The Real Threat of AI in Web Design

Setting the Stage: Not Another Doom Presentation

Tiffin opened by explicitly distancing himself from what he called the dominant mode of AI discourse in the industry. He acknowledged that audiences are, in his words, "a little bored of AI presentations," and framed his session as one that would "cut through the hype, the bullshit, the doomering." The talk, he emphasised, was not about whether AI is coming for web designers (a question he considers settled and exhausted) but about what is actually happening on the ground and what practitioners should do in response.

His tone throughout was that of someone who has been in the industry long enough not to panic, someone who has, as he put it, "watched this movie before."

Historical Perspective: The Wix and Squarespace Panic

Tiffin anchored his argument in a historical parallel that he clearly finds instructive. Around 2013 to 2014, Wix and Squarespace were framed as existential threats to agencies and freelance designers. Forums, conferences, and industry conversations at the time sounded, he argued, almost identical to what is being said about AI today. Agencies were going to die.

Designers would be made obsolete. Drag and drop would replace skill.

He described a personal anecdote from that period: a client left his agency to build their own website on Squarespace, only to return within two months reporting that conversions were down, they could not do what they needed on the platform, and they had been oversold on what it could deliver. He also described friends in his social circle who joked about him needing to visit the job centre or retrain as an electrician.

The punch line: Wix and Squarespace did not kill agencies. What they did was filter the market. The clients who left were, as Tiffin described them, the "plumber who wanted a four hundred pound website," the local handyman, the gardener who wanted to do a trade deal, "your grandmother selling chestnuts at the village fair." These were low-margin, high-friction, time-consuming clients. When they migrated to DIY platforms, it was, he said, "frankly a gift" to agency owners. The serious businesses stayed with agencies because they needed more than a template.

His conclusion: the collective panic rarely matches the actual outcome.

The Current Landscape: The 30-Second Threat

Tiffin made the case that AI represents a genuine and significant disruption, particularly for the lower end of the market, but he was precise about what that disruption actually is.

He outlined the traditional website production process: discovery calls, wireframing, design mock-ups, development and coding, content creation and integration, testing, feedback loops. A decently sized project, he noted, would historically take eight to twelve weeks.

Today, a client can generate a website in approximately 30 seconds using AI tools. He was careful to note that the result is not necessarily a great website, but for many clients at that end of the market, it is "good enough."

He broke down the economics: AI-generated website: approximately $15 per month ongoing cost, 30 seconds of time.

Traditional build: approximately $5,000 upfront.

Cost saving: 97%.

He then demonstrated this live using Lovable, an AI website builder. He typed a single prompt requesting a bakery website for a fictional business called "Ben's Bakes" in Tower Dwells, Kent, with opening hours, a menu of freshly roasted coffee, handmade sandwiches, and pastries. Within the presentation, the tool generated a functional website from that prompt alone. He showed how subsequent prompts could expand individual sections, such as adding a full page about the coffee philosophy.

His assessment was honest: "This is not a five grand website, I admit," but he noted that with additional prompting and refinement, the gap between what AI produces and what agencies produce is narrowing for basic brochure sites. He also shared a personal example: when his wife started a business and needed a website to accompany flyers she had already had printed, he built one in approximately an hour using AI tools, a process that would otherwise have taken him three hours starting from a blank WordPress install.

The conclusion he drew: agencies whose entire offering is "we build nice brochure websites"

are competing against a $6.3 billion AI website builder market. His characterisation of that situation was stark: the word "slaughter" was used without apology.

What AI Is Actually Good At

Tiffin was careful not to position AI as purely a threat. He outlined how his own agency, The Outsourcing Agency, is actively integrating AI into its workflow, and he was specific about which tasks AI genuinely improves.

First drafts. AI can generate initial versions of copy, layouts, and content structures quickly, freeing up human time for refinement and strategy.

