountain of stress to produce and understand an unfamiliar language. That could destroy their engagement and motivation.
Our original idea was to provide unlimited tutoring, interactive lessons, video classes and notes before each chat — enough content, we thought to enable some kind of meaningful exchange for each session. Now that our BIG, expensive idea had been soundly rejected, what small part could we salvage and refine into pure gold?
We got this idea of providing scripts to learners, allowing them to chat by reading translated material. Naturally, providing inflexible dialogs would get boring really fast, so we tried to think of ways to create branching conversations that empowered user choice at each step.
I practiced a few of these branching scripts (English → Mandarin) with Erick. They seemed to work, and I started to feel some sense of proficiency for basic expressions. We felt that there must be a way to present this idea in a way that both engaged users and, more importantly, produced real speaking proficiency.
Naturally, there were unique problems with this approach. If each user had 3 choices at each turn, and each choice generated 3 new ones, they’d reach 1.65 million choices in just 13 sentences. We figured we‘d develop a chat flow system, a set of rules, and a taxonomy (breaking it into more bite-sized chunks) that avoided us having to spend the rest of our lives creating navigable forests out of all these proliferating branches armed with only a machete.
We also decided focus entirely on chat and on mobile chat, in particular, to take advantage of the fact that the vast majority of all digital communication is done through mobile and because chat across 12 time zones (e.g. NYC/Beijing) would be best accomplished if users could do it pretty much anytime, anywhere.
Realizing that it’s tough to get users committed to practicing a Second Language regularly, we gave the app a dating-style look and feel that encouraged personal bonding as much as language practice. After all, if people were motivated as much by making friends or even dating as by learning a new language, they’d learn it faster anyway due to higher engagement and motivation over the long haul.
We showed this to a few of our “tough” friends and they were like — “OK, we get it now. This could work”. We were psyched and ready for any travels, troubles, and tribulations to come. Which was a good thing because they were ready and waiting.
We also picked up another co-founder, Christopher Davis, an old friend who’d served as a Creative Director for some very big Ad Agencies and founded his own as well. Christopher had won multiple awards and had a brilliant and supple mind, which took him far outside the box at times. He’d also spent 9 months as a monk in a Burmese monastery. His 30,000 ft. view was valuable as we developed a larger framework in which to understand our entrepreneurial journey.
Together we developed a vision for where Language Hero was now and where it needed to go. What were the “big ideas” propelling it? How would we market it? What kinds of benefits might be realized through its development cycle?
“When you see a fork in the road, take it”
In July of 2016, we were accepted into a New York-based accelerator program called 2020 Startups. It was a new program that lacked the cachet, financial support, and connections of some of the more established accelerators, and it actually charged each participating team a few thousand dollars to participate but, we were looking for a “game changer”, and Y-Combinator certainly wasn’t looking at us. We took a chance on 2020 Startups.
And…as it turned out, in that moment, it was precisely what we needed. It laid some more hard truths on us, gifted us with our future “nerd”, and introduced us to a hustler-type manager who’s hacking-based approach to entrepreneurship would eventually rub off on us in a big way.
We’d also begun reaching out to academics in order to develop an English and Mandarin curriculum. We’d met Kate, an ESL Program Director at a local college and NYU Instructor through Linked In. Kate found the conversation tree an interesting concept and agreed to work on it to create an English–Mandarin demo version once we could locate a Mandarin teacher to work with.
Good fortune struck again in the form of Ramiro, a Colombian engineering student and serial world traveler, whom we met through a chance encounter at Taoyuan Airport, in Taiwan. Ramiro was then studying in Taiwan and had a Mandarin teacher named Melody. He introduced us to her by email and we immediately hit it off. Melody found our unique chat method intriguing enough to agree to work on the demo curriculum with Kate and…we were off!
Melody also set up a meeting with Dr. Chin-Chin Tseng, Chair of the Mandarin as a Second Language Department at National Taiwan Normal University, one of Taiwan’s top schools. We met her briefly just prior to flying back to New York to start the 2020 Program. Dr. Tseng felt that the App (just screen shots at that point) had some promise and agreed to follow our progress.
Getting to 20–20
Mark Gold’s watchwords were “Credibility, Visibility, Profitability”. As 2020 Startups’ Managing Director, Mark was a big advocate of hacking into any and every type of ecosystem that can advance one’s company’s interest. He recommended that we create a series of events for Language Hero at Ivy League Universities, which would at least imply a relationship with them and, therefore, a measure of (faux) academic legitimacy. Lacking the money or relationships to pull this off at the time, we kind of put it on the back burner. In the end, it turned out to be extremely good advice.
Making our time at 2020 worthwhile it all came down to what we could show on Demo Day. At minimum, we needed a working prototype that communicated the look, feel, and function of the actual realized product. It took several near-disasters with would-be app developers (including one who ripped us off for $2000 and produced nothing of value) to find Jason Chang, a Taipei-based iOS developer. Jason intuitively grasped what our chat process needed to feel like for users, and he executed it beautifully. His work has stood the test of time.
While at 2020, we also met our future (and current) CTO, Evgenii Podkovyrov. Evgenii was the CEO of another company, who’d been reluctant to take on that role. He had been working in the AI space for almost 10 years, since building a still profitable trading platform with a college roommate. In Language Hero, he saw a way of using AI to develop a more natural feel to conversation than what NLP (Natural Language Processing) alone could possibly provide, leading to far higher engagement, which was and is our “holy grail”.
Evgenii felt that the app’s approach to SLA “[had] to work”, as he put it. By the end of our 2020 program Evgenii was on board. More importantly, he committed resources from “Quantum Brains” the FinTech firm at which he was the managing partner to building a finished product.
The addition of Evgenii to the team solved two distinct issues — having enough “runway” (financial resources to continue operating),and having the in-house capacity to make the endless rounds of changes and tweaks that characterize virtually all app development. The first issue fosters investor confidence in a startup’s ability to marshal and conserve resources — the second guarantees they won’t burn them up anyway.
As Demo Day approached we had more good news. Jason Chang had completed the demo right on time and it worked beautifully with the content provided by Melody and Kate. Not only that — due to a scheduling issue we had an opportunity to officially launch Language Hero at the Taiwan Economic and Cultural Office in New York (equivalent to their Embassy), where Dean Tseng and several other professors were visiting. The launch went very well. Dean Tseng was impressed enough to request specific changes to the interface for our next meeting, which we set for February, in Taipei.
We psyched when Demo day finally arrived a few weeks later. We felt that we’d finally arrived and that the world (OK, the investment world, at least) would now start to take notice.
Read Part III to find out what went down on Demo Day and beyond.