Enterprise background
Fifteen years of test results, and what they're worth to a shop with one phone line.
The brands can't all be named. The lessons transfer anyway — and they're the reason this practice charges for a retainer rather than just a build.
What the work actually was
Eight years at Adobe as a Senior Business Consultant, then Director of UX Strategy at Evolv AI. The job was to take a client's website, find where visitors were dropping out, form a real hypothesis about why, and run a controlled experiment to find out whether the fix worked. Then do it again, several hundred times.
It is unglamorous and it is the only honest way to know anything about a website. Everything else is taste with confidence attached.
Four things that turned out to be true everywhere
Most opinions about websites are wrong, including expert ones. A large share of confidently-predicted winners lose when you actually test them. This is the single most useful thing I know, and it's why I'd rather test a change on your site than argue about it.
Removing things beats adding them. The reliable wins are a deleted form field, a removed choice, a shorter path. Almost nobody's first instinct is to take something away.
The headline is doing more work than the design. If it answers the question the visitor arrived with, everything below it gets read. If it doesn't, nothing does.
Speed is a conversion feature, not an engineering one. On a phone, on mobile data, in a parking lot — which is where your customers actually are.
Why this matters more for you, not less
An enterprise site can absorb a bad decision; volume hides it. A local business with a few thousand visitors a month cannot. Every avoidable point of friction is a call that didn't happen, and you feel it in the schedule that week.
The difference at your scale is that the fixes are cheaper, the feedback is faster, and one person can do the whole job.
The framework
The audit I run is a framework I built called CONVERT — a structured pass over a site that produces a prioritised list of changes ranked by expected impact against effort, rather than a hundred-point checklist nobody acts on. It's what the free teardown is a short version of.