A year in review
2017 didn’t go quite as I hoped it would. In January, I imagined I’d spend the better percentage of the year on Kiwi, an AI natural language bot that I’ve been excited to start for years. And in some ways that’s happened – I’ve built a proof-of-concept grammar parser for speech input, and I’m nearly done building my own wikipedia text corpus that I’ll be able to use for AI training, but I’m conflicted about my progress this year.
Here’s my problem: with Jotlet (and Remotely, and Loose Leaf, and just about every other project I’ve ever started) I spend all of my energy on building the product, with little regard for customer development. I’m an engineer, I love building. I’m not a sales guy or a marketer, I’m not nearly as comfortable reaching out to potential customers. I’d rather build something than spend the time making sure it’s worth building.
With Jotlet, I spent years building the product and hoping that it’d somehow naturally go viral. Barely out of college and reading Techcrunch all day, I thought that’s how all successful tech companies made it big. We did eventually adjust course, focus on marketing and reaching customers, and that led to Jive. We certainly should have started sooner.
With Loose Leaf, I spent nearly 2 years building the app before showing it to anyone. Shortly after launch, I realized I was making the same mistake all over again, and shifted to spend the next 6 months solely on marketing. I learned a ton – much of it documented in the blog posts here and in the App Launch Guide – but even with all the effort it was too late to salvage my mistake.
This year, with Kiwi, I vowed to break the pattern. So instead of coding, I spent my time looking for a co-founder, shopping my idea to anyone who’d listen, even applying for accelerator programs. Ironically – I think this was the exact wrong time to focus on customers and right time to focus on coding. The AI market is so hot, that anyone accomplishing anything in AI is a big deal. There’s just too many people with AI ideas that trying to focus on the market/customer at this point without a proof-of-concept is foolish. I lost myself in the crowd of what-if-I-build-this, instead of standing out in the crowd of I-actually-built-somethings.
It’s taken me a bit over 10 years, but I think I’m finally understanding my weakness: I focus on either building or marketing, instead of both at once. It takes two legs to stand.
My goal for 2018: Build Kiwi. Build something awesome. Build it for myself. Build it for anyone. Open source cool stuff. Just build build build build build build build. And as soon as I have a proof-of-concept: show someone. Don’t wait until it’s “done,” show my progress at every step.
Building is what I’m good at. It’s what I enjoy. It’s what I need to be doing. And as I’m building, and after I’m building, and while I’m building, I need to show my progress even with the rough edges. I’m good at building, it’s what I’m best at and it’s what I enjoy the most. I need to refocus 2018 on building, and keep myself open and honest about my progress.
My favorite part of 2017
With all that said, I did focus on some very meaningful things in 2017 that I’m very eager to continue into 2018. I spent more time volunteering in 2017 than ever before, and it’s been extraordinarily rewarding.
I started volunteering with the Prison Entrepreneurship Program, and I’ve loved every minute of it. If you’re in the Houston area – I’d love for you to get in involved too – please do reach out I’d love to talk to you about it. If you’re not in the Houston area, please donate. The results of PEP are absolutely inspiring: < 7% recidivism (compared to 50% national avg), 100% of participants are employed w/in 90 days of release, nearly 100% still employed a year later (compared to 50% national average), and much more.
Let’s all build a better year than we did last year.
3 thoughts on “Making every year better than the last: my goals for 2018”
I have gone through a similar over ten year process.
One related insight, by building smaller projects it makes it easier to have the time to do both coding and marketing.
That’s a good point – I have a tendency to dive into huge multi-year projects, and I might be able to better balance with smaller timelines. I’ll have to do some thinking about how to apply that to my new AI project this year.
Show me! I’m interested! 🙂 Unfortunately, I have zero contacts that would be useful for a builder-of-bots. But always happy to catch up!