Prior Art

Re the Two Ventures I’ve Gone Through
February 13, 2024 by ben - [lore, startups]

Goal and Audience

I have co-founded and co-failed two startups thus far. What follows is half post-mortem, half reflection, and half contextualization for my current practices.

This piece might be relevant if:

1. you’re curious about the two startups I’ve failed thus far
2. or you’re curious about what I’ve learned about being a founder.

I’ve written a more terse post-mortem on the first startup. If you’re looking for something like that, you should read Post-mortem: Ours.

Update: I’ve also written a post-mortem on the second startup. If you’re curious, you should read Post-mortem: Pallet Materials.

The First Try: Ours

The first time I got into a startup was mostly on accident.

It was Summer 2022 and I working at Instagram doing some very serious end-user experience research. (Read: I was swiping through stories.) There was a post by my high school classmate asking if anyone knew someone who’d want to do backend dev for a startup. Being a big CS nerd with lots of CS nerd friends, I swiped up and asked about the project.

Madeline gave me the story. Along with her sister (Adrienne) and her sister’s friend (Avery Chen), she was founding a startup. The vision: build an app that connects students running errands with other students that they might go with. It’d be a social-practical experience. A way to meet new people, under the guise of dealing with necessary errands and the nice pragmatic excuse of splitting the Uber cost.

The three of them had already piloted this with airport rideshares, which worked well in NYC where the Uber from JFK to Columbia can take an hour and cost over $150. They’d done interviews with students to talk about pain points and this is what they landed on. They’d gone to a Columbia pitch competition and won. Now they wanted to build a working MVP. While Maddie was (and still is) a very competent programmer, she was the only one who could code — and she was already designated as the team leader, with a plate full of responsibilities (startup and otherwise). So they needed another technical cofounder. Hence her story post.

We were on this call and it became clear at some point that Maddie wasn’t just interested in who I had to refer for the job, but if I would be willing to join the startup. I don’t really know how or when that became the case. It was just a very natural progression from “Let me show you the code we have so far” to me saying “ah okay so here are the next steps,” and then I was in the Ours group chat and we were having regular meetings.

In fairness, I thought for a long time about whether or not to join. But I knew all of them at least to some extent, because our high school was just around four hundred people in total and at that scale how can you not know everyone? I knew Maddie as social and well-liked and very academically and professionally driven. It quickly became clear that Adrienne and Avery were just as competent, despite being two years younger.

The four of us worked well together. Maybe it was because we had the same weird education? Nueva was not a normal school by any means. They taught without tests and with lots of projects, and put a lot of resources towards teaching “design thinking.” They taught us how to do interviews and find underlying problems and how to go to the Stanford d.school to teach grad students. Avery and Adrienne were so on top of interviews. I’m still learning how to be so outgoing.

Founding with two siblings worried me. I wasn’t sure how that’d affect the dynamic, especially around responsibility and decision-making. It turned out to be a fine thing. At least as far as I can tell — I’m sure it must’ve put stress on the two of them, but they didn’t let it show.

So anyways I said yes and joined the group chat and started figuring out how to build an app. I had never actually done that before, save for a little Flask app that I built when I was TAing at Nueva. Cue me digging through Firebase docs furiously trying to figure out document databases and authentication and image storage and oh my god why have we just hit our free rate limit so quickly? (Ah it’s because I mapped before filtering instead of filtering before mapping.) Somehow I got something working together, and we were like okay this is great! Let’s run a test at Columbia and see how it goes!

And it went nowhere. Maddie and Adrienne were actually quite good about getting a decent number of Columbia kids, but the app fizzled almost immediately. You could say that it was too small to get proper network effects going, but I think it was more of a value proposition problem. We knew that you can’t advertise anything as “a way to meet new people!!” because that’s cringe, so we’d gone for errands instead. Because then you can meet new people while just saying “oh it’s convenient and I need to get these things done… might as well go with someone else.” Except it’s not really convenient to have to coordinate with someone else. Especially if you don’t know them already, you don’t know when they’re free, you have to coordinate locations and plans… now you’re texting a stranger about where you’ll meet them in two hours, and wow all you really wanted anyways was to get your dry cleaning done and a bag of Trader Joe’s orange chicken. Why should you wait for someone else? So that didn’t work out.

Nueva’s athletics motto was “We win or we learn.” (Often followed by: ”and we learn a lot.”)

Back to the whiteboard and back to our interviews. It hadn’t been the ideal outcome, but we could say that we learned things: errands are too short term, and the social appeal of meeting new people wasn’t particularly convincing. So we pulled up our interview notes and looked for more pressing needs.

We landed on subletting. I was looking for summer subletters. As was Madeline. As were our friends and classmates. Plenty of people were looking for summer housing, too, because they had summer internships and needed somewhere to live. “Needing somewhere to live” seemed like one of those fairly pressing, non-negotiable needs. We didn’t see any good solutions, either. Finding sublets sucks. Facebook groups are awful. Finding sublets via facebook groups? Surely we could do better. (And wouldn’t it be nice if we could kill Facebook.) So we put together a google sheet and started posting about it and tracked our growth and got like 400 entries on our spreadsheet with a lovely 10–30% week-over-week growth rate.

Took that and applied to a Pear startup competition, because we’d need money to take our startup dreams further, because we figured asking for funding would be a good test, and because Madeline was working at Pear and was organizing the competition. Got through the first round because we presented decently (and also Maddie had good rapport with the interviewer, believe it or not). Then we were there at the final round! But we didn’t have a super clear focus. Our business model was unclear, as was our target market and initial value proposition. We did not win the competition.

