The idea of a single "eureka" moment is misleading. True insight comes from deep immersion in a problem space over time. Eventually, you gain so much context that a better way of operating seems obvious, not like a sudden stroke of genius.
Quanta's engineers performed manual bookkeeping, a practice they called "engineers as bookkeepers." This forced immersion into the domain's deep complexities and edge cases, leading to a far more robust and effective automation product than if they had worked from a spec sheet.
Early-stage companies don't want to buy another piece of software; they want a problem solved. Quanta succeeded by providing a complete accounting service ("the work to be done"), which is what customers truly valued, using that as the wedge to build its underlying automation platform.
Before their product was ready, Quanta partnered with an outsourced accounting firm to service its first design partners. This allowed them to immediately start selling, charging customers, and learning the operational complexities of the service, de-risking the business while building their own technology.
Finding PMF is like pushing a heavy boulder uphill. True PMF is the reverse: the boulder is rolling downhill, and you're chasing it. Demand outstrips your capacity, customers stick with you despite imperfections, and the momentum feels like it's pulling you forward.
When building an AI-enabled service for a mature market like accounting, customer demand is a given. The core business risk shifts entirely from sales and marketing to engineering. The key question becomes: can you automate enough of the manual service delivery to achieve venture-scale gross margins?
When conducting user research, begin with high-level, open-ended questions about the user's biggest pain points. Introducing your specific idea too early will bias their feedback, as people naturally want to be helpful and will focus on your concept, even if it's not a real problem for them.
Unlike pure SaaS, an AI-enabled service has a manual component that can be overwhelmed by demand. Quanta had to pause onboarding new customers because saying "yes" to too many slowed down engineering and hurt service quality. Throttling growth is critical to long-term success.
Instead of engineering complex solutions for every possible edge case upfront, Quanta's team wrote code that would simply ping a human on Slack when a rare event occurred. This "human-in-the-loop" approach is a massive mindset shift that allows for much faster initial product development.
Quanta's founder waited nearly two years to announce her seed round, timing it with the product's public launch. This strategy bundles a funding announcement (which is easier to get press for) with a product launch, creating a single, more powerful PR moment that drives user sign-ups.