Almost every redesign begins the same way. Someone opens Pinterest, pulls together a moodboard, screenshots three competitors, and forwards the lot to a designer with a note that says "something like this, but for us." The new site ships six weeks later. It looks better. It performs about the same.
This is the part nobody wants to say out loud: aesthetics-first redesigns reproduce strategy mistakes in a new colour palette. If the old site did not convert, the prettier version will not either. The problem was never the font.
A redesign should start with evidence. The moodboard comes much later — after the numbers have already told you what the site is failing to do.
Your analytics already know what you do not
Before you choose a typeface, open your analytics. The data is rarely flattering, but it is honest. Which pages do people actually land on? Where do they drop off? How far down the homepage does the average visitor scroll before leaving? What is the bounce rate on the services page versus the about page?
In most small business sites, two things become obvious within an hour. The pages you spent the most time on are the ones nobody reads. And the moment of drop-off is almost always the same — a tired hero, a vague headline, or a contact form that asks for too much.
You cannot redesign around problems you have not named.
The inbox is qualitative data
The other half of the evidence lives in your inbox. Every "do you also do…" email is a signal that your services page is unclear. Every "how much does it cost?" is a signal that you are not framing value early enough. Every "can we hop on a quick call?" from someone who is plainly not a fit is a signal that you are attracting the wrong audience.
Read the last fifty enquiries before you brief a designer. The patterns will tell you exactly which beliefs the new site needs to build, and which wrong-fit conversations it needs to filter out.
Before-and-after only matters if you measured the before
The most common regret after a redesign is that nobody can tell whether it worked. The site looks newer. The team feels better. But leads are flat, and there is no baseline to compare against, so nobody is sure.
Before you touch the design, write down the numbers you want to move. Monthly enquiries. Conversion rate on the contact page. Average time to first message. Quality of leads on a one-to-five scale. If you cannot say what success looks like in numbers, the redesign is a vanity project — even if the result is genuinely beautiful.
Competitor moodboards average you into your category
Reference-driven design has a subtle failure mode. If you pull from ten competitors, you end up looking like the average of those ten competitors. The new site is modern, but it is also indistinguishable. You have spent six weeks and a real budget to blend in more cleanly.
The businesses that win on the web do not look like their category. They look like themselves. The way to get there is not to study competitors harder — it is to study your own customers, your own enquiries, and your own positioning until the brief is specific enough that no competitor's site could have produced it.
The brief writes itself once you have looked at the numbers
The reason most design briefs are vague is that they were written before the diagnostic work. "Make it more modern" is what you write when you have not looked at the data. "The services page has a forty-eight per cent drop-off in the first scroll, our top three enquiries are from the wrong budget range, and the headline does not mention the outcome our best clients actually buy" is what you write after you have.
The second brief produces a redesign that moves numbers. The first produces a moodboard.
The one case where the common advice holds
There is a real exception. If you are launching a brand-new business with no traffic, no inbox, and no customers yet, there are no analytics to read. In that case, references and moodboards are the right starting point — not because aesthetics come first, but because direction has to come from somewhere. The moment you have ninety days of real visitors, the rules change.
The better rule of thumb
Spend the first week of any redesign in your analytics and your inbox, not on Pinterest. Write down what the data says, what the enquiries say, and which specific numbers the new site has to move. Only then open the moodboard. Aesthetics done on top of evidence are powerful. Aesthetics done in place of evidence are expensive decoration.
If your current site is due for a redesign and you are not sure where to start, send me the analytics and the last month of enquiries — that is where the real brief lives.