My approach changes with the problem. The thinking behind it doesn't.
If you ask me how I design, my first question is: what are we trying to solve? My approach depends on the problem.
I start by understanding what isn't working and why. A heuristic evaluation gives me a baseline. Then I go deeper into how people actually use the product.
Where are people getting stuck? What are they trying to accomplish? What information is missing or hard to find? Does the workflow make sense? Is the architecture supporting the experience, or is the interface compensating for problems underneath it?
The goal is to find the actual problem before designing the solution.
From there, I explore several directions instead of jumping to one answer. Each direction represents a set of assumptions about what might work. I prototype, put those assumptions in front of users, learn what holds up, and refine.
Understand → Explore → Validate → Refine
In 0→1, I spend more time on why the product should exist, what problem it solves, and what foundation it needs to grow.
That work usually starts with discovery: understanding the goals, the people involved, and how the work actually happens today. I dig into the business documentation, including requirements, process flows, policies, and existing data, because the rules of the business often shape the product more than the screens do.
From there, I think in systems. I use Object-Oriented UX (OOUX) to identify the core objects, how they relate, the actions people take on them, and the information each one holds.
And I design outward and inward.
I often visualize an experience before I know every detail behind it. That gives the team something concrete to react to, and it shows me the framework the product needs. Then I work inward, connecting the objects, relationships, data, and workflows that make the experience possible.
Then I move outward again.
That back and forth connects the dots while the product is still taking shape. In 0→1, you rarely know everything upfront, and you don't need to. You need a foundation strong enough to absorb what you learn next.
Discover → Envision → Connect → Learn → Expand
HorseKeeper was built this way. Take away the subject matter, and the system underneath is familiar: profiles, records, events, services, documents, people, roles, transactions, inventory, notifications, expenses, and reporting.
The terminology changes. The patterns don't.
The subject matter may be horses. The product architecture is enterprise.