I have started a podcast. It is called The Build Brief, and episode one answers the thing every client has asked some version of this year: has AI made all of this pointless?

No, but it moved where the value sits, and a lot of people are still standing in the old spot.

What the episode actually argues

AI did two things. It made the generic website nearly free, and it made the generic article nearly worthless. Almost everything else people are worried about is a consequence of those two facts.

That sounds bleak until you follow it through. If the generic got cheap, the specific got rare, and rare is where the money is. This is a better market than it was three years ago if you sell something particular to people with a particular problem, and a much worse one if your offer was "I make websites".

The first forty percent of design is gone, and I do not miss it

Sitemap, wireframes, a type pairing, a colour system, a hero section, the same contact page you have built four hundred times. That is now a prompt and nine seconds of waiting, and it was never the part that decided whether a site worked.

The part that decides is knowing which of the seven things a business does is the one that pays the bills. Knowing the real objection is price, and handling it before it is raised. Knowing the customer is on a phone in a car park with one bar of signal, and deciding what loads first.

None of that is in the prompt. When somebody says "I want it to look modern", that is not a brief. It is an anxiety, and the job is working out what it is about.

The counterintuitive bit about search

This is the part of the episode I would most like people to hear.

A large share of searches now get answered on the results page. The user asks, the engine answers, nobody clicks, so the obvious conclusion is that SEO is finished. For small sites the opposite is true.

Every keyword has a difficulty score from nought to a hundred, and every domain has an authority score on a similar scale. Take "how to install WordPress": roughly twelve thousand US searches a month, difficulty around forty six, page one occupied by WordPress.org, Reddit, Ubuntu, Hostinger and cPanel. Those sit in the eighties and nineties. A newer site might be at six.

Six against ninety is not a ranking strategy. It is a portfolio piece.

The strategy is the terms at difficulty ten or below that the big domains never bothered with, because the volume is too small to interest them. Seventy searches a month, difficulty nine, page one is other small sites. Seventy sounds like nothing until you have sixty of them.

And the head term now gets summarised above the results, while the long tail term is too specific for the summary to be reliable, so the click survives. AI made low competition long tail keywords more valuable, not less. That is the single most useful thing in the episode.

The platform argument, settled

"Which platform is best" is the wrong question, which is why the answers are always rubbish. The right question is who maintains this in two years, and how much do they want to think about it.

WordPress is the one you own, and ownership is a job. Updates, backups, security. Skip it and in eighteen months you have outdated code and a contact form that quietly stopped sending. That failure is close to a genre, and it is why serious WordPress development and SEO work is still the bulk of real freelance demand.

Squarespace is the one that maintains itself, and the tribal claim that it cannot rank is mostly nonsense. I went through the evidence properly in this piece on Squarespace and search, and if you want it handled by somebody who works in it weekly, that is what my Squarespace design and SEO service is for.

The honest summary: the platform is almost never why a site is not ranking. Nobody wrote anything worth ranking. That is also the least profitable thing I can tell anybody, because "migrate your platform" is an invoiceable project and "write forty useful pages" is a slog.

Also in this episode

Why a system that is ninety five percent right becomes useless once you chain five steps together. What a website costs in 2026 and where the money goes. Why redirects are the most commonly skipped item on a redesign and the fastest way to throw away your rankings. And why accountability is the one thing nobody has automated.

Three things to do on Monday

Open your own site on your phone, on mobile data, and try to hire yourself. Find the price. Send a message. Time it.

Check your contact form actually sends, by sending yourself one. Do this first, because if it is broken every other fix is wasted.

Write down the twenty questions people ask you before they buy. That is your content plan for five months and it cost you nothing.

*Adil Makhdoom is a freelance web designer and SEO consultant working with WordPress, Squarespace and Showit.*