This question usually arrives disguised as a website question. Someone doesn't like their site, they start planning a redesign, and somewhere in the first conversation it becomes clear the site isn't really the problem.
Sometimes it is. Here's how I tell the difference.
The symptom test
Ask what specifically bothers you about the current site.
If the answers are structural — it's hard to update, it's slow, people can't find the contact page, it looks broken on a phone — that's a website problem. Your identity may be perfectly serviceable. Rebuild the site and keep the brand.
If the answers are about feel — it looks dated, it doesn't reflect who we are now, it makes us look smaller than we are — the site is the symptom. You can redesign it, but you'll be dressing up an identity that's already fighting you, and you'll be back in two years.
What a brand has to provide for a website to work
You don't need a full identity system to redesign a website. You need enough decisions already made that the site isn't inventing them on the fly:
- A logo that works small, on dark and light, without modification
- A defined color set, including something that works as an accent and something that works across large fields
- Typefaces chosen for headlines and body text, licensed for web use
- A clear sense of tone — how you sound in a sentence
- Photography that exists, or a plan for making it
With those, a redesign has something to build on. With a logo and nothing else, the website project ends up doing brand work anyway — informally, mid-build, without anyone calling it that. That's the expensive version.
When brand should come first
I'd do identity work first when:
- The business has meaningfully changed what it does or who it serves
- The current identity was never designed — it was assembled, inherited, or made in a hurry
- You're merging, renaming, or repositioning
- Different materials look like they belong to different companies
- You can't articulate what makes you different, which is a strategy problem wearing a design costume
That last one matters most. A website can present a point of view. It can't manufacture one.
When it shouldn't
Plenty of businesses have a perfectly good identity and a bad website. If your brand still fits, redesigning it is a distraction — an expensive, disruptive one that resets recognition you've already paid to build.
Old is not the same as dated. A mark that's been in market for fifteen years carries recognition a fresher one won't have for a long time. Refinement usually beats replacement: tighten the logo, resolve the palette, choose better type, define the rules — and keep the equity.
The middle path, which is where most projects land
In practice most projects aren't purely one or the other. They're a website build with a light identity pass folded in: the existing mark kept and cleaned up, the color set properly defined, type chosen deliberately, and the whole thing documented well enough that the next person can stay consistent.
That's often the right scope. It respects what's working, fixes what isn't, and doesn't turn a website project into an eighteen-month rebrand.
How to decide in one question
Imagine the site rebuilt exactly as you'd want it — good structure, fast, well written, well photographed — using your current logo, colors, and type.
If you picture that and feel good, you have a website project.
If you picture it and something still feels wrong, you have a brand project, and the website is just the thing that finally made it visible.