Jul 23, 2015

My First Big Software Startup Failure

As much as I enjoy reading about young people’s success in the software startup space, I never can get enough of people’s failures. After all, you learn more from your failures than from anything else. So here’s a bit on mine.

Back in 2012 I started learning web development for the sole purpose of building myself a better teacher website (I was teaching high school math/science/engineering at the time). After taking Udacity’s inaugural “web development” course, I immediately began cranking out a series of iterations toward my goal of making my teacher website better.

qCalculus

My first stab went way overboard - I ended up creating a collaborative online calculus course, complete with a discussion board and an awards mechanism. I launched the course on the first day of school and had abandoned it by mid-semester. My students loved it, but honestly it was too much work for me to maintain a complicated web application and teach high at the same time. So I went back to the drawing board, open-sourced my code, and moved on to the next iteration.

Calculus On Chrome

My next attempt was to create a calculus curriculum using a markdown + MathJax editor. My goal here was to put in place a curriculum editor which would allow me to quickly prototype lessons and publish them to my students. I also wanted no part in a complicated, “collaborative” web application which I knew would end up consuming way too much of my time. I was able to put this new curriculum in place by the end of the same school year. I consider this a success because it didn’t add much new additional work to my daily routine - I was already writing a curriculum and lessons, this just made writing them easier.

You’ll also notice that I started incorporating my love for python/programming into my lessons - much of the numerical bit of calculus can discovered using programming.

A strange thing happened by the end of that year - a couple teachers on my campus told me they’d like to have a website that was similar to my calculus one. Aha! I’ll make this a software product!

Lessonwell

Based on my success with Calculus On Chrome, I realized that I needed to make a teacher website that made writing lessons easier, not more time consuming. So my idea was to create a lesson platform that would allow teachers to quickly prototype lessons (especially math lessons). I spent much of the summer of 2013 at my desk working away at lessonwell.com.

This was my first venture into many things: marketing, collecting money, customer support chief among them. Building this software product consumed every single hour of that summer, and I had a hard deadline: if it wasn’t done before the start of school, I would have to wait until the next summer to hack on it. Not an option. Plus, since this was such an awesome idea, I’d be rolling in the Benjamins before too long and I could quit teaching high school altogether, mwahaha!

School started and I launched lessonwell to my own classroom. Simultaneously I launched another project I’d been hacking on called the badge project. Why? Well, I loved the award mechanism of qCalculus and wanted to reiterate on that. Plus, I couldn’t get enough web development experience; so I was hacking in every waking moment.

Students enjoyed it. Some teachers also loved it. Some of them even signed up, paid me money, and started using it for their own classrooms. How many? Well… only one.

That’s right, only 1 other teacher in all of the entire internet bought (and continued to pay) for a subscription to lessonwell. Eventually I realized that I loved software development more than teaching and open-sourced lessonwell to add to my portfolio. I also realized that I had made a ton of mistakes in building this business and needed some more time before I attempted another software product.

What Went Wrong

In hindsight, it’s painfully obvious to me what went wrong, and none of it had anything to do with the product itself.

  1. I spent zero time understanding my customer. I made the incredibly naive assumption that if the software was useful to me, it would be useful to everyone else! What I should have done is sat down with my teacher friends every day and asked them what they wanted in a teacher website.
  2. I spent zero time marketing. Well, not zero, but next to zero. I was on Twitter and Facebook and emailed people. But I made little-to-no attempt to meet people in person, pick up the phone, or otherwise made face-to-face meetings a priority.
  3. I had way too much going on at the same time. Software is addicting; building cool stuff is addicting. But I was building too much stuff. I was also a full-time high school teacher, was running after school astronomy and robotics courses (one of them turned into my next startup failure), and was the coordinator for the DFW EdTech Meetup.

In the end, it was a failure. But a glorious failure - the process of creating a teacher website for myself turned into a new career instead. I still like the idea of building a useful teacher website, and may revisit it in the future.

I’ve now got to figure out how to be a better business person.