Why SEO is decided before the design, not after

It happens often: the site is ready, the client is happy with the design, and at that point someone asks “shall we do the SEO now?”. It is the same logic as building a house and then wondering where to run the pipes.

It can be done, of course. But it costs more, it delivers less, and it means going back over things that have only just been finished.

What optimising for search engines really means

There is a widespread misunderstanding: that SEO is a set of tweaks to apply to an existing site — a few keywords in the copy, the meta tags, maybe a few links. That is the final part, and the least decisive.

The part that counts concerns structural decisions: how the information is organised, which pages exist and which do not, what the addresses are called, how they link to each other, how fast the site is, how the heading hierarchy is built. These are all choices made before anything is designed, because they determine its shape.

The decisions that come before the design

Which pages should exist

Every page should answer one specific search intent. If you group five different services onto a single page, that page will rank well for none of the five. If you create twenty near-identical ones, they will compete with each other. This is information architecture, and it determines the site map before layout is even discussed.

What the addresses are called

URLs should be decided once. Changing them after launch means managing redirects, losing part of the value accumulated and risking errors. It is one of those things that costs nothing if done at the start and costs time if done later.

How the headings are structured

The heading hierarchy is not a graphic choice: it is the logical structure of the document. One main heading per page, subheadings following a sensible order. If the design is created without that logic, you either rewrite the design or accept a confused structure.

How much room the content has

Many elegant layouts are designed for pages with little text. Then the need arises to explain something in depth and there is nowhere to put it, or it gets squeezed into a place that was never intended. Knowing in advance which pages will have to carry long content changes the way you design them.

The hidden cost of adding it later

Retrofitting almost always means going back over three things at once: the page structure, the copy and the code. Meanwhile the site is live and accumulating signals on a structure you are about to change.

SEO done afterwards is not the same SEO done beforehand, only later. It is a different job, more expensive, with a poorer result.

How I handle it on my projects

Optimisation belongs to the design stage, not the build stage. By the time I draw the structure of a site I have already decided which pages exist and why, which search intents they cover, how they link together. The design is built on top of that structure, not before it.

The build stage adds the rest: technical SEO on speed, code, URLs and structured data; on-page optimisation on headings, copy and images; configuration of the measurement tools. That is the visible part, but it rests on decisions already taken.

The practical advantage is that there is no moment at which “the SEO gets done” and no separate cost for doing it: it is part of the project, just as accessibility or speed are.

If your site is already live

Not all is lost, and you do not necessarily need to rebuild everything. The starting point is an audit that says what is working, what is broken and what is best left alone: sometimes the structure is sound and targeted work is enough, other times the problem is further upstream and carrying on tweaking is simply spending at a loss.

What makes the difference is understanding which of the two cases you are in, and that means looking at the data before proposing solutions. If you like, write to me and we can talk it through.