The loop

01 — Identify

What's actually worth solving

Separating real problems from requests and noise. Most things that come in as feature asks are symptoms.

02 — Evaluate

What holds up under pressure

Pressure-testing from multiple angles — business, technical, operational — before committing direction.

03 — Design

What can work as a system

Not just what solves this problem once, but what solves this class of problems repeatedly.

04 — Deliver

What we can actually ship

Sequenced incrementally so value ships before the full system is built. Delivery thinking starts in discovery.

05 — Validate

What proves itself in reality

Based on real-world impact, not assumptions. The loop restarts from what we learn here.

Product isn't just about getting to delivery — it's about making sure what you're building can survive contact with reality, both technically and operationally.

How I think

Where most product work breaks

The gap between a good idea and a real system. That's where constraints show up, edge cases multiply, and delivery gets messy. That gap is where I spend most of my time.

A feature solves a problem once. A system solves a class of problems repeatedly.

The shift usually happens when edge cases pile up, manual work increases, and behavior becomes unpredictable. That's when you're no longer building features — you're designing a system.

Delivery has to start in discovery

I stopped treating delivery as something that happens after discovery. Because that's usually where good ideas fall apart. A solution can look great on paper — clear problem, strong logic, aligned stakeholders — but if you haven't pressure-tested how it ships, it's still fragile.

Now I pull delivery thinking into discovery early. Not to slow things down, but to avoid designing things that can't survive real-world constraints. In practice this means constantly asking:

  • Could we actually ship this?
  • What breaks when this hits production?
  • What's the smallest version of this that creates value?

I don't memorize — I map

From early math to relational database design, I kept noticing the same thing: everything connects if you're looking for structure. I'm almost always running a parallel thread asking how things connect, where they break, and whether they're efficient. That applies to systems, workflows, art, and everyday decisions.

Product has no single truth

Product isn't an objective discipline — it's a collection of perspectives. Most people are reacting to one slice — discovery, delivery, growth, UX — and presenting it as the whole. That realization changed how I work: I don't follow a single approach, I connect multiple perspectives at once.

Product is like the sun. Everyone is reacting to it, but most people only see one ray.

How this thinking has evolved

Apr 2026
Mapped my process into a loop instead of phases
Started documenting thinking publicly
May 2026
Realized most product advice is perspective-biased
Leaning more into system design vs feature thinking