Not every redesign needs to start with a redesign.
Quite often, clients come to me with an existing product, MVP, or website and pretty much the same concern:
“We know something isn’t working, but we’re not exactly sure what”
And in cases like this, I almost always suggest starting with an audit.
It doesn’t have to be a huge 30-page document. It really depends on what the product needs. Sometimes it’s a full product or website audit. Sometimes it’s a quick landing page review, a deep dive into a specific flow, or simply looking at where users are dropping off and why.
What I like about starting with an audit is that it quickly turns “we need to improve something” into a clear list of actual problems, hypotheses, and next steps.
From there, everything gets easier. We know what’s actually worth changing, what the scope looks like, and what should come first. And if there’s a lot to improve, we can prioritize it and break the work into manageable iterations.
It also means much less back and forth once we get into design, because we’re not trying to figure out the core problems along the way.
So if you already have a working product or website but aren’t quite sure where to start improving it, you might not need a redesign just yet.
You might need a good audit first.
And if you need a fresh pair of eyes on your product, website, or a specific flow - feel free to DM me!
What sells it for me is the refusal telling the agent exactly where the token should go. A bare "no hex" error usually just sends it looping. Does guard.mjs run only as an agent hook, or in pre-commit too so it catches humans as well?