Building a product from zero is different from every other PM job. Not harder, exactly — different. The skills you need in a 0→1 environment aren't the same skills you need when you're optimizing a mature product. And if you don't recognize that difference, you'll bring the wrong tools.
I've taken AvaSense — an AI-driven contract and supplier governance platform at Avasant — from a blank Figma file to a live SaaS product with six enterprise clients and $10M+ ARR. Here's what I learned.
Conviction is the most important early-stage product skill.
There are no metrics when you're at zero. No A/B tests. No cohort data. What you have is a hypothesis about a problem, and a set of signals from discovery research. The job is to hold that hypothesis with enough conviction to make decisions — while staying genuinely open to being wrong. The PMs I've seen fail at 0→1 were either too attached to their initial concept, or so uncertain they couldn't make a call.
Discovery is never done. It just changes form.
Before launch, discovery means understanding the problem. After launch, it means understanding *your* product in the problem space. The 20+ client interviews and 5 SME sessions I ran before writing the first PRD were not a phase — they were the start of an ongoing practice. The form changed: in-app analytics, support patterns, renewal conversations, executive demos. But the listening never stopped.
Enterprise procurement is a product problem.
Your product doesn't just need to work. It needs to survive procurement. That means audit trails, role-based access, data residency controls, and documentation that answers security review questions before they're asked. I learned early to design for the security review, not just the user. Both have to win for you to close.
Velocity compounds.
The 30% improvement in delivery velocity we drove through Agile/OKR frameworks and standardized SOPs wasn't just an operational win. It changed what the team believed was possible. Speed is a morale strategy as much as an execution strategy.