I recently posted on LinkedIn about the pressure AI is putting on web design agencies. It generated a lot of discussion, and some of the responses were more interesting than the original post. So I wanted to go deeper on this, because the underlying issue matters well beyond web design.
Let’s start with what prompted it.
The £299 WordPress site is already under pressure
There are web design agencies across the UK right now selling WordPress brochure websites from around £299. Some charge more, some less, but the point is that there’s a huge market of agencies whose core offering is: give us some money, we’ll build you a website.
The problem is that AI website builders like Wix, Squarespace, Webflow, and a growing number of newer tools can now generate a complete, responsive site from a text description. In many cases, the result is perfectly adequate for a small business that just needs a functional online presence. Plans typically run from free to around £30 a month.
This is already happening.
When I shared this observation on LinkedIn, the responses fell into some predictable patterns, and a few that were genuinely useful.
The defensive response: “AI websites are terrible”
Several web developers pushed back hard. The argument was essentially that AI-built websites are poor quality, have no real UX thinking behind them, and that the whole “agencies are doomed” narrative is overblown.
There’s some truth in that. AI-generated sites can be generic. They don’t understand your brand the way a human designer does. The code isn’t always clean.
The weakness in that argument is that the quality bar for a basic brochure site is lower than many web professionals think. A local plumber needs a site that loads fast, looks professional enough, shows their phone number clearly, and has some decent reviews on it. A custom design system adds little for them. AI can do that already, and it’s getting better quickly.
The defensive response is understandable. Nobody likes being told that what they sell is becoming commoditised. But ignoring the change won’t protect anyone’s margins.
What people in the industry are already seeing
More telling were the comments from people in the industry who shared real examples. One commenter described quoting against agencies selling basic websites for $500 to $800, only to find the same clients asking why they’d pay that when AI tools can generate something in 30 seconds.
Another pointed out that hiring a developer for a brochure website will soon be a thing of the past, and that the focus needs to shift to outcomes, problem solving, expertise, and personal relationships.
These are observations from people watching it happen in their own businesses right now.
AI is exposing weak agency offers
One comment summed it up well: AI is exposing weak offers, and that is a different problem from agencies being replaced.
I agree.
If your web design agency sells “a WordPress website” as the product, you’re selling a deliverable. A thing. And when AI can produce a comparable version of that thing for a fraction of the cost, your pricing collapses. It doesn’t matter how good your code is, how clean your CSS is, or how many years of experience you have. If the client can’t tell the difference between your output and what AI produces, you have a positioning problem.
The agencies that will thrive are the ones selling something AI genuinely can’t replicate: strategic thinking about what the website needs to achieve, deep understanding of the client’s customers and market, conversion expertise, ongoing optimisation based on real data, and a trusted advisory relationship where the agency acts as a partner to the client.
Clients will keep paying an agency for the business results a website produces. My Value Transformation Assessment looks at this in more detail for your own business.
Can AI handle the strategy work too?
One of the more challenging responses raised an interesting point. What happens when AI can plug into heatmaps and analytics, advise on why visitors aren’t converting, build strategy around commercial goals, and proactively adapt based on performance data?
It’s a fair question, and the honest answer is that AI is already doing some of this. Tools exist that can analyse conversion paths, suggest layout changes, and run multivariate tests automatically.
But analysis and judgement are different skills. AI can tell you that visitors are leaving your contact page without getting in touch. It can even suggest changes. What it can’t do is sit across the table from your client, understand that their real problem is that they’re targeting the wrong customer segment entirely, and help them rethink their go-to-market strategy.
The client doesn’t necessarily care whether you’re using AI tools in the background. They care about results. If you’re the person who helps them understand why their business isn’t growing and what to do about it, you’re a trusted partner. If you’re the person who builds them a website and sends an invoice, you’re a supplier who’s about to be replaced by software.
Functional and premium purchases: Premier Inn and The Dorchester
Another commenter drew a useful comparison between functional purchases and emotional ones. A Premier Inn room does the job. It’s clean, consistent, and reasonably priced. The Dorchester is a different product entirely. There you pay for the experience, the status, and the relationship as well as the bed.
AI website builders are the Premier Inn of web design. They’ll do the functional job well enough for many basic needs. The question for agencies is: are you selling Premier Inn rooms and trying to charge Dorchester prices? Or have you genuinely built a Dorchester-level service?
Many agencies sit somewhere in between, and that’s uncomfortable for them. They’re charging more than the AI tools but not offering enough differentiation to justify it. Margins get squeezed hardest in that middle ground.
What this means if you run a web design agency
If you’re reading this and recognising some of these patterns in your own business, here are a few things worth considering.
First, look honestly at what you’re actually selling. If your proposals are structured around deliverables (homepage design, 5 inner pages, contact form, SEO setup), you’re selling a list of outputs, and that’s the offer AI can undercut. Restructuring your proposals around business outcomes (increased lead generation, improved conversion rates, measurable revenue impact) moves the conversation on to the value you create. I cover how to do that in how to position your services as an investment.
Second, think about your client relationships. Are you a supplier who gets a brief, builds a thing, and moves on? Or are you a partner who understands the client’s business, tracks what’s working, and proactively advises on improvements? The first role is easy to replace. The second is much harder to replace.
Third, consider your pricing model. If you’re still charging fixed project fees for website builds, every improvement in AI tools puts downward pressure on what you can charge. Retainer-based pricing tied to ongoing value, where your fees are linked to the outcomes you deliver, creates a much more defensible business.
Finally, use AI yourself. The agencies that will do best are the ones using AI to speed up the commodity parts of their work (initial layouts, first-draft copy, code scaffolding) while investing more time in the strategic, relationship-driven work that AI can’t do.
The same pressure applies to other service businesses
Web design agencies are just the most visible example right now because the AI tools are so tangible. You can watch an AI build a website in real time. It’s dramatic and it gets attention.
The same dynamic is playing out across many service businesses where the core offering can be described as a deliverable. IT support, bookkeeping, marketing, recruitment, consulting. If you can reduce what you do to a checklist of tasks, AI will likely do those tasks cheaper and faster in time.
The businesses that survive and grow will be the ones that have moved from deliverables to outcomes, from supplier relationships to trusted partnerships, and from cost-based pricing to value-based pricing.
It’s an uncomfortable message, but an honest one.
Check how exposed your business is to AI
I’ve built a free assessment specifically for service businesses that want to understand how defensible they are against AI disruption. It takes about five minutes, and you’ll get an immediate score across four areas: replaceability, pricing resilience, supplier vs trusted partner positioning, and adaptability.
It gives you an honest view of where you’re strong and where the gaps are.