SEO · Strategy

Why SEO matters more now than it did five years ago, not less

AI answers were supposed to end search. They are built out of the same web, which makes the groundwork more valuable rather than less.

We get asked a version of this most months now. Assistants answer questions directly, people click through less, so is there any point still doing SEO. It is a fair question and the honest answer is not the reassuring one an agency is supposed to give.

Some things about search genuinely did get worse for businesses. Fewer clicks reach you for questions that can be answered in a sentence. If your traffic was built on definitions and simple how-tos, that traffic is going and it is not coming back.

But the work underneath is worth more than it was, not less, and the reason is mechanical rather than hopeful.

Assistants are reading the same web

An assistant naming three letting agents did not invent them. It read the web — your pages, your reviews, directories, press, forums — and assembled an answer. Everything that decides whether a crawler can find you, understand you and trust you decides whether you appear in that answer too.

So the list has not changed much:

  • Pages that can actually be reached and indexed. A stray noindex or a blocked section makes you invisible to both.
  • Pages that load. A category page that takes nine seconds to become useful on a mid-range phone is a page most people never see.
  • A consistent, verifiable identity — the same name, address and details wherever you appear. A system trying not to be wrong has a reason to name someone else when three directories disagree about your phone number.
  • Being mentioned somewhere other than your own site. This was always the hard half and it is now the bigger half.

None of that is new. What is new is that skipping it now costs you twice.

What did change, and it is a real change

The unit is smaller. A search engine ranks pages; an assistant lifts passages. A ten-thousand-word guide that buries the answer to one question in the middle is worse than a short page that answers it plainly near the top.

And the question is longer. Not "letting agent Manchester" but "I own a flat in Manchester and my tenant is leaving, what do I need to do before re-letting". Pages written for the two-word version do not match the twenty-word one, and most service pages on most sites are written for the two-word version.

Why the groundwork got more valuable

Three reasons, and they compound.

There is no second page

In a list of ten results, position six still earned something. In a generated answer naming two or three businesses, position six earns nothing. The gap between being included and being nearly included got much wider, which raises the value of everything that decides inclusion.

Specifics survive summarisation and adjectives do not

"We handle around two hundred tenancies across Greater Manchester" can be carried into an answer. "Leading provider of premium property solutions" cannot, because there is nothing in it to restate. Sites built entirely out of adjectives were always weak. They are now unusable as source material.

Competitors are quietly giving up

A fair number of businesses read the headlines about AI killing search and stopped. In competitive local categories we are watching sites stand still. That is an opening, and openings in search close slowly but they do close.

What we would not spend money on

Being straight about this matters more than the rest of the post.

  • Content written to catch questions an assistant now answers in one line. That traffic is gone and no amount of word count brings it back.
  • Crawl budget work on a site of a few hundred pages. It gets raised constantly and it is never the reason for anything.
  • Chasing a perfect performance score while a genuinely slow page elsewhere goes untouched.
  • Any fixed monthly report that is the deliverable rather than a description of work that happened.

So: is it worth starting now?

If you want enquiries this month, no — paid search does that, and we would say so. SEO is slow and it is honest about being slow: in a competitive local category, expect profile and citation work in month one and meaningful movement somewhere between month four and month six.

What it buys is a position that does not switch off when the budget does. That was true before assistants and it is more true now, because the same groundwork is feeding two doors instead of one.

More on SEO