Deep research, with human validation. Tiffin specifically recommended Notebook LM as a research tool. Rather than treating AI as a source of truth, he described using it as a research assistant: you load in sources you have already identified as relevant and correct, then query the AI to validate and expand on that material. He noted that a significant portion of the research underpinning his own presentation was done using this method.

Competitor research. AI can process and synthesise competitive landscapes quickly.

Layout exploration. Rapid generation and iteration of layout options before committing to a design direction.

Code scaffolding. AI can generate the structural skeleton of code, which developers can then build on.

Mock-up images and copy. Rather than waiting for client-supplied content before beginning design work, AI can generate placeholder images and copy that give designs a realistic appearance during first draft presentations.

Wireframing. Tiffin mentioned tools including ReLUM and indicated that his agency was beginning to experiment with Claude's design capabilities for wireframing based on specified sections and requirements.

The business case he made for agency owners: all of these tasks save real time. For a white label agency like his, those savings can be passed to clients as cost efficiency. For individual agency owners and freelancers, the time saved is time that can be redirected toward the higher-value strategic work he went on to describe.

The Value Iceberg: Where the Real Money Lives

Tiffin introduced what he called the "value iceberg" as a visual framework for understanding where value actually sits in the current market.

Above the surface (visible, commoditised by AI): Layouts, code generation, basic copywriting, speed of delivery. These are the components of web work that AI has effectively commoditised. They are visible, easy to compare on price, and increasingly cheap to produce.

Below the surface (90% of the iceberg, where the real money is): Business strategy, user psychology, trust and credibility, complex third-party integrations (CRM systems, custom applications), stakeholder alignment, and brand nuance.

Tiffin's argument was that the mistake many agencies make is selling the tip of the iceberg, competing on visible, production-based deliverables, without adequately surfacing or charging for the deep strategic value underneath. His prescription: reposition and sell the depth.

The Sea of Sameness: A Market Opportunity Hidden in Plain Sight

Tiffin cited research from Nielsen Norman finding that 80% of AI-generated websites share almost identical structural logic. The reason, he explained, is that AI optimises for the average: the most statistically typical arrangement of elements that performs adequately across generic use cases. The result is that every AI-generated bakery, law firm, gym, and dental practice looks like every other one in its category.

He framed this not as a minor inconvenience but as a significant market opportunity. For practitioners who can deliver genuine differentiation through bespoke brand design, strategic thinking, and human judgment, the value of that work has increased, not decreased. Brand design and strategic thinking, he argued, have become more valuable, not less.

The Three-Way Comparison: Pure AI vs Traditional Development vs Hybrid Agency

Tiffin structured a direct comparison across five dimensions: Speed to market: Pure AI is near-instant. Traditional development is inherently slow.

Hybrid agencies can leverage AI to accelerate delivery significantly, retaining speed while maintaining human oversight.

Cost structure: Pure AI costs approximately $15 per month. Traditional development involves high upfront costs. Hybrid agencies can move away from time-and-materials pricing toward value-based pricing, charging on the basis of the outcome delivered to the client rather than hours worked.

Brand authenticity: Pure AI produces generic output. Both traditional and hybrid agency approaches can produce bespoke, differentiated work that reflects genuine brand identity.

Handling complexity: Tiffin was candid that current AI tools frequently fail when confronted with complex requirements, often fixing one issue while breaking something else, or losing context across long sessions. He acknowledged this may change, but noted it as a current limitation. Traditional and hybrid agencies are not similarly constrained.

Scalability: Returning to an AI-built project after a period of time typically requires several prompts to restore context before meaningful work can resume. Traditional and hybrid agency projects do not have this limitation; they can be rebuilt, expanded, or rearchitected without constraint.

The Mindset Shift: From Pixel Pusher to Business Consultant

The core strategic argument of Tiffin's session was a mindset shift he believes is now essential for practitioners who want to remain competitive.

The old positioning: "Pixel pusher." The question agencies answer is "Can you build this?"

This is a commodity transaction: the client knows what they want and they are shopping for the cheapest or fastest producer.

