Six things you can do right now to audit, understand, and improve your own product
Most founders I talk to know something is off with their product.
The UI feels inconsistent. Users take longer to onboard than they used to. A feature that seemed obvious to build isn’t being used. But without a designer on the team, it’s hard to know where to start — or whether what you’re seeing is actually a problem, or just noise.
Here’s the thing: the first things a designer does when joining a new team aren’t magic. They’re structured observation. And with the right questions, you can do most of it yourself.
1. Stop looking at your product like someone who built it.
The hardest part of auditing your own product is that you know it too well.
You know where everything is. You know why every decision was made. You’ve long stopped seeing the rough edges because you’ve internalized them.
So before anything else: find someone who hasn’t used your product before. A friend, a colleague from a different field, someone in your network. Ask them to share their screen and try to complete one core task — without you explaining anything. Don’t help. Just watch.
Notice where they pause. Where they click the wrong thing. Where they ask “wait, what does this do?” That’s your first audit.
2. Map the flows — starting with what the user is trying to achieve.
Designers think in goals, not in features. That’s the shift worth making.
Take your three most important user tasks — the things your users need to do regularly to get value from your product. Before you trace a single screen, ask: what is the user actually trying to accomplish? Not “click here to generate a report” — but “understand whether my team is on track before Monday’s meeting.”
That goal is the lens. Now trace each flow, screen by screen, click by click, through that lens.
Ask yourself at each step: does this screen move the user closer to their goal? Is it clear what they should do next? Could they arrive here from somewhere else and lose the thread entirely?
I’ve seen products where every individual screen looked reasonable, but the end-to-end flow had a gap that nobody noticed because everyone was looking at features in isolation. Mapping from the user’s goal reveals what the feature view hides.
3. Do a consistency pass — it builds more than aesthetics.
Inconsistency isn’t just a visual problem. It quietly erodes trust.
When a product feels consistent — same button logic, same spacing rhythm, same tone in the copy, same way of handling errors — users don’t notice it consciously. But they feel it. It signals that someone thought carefully about this product. That it was made, not just built.
Go through your product looking for one thing: does it feel like the same product everywhere? Note the moments where something feels slightly foreign — built at a different time, by someone with a different mental model. Those are usually where the debt is, and where trust starts to slip without anyone being able to name why.
4. Read your support tickets as a UX document.
Your support queue is the most honest feedback you have.
Pull the last 30 support tickets or user questions. Group them by theme. If three or more people asked the same question, that question shouldn’t exist — the product should answer it.
“How do I find X” is a navigation problem. “Why didn’t Y work” is a flow problem. “What does Z mean” is a copy or onboarding problem. Each cluster points to something specific.
5. Talk to a user who almost churned.
Power users will tell you what they love. Almost-churned users will tell you what’s broken.
Find one or two users who reduced usage, downgraded, or needed significant support to stay. Ask them to walk you through a recent frustrating experience. Don’t defend the product. Don’t explain why it works the way it does. Just listen.
One of these conversations surfaces more actionable insight than ten happy-customer calls.
6. Prioritize by friction, not by appearance.
The goal isn’t to make the product look better. It’s to make it work better.
After you’ve done the steps above, you’ll have a list. Resist the urge to fix the things that look worst first. Instead, ask: what’s causing the most friction for the most users? What’s slowing down deals? What keeps coming up in onboarding calls?
Fix those first. The rest can wait.
None of this requires a designer. It requires structured attention — the willingness to look at your product through someone else’s eyes, follow a real user’s path, and sit with what you find.
Most founders who go through this come out with a clearer picture of where the real problems are — and a much better brief for when they do bring someone in.
Lia Dolev is a senior UX/UI designer with 20 years of experience in complex B2B products. She works with B2B startups that have outgrown their early design decisions. → liadolev.studio