logs

Most Orchard work will be put on pause. We have shifted our startup’s fundamental philosophy to instead pick a specific customer and solve their most pressing problems rather than trying to work with AI to build a one-size-fits-all solution in hopes of chancing upon PMF. This way, even if we are unsuccessful in a product or two, we should come out of the experience having much more fundamental learnings about a customer segment, which will be a nontrivial knowledge moat.

Day 378: Monday, May 1, 2023

It turns May today, which is a sobering realization as we’ve been working on our startup for quite some time now. While we’ve improved our processes quite a bit, it still feels like we’re back at square one in some aspects—before starting the entire startup journey, I (optimistically) would have thought that we’d be much further along in the PMF journey.

We have roughly 70 weeks of runway left at this point—to hammer this point home, we’re going to assign every single week a distinct objective that’ll move us closer to finding PMF. Our goal for this week is to identify a customer that is presumably worth us sinking 35 weeks into learning about.

Day 379: Tuesday, May 2, 2023

John brought up a new thread/space in supply chain. It’s honestly a cool space with lots of legacy incumbents—maps of supply chain networks kinda look like Starcraft, which I can definitely get behind in terms of idea excitement.

Day 380: Wednesday, May 3, 2023

Honestly, slower day for me personally. I have trouble doing nontechnical, unstructured work for long hours (versus coding, where it’s easy to just get lost in the task at hand).

Kevin has been working on scraping SEC data as a side thing. Tae is exploring edtech as a whole today. John has been looking into supply chain stuff.

Day 385: Monday, May 8, 2023

I forgot to write logs for a few days in a row, but the primary objective for the last week was to settle on a specific vertical to focus the remainder of our startup runway on. The idea is that we want to narrow our focus as staying in the realm of all possible ideas ends up in a lot of wandering and little progress. The hope is that picking a specific vertical will cause us to build up customer insights over time such that even if our initial idea fails, we’ll have enough to go off of to maneuver to new pivots within the same customer segment/vertical.

Today, we spent most of the day discussing as a team all the possible verticals we’ve explored over the past week. We created initial hypotheses for all of them, and listed out pros and cons. This meeting took several hours, but at the end of it, it felt like we all had a pretty good grasp of which verticals were most promising and which ones we should table for the sake of focus.

We ultimately didn’t commit to a vertical by the end of the day like we originally planned, but this is a rather big decision. We’re hoping to reach a conclusive commitment by tomorrow.

Day 386: Tuesday, May 9, 2023

We have narrowed it down to two verticals: corporate intelligence and education. For corporate intelligence/financial services, there are a lot of tools that we can quickly build out and sell to contacts at various firms—this is the major pro, but it’s hard to envision exactly what the bigger picture looks like after these initial small sales. But perhaps that’s besides the point. For education, we have strong conviction that AI will fundamentally change learning, and that’s an exciting concept for us. However, we’d probably be building out a consumer app in the beginning, and there’s all the execution risk that comes along with that—it’s a slow to fail idea, and it’ll be hard for us to get any early revenue.