The new positioning: "Business consultant." The question agencies should be answering is "Why are we building this?" This is a strategic engagement. It commands a strategic fee.

Tiffin was explicit that this is not about acquiring new skills so much as reframing existing ones. Same person, same skills, just a different conversation, and ultimately ten times the value.

He gave three concrete examples of how this reframing plays out in practice: "We need a new website" becomes "What is the business outcome this website needs to drive?"

"Here's our quote for eight pages" becomes "Here's our discovery process to figure out exactly what you need."

"We charge by the hour" turns into "We charge by the value of the outcome."

Who Actually Pays for Premium Web Work, and Why

Tiffin delivered what he framed as a hard truth about the real economics of high-value web projects: nobody pays 25k for a website. What they pay 25k for is more leads, less risk, better conversion, protection from a botched launch, and competitive advantage.

Code, he argued, is the medium, not the product. When practitioners reframe what they sell through the lens of business outcomes rather than technical deliverables, they are able to command significantly higher fees for the same underlying work. He urged the audience to apply this lens to everything they sell.

He also made a specific observation about the marketing manager as a buyer: he does not believe there is a marketing manager in the world currently prepared to sign off on a web project delivered entirely through AI. Marketing managers, in his experience and view, understand risk and accountability in ways that make them natural buyers for human-led strategic web engagements.

Who to Target: The Strategic Client Profile

Tiffin was direct about the client profile he believes agencies should be actively pursuing in the current environment.

The target: businesses with annual turnover of one million pounds or more that have a dedicated marketing manager on staff.

His reasoning had two components. First, these businesses have budgets that align with the strategic fees practitioners can command once they reposition themselves. Second, and

perhaps more importantly, they have people who understand strategy. A dedicated marketing manager is someone who will buy into strategic thinking rather than simply comparing your quote to the $15 per month alternative.

He contrasted this with the smaller business owner, who is increasingly being targeted by major AI platforms as those platforms scale their marketing budgets. He predicted that the leads from small businesses will dry up as AI companies run large-scale campaigns (he imagined television and Super Bowl-style advertising) directly competing for those clients.

By the time that happens, he argued, the question small business owners will be asking is not "who should I hire?" but "why would I spend $5,000 when I can do this myself in 30 seconds?"

The value migration, as he described it, has already happened. The crossover point where commoditised production trends downward and strategic consulting trends upward has passed. The question is which side of that crossover practitioners are building their business on.

Practical Advice: Three Actionable Tips

Tiffin closed with three distilled recommendations: Become a digital consultant whose commodity happens to be website development. Lead with strategic thinking. Position the website as the output of a consultative process, not the product itself.

Sell business outcomes, not solutions. Every conversation with a client should be framed around what the client is trying to achieve, not what you are going to build.

Use AI to save time and deliver more value. Integrate AI tools into your workflow for the tasks they are genuinely good at, freeing up your human capacity for the work only humans can do.

He summarised the opportunity clearly: practitioners who can walk into a meeting and confidently discuss conversion rates, user psychology, and business goals will command a significantly higher premium than those who only talk about fonts, colours, and code.

His closing sentiment was optimistic. The people working in web agencies have relationships, taste, and judgment that AI cannot replicate. The opportunity, he concluded, has never been bigger for those willing to evolve.

Integrate AI tools into your workflow for the tasks they are genuinely good at, freeing up your human capacity for the work only humans can do.

BEN TIFFIN 1.

Audience Poll

During the session, Tiffin ran a live audience poll (via QR code) asking attendees how secure they felt about the rise of AI in the industry. He returned to the results later in the session and noted a spread of responses: some attendees described themselves as very concerned, some as somewhat secure, and a smaller group as very secure. He addressed the very secure contingent with mild humour, suggesting they had probably already logged off, while expressing hope that his presentation had been useful to those with concerns.

About the speaker Ben Tiffin The Outsourcing Agency

Ben Tiffin runs The Outsourcing Agency, where he works with web professionals on how AI is changing the competitive landscape for design and development work.