How to Approach a Website Renewal: Decide What Stays Before You Rebuild

It looks dated. It breaks on a phone. Nobody can update it. Those are real problems, and they are usually what starts the conversation. But start there and the brief quietly becomes "make it newer" — and newness cannot be the goal. Whatever you choose for being new will be old again in three years.

Sites get rebuilt and the enquiries do not move. Most of the time, something that should have been kept was thrown out with everything else.

"Make it new" is the goal that strands you halfway

Set the purpose on the other side of the screen, not on your own. Who has to believe you, what they need to believe, and in what order. Write that as one line and most later decisions make themselves.

When the goal is newness, the test becomes whether something looks good — a test with a different answer for every person in the room. Opinions split, sign-off stalls, compromises stack up, and what ships is nobody's taste.

Name the person you want and you always have somewhere to return to. Would this photograph reassure them? Is this a word they actually use? Would they read this far in this order? Decisions get faster, and they hold even when the people making them change.

One line is enough. Long planning documents are re-read only by the person who wrote them.

The first step is counting what already works

Look at your current site as an ally for once. It is almost never all bad.

Four things are worth reading. Which pages are actually read. Where the people who enquire came from. Which words bring anyone in from search. And what real clients have told you they saw. Use numbers where numbers exist and memory where they do not.

Whatever surfaces there is what stays. An article people finish, one photograph they remember you by, a sentence that gets quoted back to you on the phone. It may look modest. It is the only equity the current site has built.

The most common accident in a rebuild is deleting that equity by moving its address. Change where a page lives and you cut the search history it earned and every link pointing at it. The site looks better and reaches fewer people. Deciding what stays is not sentiment; it is the step that prevents this.

Cut the exceptions, not the pages

When people reduce, they count pages. Page count is rarely the weight. The weight is the number of exceptions.

Any site that has been run for a few years has them. One page where the headings are larger. One notice in red. One index sorted differently from the rest. Each was the right call on the day. Ten of them together tell a visitor that nothing here is governed by a standard, and that reading is where most of the cheapness comes from — not from decoration.

So subtract inconsistency, not content. Heading sizes, the distance photographs are taken from, the amount of air, the wording on buttons. Even with nothing removed, aligning those makes a site quieter, and the quiet is what reads as expensive.

Then retire the pages whose work is done. Retire them with somewhere to go, not into nothing. How carefully you handle that decides what the months after launch feel like.

When to do it: not when it looks old, but when what you say has changed

You do not need a fixed number of years. Looking dated is a weak reason on its own.

There are three real signals. What the business now says no longer matches what the site says. The client you want has changed. And nobody inside the company wants to show it to anyone. The third one never appears in analytics and is the most honest of the three.

If what you say has not changed, you probably do not need a rebuild. New photographs, a different order, shorter sentences — the difference that makes is larger than people expect. Rebuilding is one option, not the first one.

And if you can choose the timing, avoid the run-up to your busiest season. The first month after launch always produces small corrections. Publish when you have no time for them and they stay uncorrected for years.

For a corporate site, the order is the same

More people have a stake in a company site. Every department has something it wants included, and including all of it produces a page addressed to no one.

The sequence does not change. One line on who has to believe you, a count of what works, then align the exceptions. What grows is not the process but the cost of agreement — which is exactly why the line is worth writing down first. In a room, what carries weight is not volume; it is having decided in advance.

When opinions split, do not argue taste. Return the question to which version is clearer for the person you named, and it usually resolves.

Decide three things before you brief anyone

You do not need an exhaustive specification. Three things are enough.

One: who you want, and the last action you want from them. Two: the pages that must survive, and the words on them. Three: who is producing the photography and the copy. The third is the one that stalls projects — design finished, content empty, months passing.

Colour, type and motion are better left open. They are answers derived from the purpose, not conditions set before it. Handed over as conditions, they only remove options.

Settle those three and your quotes get sharper too. A brief that is clear about what is being handed over does not come back padded with contingency.

It is only a question of order

So the whole approach is this. Decide in one line who has to believe you. Count what works and keep it. Align the exceptions. Retire what is finished, with somewhere for it to go. Building comes after that.

Follow that order and a renewal stops being a rebuild and becomes a continuation. A site that inherits what was earned is strong from the day it launches.

New is not a goal.

You can start building once you know what stays.