Intro

My name is Hiro Yamada, and I’m the co-founder and CEO of First Mate Technologies, a software development studio. In May 2023, my co-founder and CTO, Mark Yao, and I set out to revolutionize language learning, using generative AI. We raised $500k from VCs and built a functioning product with about 30 paying users.

Despite our efforts, we ultimately pivoted from our original vision and transformed into a development studio assisting other startups.

Why? Reflecting on our journey, we made several mistakes along the way and identified five key lessons to help others avoid the same.

‍

What We Were Building: Fluency

Fluency was a Zoom/Google Meet plugin designed for non-native English speakers. It analyzed users’ spoken English in real-time, offering feedback on grammar, vocabulary, and pronunciation. Our ultimate goal was to help professionals improve their English fluency and confidence.

The technology worked, and we have one paying customer that purchased 30 licenses—a premium language learning academy for business executives. However, scaling beyond that proved challenging.

‍

Why We Failed

1. Losing Passion by Moving Away from the Problem

Fluency was born from my personal experience as a non-native English speaker. While living in the U.S., the language barrier was a daily pain point. I often felt my English skills caused me to be undervalued by 20-30%, which fueled my passion to solve this problem.

However, after starting the company, I moved back to Japan. The move made sense: Japan had a large potential customer base, my co-founder Mark was based in Asia, and lower living costs extended our runway.

But an unexpected side effect of the move was losing touch with the problem. In Japan, I was often the most fluent English speaker in the room, diminishing the sense of urgency and personal connection to the issue. Without that passion, it became harder to persevere through challenges.

One example of a challenging moment was a product rollout at a Japanese high school. The students laughed at how poorly the AI performed, and a teacher made a snarky comment about how they “expected better from a Harvard grad.” Experiences like this happened almost daily for ten months, and without strong passion, the resilience to keep going waned.

Takeaway: Passion is often rooted in personal experience. Founders should stay close to the problem they’re solving to maintain their drive and resilience.

‍

2. Our Product Was a Vitamin, Not a Painkiller

The pain of struggling with English fluency is real, but Fluency addressed the problem indirectly by pointing out areas for improvement rather than solving it outright.

While users appreciated the feedback, most weren’t willing to pay significantly for it—especially with free resources like YouTube, Duolingo available online. Businesses also didn’t see enough value to justify purchasing it, leaving us with just one paying customer.

Takeaway: Products that directly alleviate pain points are easier to sell and sustain than those offering incremental or secondary benefits.

‍

3. Not Paying Ourselves

When we raised $500k, many mentors emphasized conserving cash. It made sense—having funds meant more time to iterate and adapt. We avoided hiring beyond a single contract engineer in Vietnam (who still works with us!) and refrained from taking salaries.

I lived in a one-bedroom apartment in Tokyo with my wife, taking 5 AM calls with the U.S. that inevitably woke her up. Instead of renting office space, I worked from home, public libraries, or budget coffee shops.

Initially, this lifestyle felt like the classic “garage startup” story. Over time, though, the discomfort of cramped spaces and financial strain eroded my energy and morale. Seeing my personal bank balance dwindle each month only added to the stress, making it harder to focus and persevere.

Takeaway: At the beginning, a startup’s most important resource is the energy and motivation of its founders. Create an environment that sustains this, even if it costs a bit more.

‍

4. We Overrelied on Trial Contracts

In 2023, processing an hour of meeting audio with our AI engine cost about $1, which forced us to price Fluency at $20/user/month—too expensive for most customers in the language learning market. For context, some human tutors charged as little as $3/hour, making our tool a tough sell.

We initially believed our pricing was reasonable because three language schools agreed to paid trials at $20/user/month before the product was even fully developed. These trials brought in a few hundred dollars in revenue, reinforcing our optimism.

However, once we launched the product, all three customers churned after their trials ended. When we dug deeper, we realized these schools couldn’t sustain paying for our tool. Their revenue and cost structures simply didn’t align with our pricing.

In hindsight, we suspect timing played a key role. In 2023, GPT-4 had just launched, and the industry was buzzing with curiosity about how AI might disrupt language learning. This curiosity likely drove decision-makers to experiment with our product despite its high price. Ultimately, though, curiosity wasn’t enough to translate into long-term value or retention.

Takeaway: Paid trials can create a false sense of product-market fit. Validate long-term demand and affordability before banking on early interest.

‍

5. Pick Your Industry Carefully

EdTech is notoriously tough. It’s a crowded space where passionate founders and modest opportunities often lead to lower valuations and fierce competition. For instance, our $5.5M post-money valuation was in the bottom 25% for pre-seed startups according to Angelist, reflecting the sector’s challenges.

This isn’t to say EdTech was the wrong choice—it simply meant we were playing on hard mode from day one.

Takeaway: Carefully consider your industry’s dynamics before diving in. Some sectors are tougher than others, regardless of your team's quality or idea.

‍

Let’s Build Together!

We’ve made our fair share of mistakes—and learned invaluable lessons along the way. Since pivoting, we've worked with over a dozen startup founders, encountered many common pitfalls, and leveraged our experience to help our clients navigate out of them.

If you’re building something and want to avoid these mistakes, let’s talk. We’d love to help!