What a Google Ads keyword taught us about web design

A few thoughts from Maria this month on what a keyword that has her thinking about the difference between fast and good. Especially in times when everyone is rushed to automate everything.  

This month Ralf had a goal: rebuild our Google Ads strategy from the ground up. Nothing too dramatic on the surface: new keyword groups, tighter targeting and better landing pages. That meant going back to our website, adjusting pages, reviewing our blog posts and the content we had up there. As weeks passed and we saw which keywords performed better, one in particular surprised me. And then it got me thinking about a longer and more important conversation we’ve been having internally, yet is argued worldwide. So, a single keyword kept pulling me into a deeper thought than I anticipated: web design.

It sounds like the most basic keyword a digital agency could bid on. But sitting with it for a few weeks made us ask a question we don’t usually say out loud: in a world where anyone can ask an AI to spin up a website in an afternoon, why does web design still mean anything at all?

The guinea pig problem

Here’s the honest take on what’s happening out there in the trenches. If you open any social feed right now, you’ll find someone promising a website in 10 minutes, or a course on how to get rich building sites with AI. (And don’t even get me started on digital products made with AI.) The pitch is always speed and simplicity, and to be fair, the tools behind it really can produce a website in 10 minutes. What the pitch leaves out is what happens after minute 11, when that business is live with a site nobody who understands the trade ever actually looked at. 

That’s not a knock on AI tools or on interns. It’s a knock on skipping the eye that comes from having done this hundreds of times: the person who has watched which layouts actually convert, which load times actually lose visitors, which “clean and modern” template actually reads as generic and forgettable. Granted, AI is very good at producing a website. Yet, it is not the same as producing your website, built around what your business actually needs it to do.

Without that eye, a website doesn’t get better over time. It just gets iterated on blindly, one guess after another, with the business unknowingly footing the bill for the experiment.

The tool we wish we’d had, and the tool we need now

Here’s the catch, if a fast, AI-assisted website builder had existed back when FutureLab was just getting started, I think Ralf or Tarei would’ve used it without hesitation. My wild guess is that ten years ago speed mattered more than nuance. Everyone needed to be online, without a website you didn’t exist and it was part of the “digital transformation” discourse.  For a digital agency it means you need something live, something that gets you in front of clients while you’re still figuring out who you are. A quick, AI-generated site would have been genuinely useful at that stage, just the right tool for where we were.

That’s not a knock on the tools.

It’s a knock on the idea that speed is the whole job. Actually, Ralf had a good analogy for this: “is a bit like watching a YouTube tutorial on how to change your own oil”. So I’m stealing the example because it’s true. The instructions are available for everyone, anyone can follow them, and plenty of people do it just fine. But the underlying question is whether your time, your own personal safety or any damage to your car if you get it wrong, is worth more than paying someone who does it for a living.

A website in 10 minutes is a template with your logo on it. What’s missing is the eye of someone who has done this hundreds of times. The one who has watched which layouts actually convert and which load times actually lose visitors. Most times “clean and modern” look actually reads as generic and forgettable the moment a real customer lands on it. 

But businesses that have been running for a while are not at the stage of a generic website anymore. They know their customers, they have a brand, a reputation, specific things that convert and specific things that don’t. What they need from a website has gotten more specific, not less. That’s the same shift we’ve seen in almost every company we’ve worked with: the tool that was perfect for year one is not the tool you need in year five. The problem you’re solving has changed shape.

That’s the real answer to the “web design” question. It’s not that AI website tools are bad. It’s that the value of specialised, considered design goes up, not down, as a business matures because the cost of a generic, guessed-at website goes up right along with it.

Will AI make us obsolete, or is it just a tool?

This is the same tension we keep running into, from a different angle every month. Is AI going to replace the work we do, or is it one more instrument in the kit?

My honest take, after watching it play out in our own work: it’s a tool, and a genuinely great one, right up until it’s asked to replace judgment instead of assist it. AI can generate ten layout options in a minute. It cannot tell you which one respects how your specific customers actually make decisions, because it doesn’t know your customers, it reads patterns. That gap is exactly where experience earns its place.

A system should fit the organisation, not reshape it.

This isn’t so different from what we wrote about CARIM last month. The same is true for AI in design work: it should fit around the judgment of someone who knows the craft, not replace the need for that judgment altogether. Used that way, AI doesn’t make the specialist obsolete. It makes the specialist faster, freeing up time to spend on the parts that actually require a trained eye: strategy, structure, the decisions that determine whether a site converts or just exists.

Actually, I can point to exactly where we drew that line this month. As part of the Google Ads rebuild, Ralf set up an n8n automation that checks in on the campaigns every day, pulls the stats, and flags what’s moving and what isn’t. I believe that part is okay to let AI handle, because the daily pulling and flagging is repetitive and mechanical. And honestly, I don’t want to be asked to do it. Yet, someone who understands how Google Ads work still looks at what it flags and decides whether or not it is right. Automatisation doesn’t do our job for us, it just means nobody has to sit staring at a dashboard every morning to start that conversation.  

The campaign was built around the “web design” keyword itself, but we did not hand it over.

The strategy, the targeting, the decision of what that keyword should even mean for us, came from sitting down and thinking it through ourselves (well, Ralf’s). Once it existed, that’s when the automation stepped in to watch it and report back. And maybe to some is an absurdity, but that order matters. AI didn’t build the thing that needed judgment. It’s watching the thing we already used our judgment to build. That’s the whole difference between AI as a shortcut and AI as a tool.

What this means for the people we work with

If you’re a client reading this, this is exactly why we don’t treat your website as a template exercise, and why we push back when a quick fix is tempting but wrong for where your business actually is. We use AI constantly, in research, in drafts, in the n8n automations checking on campaigns while we sleep. What we don’t allow is to make the calls that require the eye:

• what your site should say,
• who it’s actually built for,
• when “fast” is the right answer
• when it’s a shortcut you’ll pay for later.

That’s the keyword that started this, and it’s also the point of it. Web design got me thinking about the ten-minute website because the two have nothing in common except the word “website.” One is a template, the other is a decision made by someone who’s made it enough times to know what’s actually at stake. This is not an anti-AI manifesto, we’re not against the speed AI can provide. I guess the point I’m against is pretending the two problems are the same problem, because frankly they aren’t. And the businesses that understand that difference are the ones building something worth having in five years, not just something live by Friday.