I’m often asked how elite teams compress the journey to product-market fit. One story I keep returning to is Jessica McKellar, co-founder and CTO of Pilot, which is the largest accounting firm for startups. For the past six years, she’s built Pilot alongside her two co-founders, Waseem Daher and Jeff Arnold — and what makes this trio extraordinary is that they’ve stuck together across three startups.
As repeat founders, the team learned a ton from their first two ventures, K Splice and Zulip, and both netted some positive outcomes. Yet there were mistakes that prevented those products from becoming an outsized success. From a product management leadership perspective, I see a clear evolution in how they approached problem selection, product discovery, and go-to-market.
With Pilot, they prioritized picking an acute problem and a huge market to tackle. That simple but rigorous reframing matters: identify a customer segment with a painful, high-frequency workflow; quantify the market; and ensure a compelling “why now.” This is classic founder-led GTM discipline and the essence of practical product-market fit lessons.
They also embraced a deliberately tedious build process for v1: looking over Waseem and Jeff’s shoulders as they manually did the bookkeeping for early customers, while she wrote code alongside them. In my experience, this “do the job, then automate” approach functions like forward deployed engineers for founders — embed with the real workflow, capture edge cases, then translate that knowledge into the system of record.
Even going back to the earliest days, Pilot had some really strong product-market fit signals, with customers agreeing to pull out their credit card and pay for the product right away when it was just an idea on paper and eventually pulling the Pilot team into expanding their product suite. That willingness-to-pay signal, coupled with pull-based requests for adjacent capabilities, is exactly what I look for before scaling zero to one B2B marketing or hiring beyond the founding team.
My playbook from this story is straightforward: choose a narrowly defined, acute pain in a massive category; run founder-led discovery inside the customer’s workflow; ship code alongside the service until the workflow is reliable; price early to validate value; and align outcomes vs output OKRs so the team optimizes for customer impact, not feature volume. Do this, and you convert messy service learnings into a repeatable product engine.
Make no mistake about it — being a founder is incredibly difficult — but choosing the right problem to tackle can drastically smooth the path ahead of you. For product creators, that choice — and the discipline to live in the customer’s workflow early — is the difference between meandering and momentum.
Leave a Reply