Killing a Start­up

Making Friends Who Can Code

To­gether with Sam Piecz , I co-founded a web ap­pli­ca­tion for vi­su­al­izing per­sonal fi­nan­cial data and test-building port­fo­lios. Since 2016, I worked as a de­signer and— even­tu­ally— as the go-to Javascript (JS) coder, for Cor­re­latePro.

Sam, who'd grown up on the Rasp­berry Pi, knew Linux and the com­mand-line in­ter­face (CLI) like the back of his hand. He showed me how to nav­i­gate the ter­minal; how to use Vim, Tmux and Github; and got me setup with basic de­vops: in our case, a vir­tual en­vi­ron­ment and Docker . Any tech­nical work I do on a com­puter today is be­cause of Sam.

Sam and I met in the Boy Scouts— even­tu­ally we both went on to earn the rank of Eagle. When we were younger, Sam and I went to punk, metal, and elec­tronic shows. We started skating to­gether, went to some weird sub­urban par­ties, and got into some light-weight trouble.

Sam started (and has since sold) an elec­tronic cig­a­rette "e-liquid" com­pany— Heating Up Vapor. I de­signed the la­bels for his bot­tles, and had started working on some custom pack­aging. As we got older, how­ever, al­most every time we went to hang out, we were re­ally just sit­ting on op­po­site sides of the room coding on our lap­tops.

Learning Slow

Sam got me started with basic CSS and HTML. He got me a job doing re­mote work in search en­gine op­ti­miza­tion (SEO). In un­der­grad, I was get­ting a de­gree in ar­chi­tec­ture. I en­rolled in (and dropped out of) a Python course. Sam kept teaching me and en­cour­aging me.

At grad school, I took a course in data vi­su­al­iza­tion. I didn't know any JavaScript, but they were teaching us to learn D3 — a JS li­brary for "data-driven doc­u­ments".

I barely passed, but con­tinued to prac­tice— even­tu­ally im­ple­menting a few basic vi­su­al­iza­tions in our ap­pli­ca­tion. I re­cy­cled some code, a force-di­rected graph, and im­proved upon it for adding stocks/​cryptos/​sec­tors to a mock port­folio. You could vi­su­alize the cor­re­la­tion be­tween those as­sets (based on dis­tance/​den­sity).

We Ac­tu­ally Built Some­thing

Be­fore we took down Cor­re­latePro, I'd also im­ple­mented a can­dle­stick graph with TechanJS . We had a free plan with a few fi­nan­cial data APIs. Sam setup some in­ternal APIs to move that data (re­al­time prices, tickers, and his­tor­ical costs) to the fron­tend. I took in the data, fil­tered it and for­matted it, and made those dates, times, prices and quan­ti­ties leg­ible. We had news, a public dash­board for viewing other users' port­fo­lios, and sen­ti­ment analysis based on Tweets.

Over time we saw a couple boosts in page-views and click-throughs as Sam con­tinued to submit new pages to Google's Search Con­sole. Working re­motely, we used Trello to keep track of tasks and com­mu­ni­cated with Slack. We posted to Hacker News and Reddit, built a com­pany Face­book page, and put some ef­fort to­ward an email mar­keting cam­paign. We were both strug­gling to con­tribute with school— and lost a third co-founder early on.

After about three years, we still weren't get­ting enough feed­back to it­erate. The few users we did have, mostly friends and family, weren't en­tirely sure how to use it; and the money we'd sunk into server costs was starting to add up— even though all our li­braries and APIs were free. We went through YCombi­na­tor's Startup School, and ap­plied sev­eral times for YC funding to no avail.

⚰ RIP ⚰

All the videos and pod­casts we'd shared— in­ter­views with founders and ven­ture cap­i­tal­ists (VCs)— told us we should have quit a long time ago. How­ever, SEO takes time. Star­tups take time. They take ded­i­ca­tion, and we weren't giving up.

There were weeks, even months (studying for fi­nals), when we made little or no progress; but we al­ways came back. We'd get burnt out, but we'd pick it up again and keep trying.

When Sam took down the server, I was ter­ribly upset. It had con­sumed our time and our at­ten­tion for years, and it felt awful... At the same time, it was lib­er­ating. We'd both been so fix­ated on fixing this thing without knowing if anyone would want it— even if it wasn't broken.

Learning Lessons

I don't want to tell you to give up, but I do want to say that— if you are a strug­gling en­tre­pre­neur— know when to move on. It might not be ob­vious, and it def­i­nitely won't be easy, but if you're ever going to suc­ceed it's cer­tainly nec­es­sary.

Sam and I learned every­thing we did about fi­nance while working on Cor­re­latePro. We wanted to make in­vesting as simple and straight-for­ward as pos­sible— for our­selves, and for everyone else who sees in­vest­ments as op­por­tu­ni­ties to rise up fi­nan­cially.

While we lost some­thing dear to us, we learned a great deal. It was an in­valu­able ex­pe­ri­ence that will give us some in­cred­ible in­sights when we move onto the next pro­ject. We did fail, but we're not giving up.