At some point in the spring the leadership role shifted from Madeline to myself, just because she had other commitments. Then summer rolled around and we all had internships and I could commit basically no time save for a once-weekly zoom call in which we talked vaguely about directions and experiments we could try. I learned then that you can’t always do everything. By the fall we called it off.

I think we all knew it was for the best. We didn’t have the time, and we didn’t particularly care to fight in the subletting space. Also, Diana Hu at YC had told us about Tarpit Ideas — those startup ideas that seem very promising but are largely not associated with real, solvable problems (and as such do not make for strong startups) — and I think we all grew wary of what we were seeing in the consumer social startup world. To this day, we still use our old group chat to talk about startups we find. Most of the ones we don’t like have pivoted. (Which of course is not statistically significant by any means. It is safe to be a skeptic. Most startups don’t make it.)

Also at some point in the spring, Christian approached me…

The Second Affair: Pallet Materials

The second time I met Christian was when he sent me a text out of the blue, saying that he was starting a startup and would I want to join and build something cool?

(The first time I met Christian was a few days after I arrived at Brown — I was with a group of people looking for the CIT showers, and we ran into him and one of his friends and went to go find the ERC showers instead.)

I was curious, so I suggested we grab lunch and went to go hear about his idea. We went to Andrews and I met him for the second time. (No recollection of the first meeting until later.) He told me about himself and his interests in architecture and sustainability. He told me about the idea of tracking carbon costs incurred during construction. There wasn’t really a clear business model or even value proposition, but he was insanely passionate about tracking embodied carbon. I think it was at this point too that I met Aiganym, Christian’s cofounder and then-girlfriend. She also seemed very excited about the idea (though slightly less talkative) but she was still very bubbly and clearly enthusiastic.

We had a good lunch and Christian sent over a list of pre-co-founding questions to me to judge our compatibility. I had a long look at them and figured they’d be worth talking about, because I had my own qualms — founding with two people who are dating? While I’m already working with Maddie/Avery/Adrienne on another startup?? I told Christian this startup would be my second priority and he was still okay with that. Which in retrospect should’ve presented at least three red flags. (First, it should’ve been very worrying that it was my second priority; second, it should’ve been very worrying that Christian was okay with that; third, it should’ve been another red flag that it was just Christian and not Aiganym-and-Christian making that invitation and that call.)

Christian, for what it’s worth, already seemed very set on me. He said that he’d talked to others, but “my eyes lit up.” (At some point I realized that people with strong personalities tend to like me. I mirror well.) We didn’t even make it through his cofounding questionnaire before we were talking about product. Aiganym and I got lunch to talk about the questions, but instead we just talked over Pokeworks about our lives. I think that made sense for us. It certainly makes sense in retrospect.

So Christian was calling people to have conversations and Christian and Aiganym and I were talking about what product should look like and Christian was sending lots of reading materials and I was doing very little work and Aiganym was left behind. I went to Chicago to work at Jump and Christian kept taking calls and I built out a website and Aiganym tried to build in processes for the team and we never stuck to it. (This was my second grave mistake, after not prioritizing the venture.) By the end of the fall we had several team crises over our dynamic. To the team’s credit, we talked it out and made some changes. But we couldn’t figure out a viable business case for reuse. We called it off before the end of the year.

We weren’t a good team. Both of them are good people and capable teammates, but we weren’t good in combination. There were a lot of reasons. We didn’t have an equal commitment to the vision. I was still hedging my time commitments, because I wasn’t ready to throw my career prospects away for this. We didn’t have a real value proposition. And we didn’t all understand why that was important. I walked away still friends with the two of them, and with the knowledge that I’d do many things differently the next time.

I read a thing once by an author (Paul Graham) who argued that you can only have one thing at the top of your mind at any given time — the thing you think about when you shower and when you dream. I never dreamt about Pallet. I had a few dreams about Ours. (One, I remember, was dream-Mar Hershenson saying that I could build the app in a day. So I did.)

And there would be a next time! Because I figured that if I could still enjoy the prospect of being a founder after all of that, then surely I’d have just as much of a shot as anyone else.

Some Things I’ve Learned

1. A startup needs to make money. A startup should have a real value proposition that can make real money. Don’t build a website before you have a value proposition if that website is not going to help you find a value proposition.
2. Interest and enthusiasm are not replacements for sales. Social benefit does not replace sales. Not even if that social benefit is climate.
3. Climate is hard. Embodied carbon is nice, but I don’t believe there’s a real value prop there yet. More critically: I don’t care enough about the embodied carbon problem to force my way into finding a real viable value proposition.
4. It’s not worth working on a problem that I don’t care the most about. (”Most” relative to all the other problems, and relative to all the other teams.)
5. The whole team needs to care the most about the vision.
6. The whole team needs to be all-in on the venture.
7. The whole team needs to care about including each other. Diversity means nothing without inclusion.
8. Process for the sake of process is counterproductive. A Slack for four people is silly… unless some of the four people need Slack in order to feel included and up-to-date.
9. Process for the sake of the team is essential. (I later read Scaling People by Stripe’s Claire Hughes, in which she says that it’s critical to have Rules of the Game so that everyone knows how to win and not get hurt.) The team needs to communicate how to work together. The game must have rules.
10. Having a team is really nice! I’m working alone now and it’s not easy and less productive.
11. I tend to be biased towards planning and not towards action. (I’ve also learned that I’m prone to shooting my ideas down preemptively. This is something I have to work on.)
12. I like working on social problems. I want to make the next big consumer app. (Nothing about that is easy or trendy, especially right now.)
13. I can only focus on one thing at a time.

I’m currently doing research & experimentation in the consumer social space. I’m not at the point where I’m ready to write about it, but if you’re trying ideas in consumer social please get in touch. I’d love to compare notes — it’s not an easy domain!