Questions

The questions you
actually want to ask.

Not an FAQ. The things you'd ask in the first coffee, answered before the coffee.

    1. Goal first. I research until I can explain the problem in one sentence: who visits, and what we want them to do. I talk to whoever knows the users best: a founder, a teacher, real users.
    2. Structure before screens. I map the pages and what each one needs to say, and agree on them with the team before any UI.
    3. Feel it, then plan it. I build a mood board in Figma until I can see the look, then plan the story: what people see first, and what comes next.
    4. Design and build, in the right order. Sometimes I prototype first in Claude and detail the pages in Figma later. Sometimes I design the detailed UI in Figma first and build after. Either way, every page works on phone and desktop, and ends up documented in Figma for the team.
    5. Ship small, learn, repeat. I launch, watch where people get stuck or leave, and fix it in the next version.
    1. I own the problem and the solution. I decide what we're solving and how, and AI helps me build it faster and test more ideas.
    2. Claude first. Most of my work now happens in Claude, because my ideas become real, working pages and prototypes in minutes. I still use Figma when a task needs it.
    3. Faster, clearer decisions. When people can click and try something real, feedback comes faster and arguments end sooner. I review every result before anyone sees it.
    4. Speed changes the work. When a new idea costs minutes instead of days, you explore more before you choose.
    1. Use it like a real visitor. Before touching anything, I use it myself on a phone and a laptop, watch how other people use it, and read the complaints.
    2. Find the weak points. Where do people get stuck? Which steps can we remove? That becomes the brief.
    3. Same problem, new solution. I don't just repaint the old pages. I look for a better way to solve the problem, even if it means a completely different structure.
    4. Ship in small versions. Agree on what "better" means up front, then release step by step and measure it against the old version.
    1. Take the note, not the offence.
    2. Ask what problem they see before we argue about the fix. Most feedback is a symptom.
    3. Say "you're right" when they are. Push back with evidence, not taste.
    1. A page or feature with no clear problem behind it. If we can't say who it's for and what it fixes, we're not ready to build it.
    2. Anything that tricks the user: fake urgency, hidden costs, a cancel button that hides.
    3. Building everything at once. I'd rather ship an MVP, test it, and learn, than spend months on something nobody wanted.
    4. Extra steps and decoration that gets in the way. Every extra click or tap is a reason to leave. Function first, then beauty.
    1. In from day one. Engineers see the scope and flows early. Technical limits are design limits, and I'd rather know them on day one.
    2. Nobody waits on design. Once the flows are confirmed, backend developers start on the database while I design the UI, and frontend developers see early drafts. An engineer waiting on a designer is the most expensive silence in the building.
    3. Prototypes as specs. My clickable prototypes show exactly how things should feel, and I write down the states and edge cases they can't show.
    4. Finish together. I answer questions fast, check the build against the design, and fix the last details with the developers.
    1. Over-polishing. I can spend too long on the last details. MVPs help me ship first and polish after launch.
    2. Trying every new tool. I love testing new AI tools, which can pull me away from the work in front of me. Now I try new ones only on side projects.
    3. Writing code by hand. I build real products with AI, not as a trained developer. I'm learning the basics so I can review what AI writes.
    1. Meet everyone who touches the product, and the people who answer its emails.
    2. Use the website and product as a new visitor would, and write down every place I get confused.
    3. Pick one small, real fix and ship it, so we've worked together once before anything big.
    4. Agree on what "better" means, so by week four we can tell.