Last week I spoke at @ycombinator DevTool Day. The focus of my talk was a summary of the merits and demerits of selling to startups. A lot of people don't realize that in the beginning, Porter never intended to have startups as users. After years of pivoting we initially built a product designed to migrate Enterprise-tier PaaS users to the hyperscalers. I remember actively turning away early-stage companies looking to use us because they seemed too small to justify the migration effort. Eventually, a couple startups hell-bent on using us managed to self-serve even though we had no documentation and had exclusively built our "onboarding" flow as an internal tool to process white-glove migrations. We let those startups stay on the platform expecting nothing to come of it. Luckily for us, we had some rocket ships in that initial sample. A few of those users rapidly grew to a significant usage scale (with individual clusters on the order of terabytes of RAM and hundreds of nodes) within months of onboarding. This was larger and higher-value than most of the customers we were running through a multi-month sales process. That got our attention. Even still, we were initially pretty reluctant to bet the farm on startups. As any VC knows, all the growth is overwhelmingly concentrated into a tiny subset of companies. This gets to a key downside of pursuing this strategy: unless you have a product that an overwhelming percentage of startups can use (like payment processing, HR, or corporate card), you run the risk of not capturing enough mindshare to effectively guarantee there are some winners in your basket. If your carrying capacity for pre-seed startup adoption is something like 15% of all companies, you could struggle to get off the ground unless you're somehow indexing for only the best of the best. For devtools in particular, you're almost always better off going truly bottoms-up with indie devs if your product allows it. In retrospect, I believe we were fortunate that two things were the case: (1) frustration with DevOps overhead (i.e. having to manage AWS/GCP/Azure) is essentially ubiquitous, and (2) cloud infrastructure is a higher-order expense for virtually all startups (probably the 2nd or 3rd largest line item after payroll and tokens). The upshot is also massive. Winning startup mindshare means you are building for the next generation of public companies. If you can keep up with your best customers and preserve sufficient pricing power, you have the ultimate NRR engine. As a devtool company, there's also nothing better than getting to work with smart, high-growth teams at the bleeding edge of technology.

After raising a $20M Series A in 11 days, some founders have reached out asking, “Justin, how did you raise so quickly?” The truth is, we weren't even looking to raise. About a year and a half ago, David from FirstMark reached out. I'm someone who generally hates talking to investors, so my instinct was to ignore it. But David kept showing up. Instead of sending generic "let's get coffee" messages, he invited me to in-person events where I could meet people doing interesting work. Through FirstMark's programming, I got to chat with the CTO of Duolingo, the founder of DataDog, and countless others. It was mindblowing how much David pushed us forward. And at that point, I wasn't even in FirstMark’s portfolio yet. They were just going out on a limb to build trust. Fast forward to last fall, and we were hiring a lot against revenue growth. We didn't need the money. But the relationship with FirstMark had gotten so strong that when David leaned in, we decided to run a lightweight process. The timeline: 1. Told David we'd consider raising 2. Pitched the partnership a few days later 3. Signed the term sheet within 10 days What made it fast wasn't some secret fundraising hack; it was a year of relationship-building that happened before we ever decided to raise. Whenever another investor came with a term sheet, FirstMark just kept matching or beating. They’d already done the work to understand our business, developed conviction, and were leaning in hard. So, it is possible to raise a Series A in 11 days.. ...if you've spent a long time building relationships with people who you know, believe, and trust. Speed in fundraising isn't about cramming the process or doing crazy FOMO. It's about doing the real work slowly so the last part's fast.