Devstars
Blog
Date: 22/01/2026
Stuart WatkinsHow design agencies can escape algorithmic dependency and build digital infrastructure that AI tools actually recommend
Your Behance following looks healthy. Your Dribbble shots get featured. Your work wins awards. Yet when a fintech startup asks ChatGPT to recommend branding agencies in London, your name doesn’t appear.
This gap between creative recognition and commercial visibility is what I call the Trust Architecture problem. And it’s costing design agencies clients they’ll never know they lost.
TLDR: Knowing how to make design portfolios convert in 2026 comes down to one shift. Stop optimising for peer validation. Start building for AI discoverability. The agencies winning new business have moved from relying on platform visibility to building digital infrastructure that AI tools understand, trust, and recommend. This article breaks down the five-part framework that makes the difference, plus a 90-day implementation roadmap.
Last month, I ran an experiment. I asked ChatGPT, Perplexity, Claude, and Google’s AI Overview to recommend branding agencies for various client briefs: fintech rebrand, sustainable packaging design, and B2B SaaS visual identity. Across 50 searches, a pattern emerged.
The agencies appearing in AI recommendations weren’t necessarily the most awarded or most followed. They were the agencies with content-rich websites, detailed case studies, and consistent messaging across multiple platforms. Visual portfolios with minimal text, however stunning, were invisible.
This matters because search behaviour has fundamentally changed. 59% of Google searches now end without a click. Users complete their research inside AI tools, then jump directly to shortlisted websites. If you’re not appearing in that research phase, you’re not making the shortlist.
Dirk Petzold recently described this as “Algorithmic Sharecropping,” and he’s right. When you build your business on third-party platforms, you work the land but the platform owns the crop. The feed controls your visibility. The algorithm decides your reach. And crucially, AI tools can’t parse visual galleries to recommend you to potential clients.
Here’s the uncomfortable truth: AI recommendation engines don’t browse portfolios. They read text.
When someone asks an AI tool for agency recommendations, the tool searches the web, analyses content, and synthesises an answer. It looks for clear expertise signals, specific experience documentation, and authoritative third-party citations. A beautiful portfolio grid provides none of these.
Consider what happens when a prospect searches “best branding agency for healthtech startups”:
The AI tool scans websites for relevant content. It finds Agency A’s case study titled “How We Helped a Healthtech Startup Increase Patient Engagement by 340%,” complete with methodology, challenges overcome, and measurable outcomes. It finds Agency B’s portfolio page showing healthtech logos with no context. Agency A gets cited. Agency B doesn’t exist in the AI’s response.
This isn’t about SEO in the traditional sense. It’s about what I call Generative Engine Optimisation, or GEO: structuring your digital presence so AI tools can understand, trust, and recommend your expertise.
To help agencies assess their AI visibility, I’ve developed a simple framework called the Trust Architecture Index. It measures five dimensions that determine whether AI tools can effectively recommend you.
1. Semantic Density (0-5 points)
Semantic density measures how much meaningful, parseable text exists alongside your visual work. A portfolio page showing project images scores 0. A 2,000-word case study explaining the strategic challenge, creative approach, client collaboration, and measurable outcomes scores 5.
Ask yourself: if someone removed all images from your portfolio, would a reader still understand your expertise and approach?
2. E-E-A-T Signal Strength (0-5 points)
E-E-A-T stands for Experience, Expertise, Authoritativeness, and Trustworthiness. These signals determine whether AI tools trust your content enough to cite it.
Experience signals include documented case studies, before/after evidence, and firsthand project narratives. Expertise signals include author attribution, professional credentials, and industry recognition. Authoritativeness signals include third-party citations, media mentions, and peer recognition. Trustworthiness signals include client testimonials, review ratings, and transparent business information.
Most creative portfolios score 1-2 on E-E-A-T. They show work but don’t document the expertise behind it. The person who did the work isn’t named. The methodology isn’t explained. The results aren’t quantified.
3. Citation Presence (0-5 points)
This measures how often your agency appears in third-party sources that AI tools reference. Industry publications, design directories, awards databases, and professional platforms all contribute.
Run this test: search Google for your agency name plus “featured in” or “mentioned by.” How many authoritative third-party sources appear? Now compare that to competitors who show up in AI recommendations. The gap tells you where to focus.
4. Content Structure (0-5 points)
AI tools prefer content structured for comprehension: clear headings, logical flow, definitive statements, and quotable insights. Long-form prose with subheadings scores higher than bullet-point lists. Specific, citation-worthy statements score higher than vague descriptions.
The ideal structure mirrors how AI tools construct answers: concise summary at the start, detailed explanation of key considerations, specific examples and evidence, expert insights and firsthand experience, clear conclusions or recommendations.
5. Platform Distribution (0-5 points)
This measures your presence across platforms AI tools reference when compiling recommendations. A website-only presence scores 1. Website plus LinkedIn authority content scores 2. Add YouTube explanations of your approach and you’re at 3. Include Reddit AMAs, podcast appearances, and industry publication contributions, and you approach 5.
Remember: 59% of searches are no-click, but that research happens across multiple platforms. YouTube processes 3 billion daily searches. LinkedIn drives professional authority signals. Each platform you occupy increases your citation surface area.
Add your scores across all five dimensions. Maximum possible: 25.
20-25: AI-Ready Your digital infrastructure supports AI recommendations. Focus on maintaining and expanding your citation presence.
15-19: Foundation Present Core elements exist but gaps remain. Prioritise semantic density and E-E-A-T signal strengthening.
10-14: Significant Gaps Visual-first approach is limiting your AI visibility. Begin systematic content development around existing portfolio work.
Below 10: Invisible to AI Your brilliant work exists in a format AI tools cannot process. Fundamental restructuring required.
Most design agencies score 8-12. Strong visual presence, weak semantic infrastructure.
Understanding why this matters requires grasping how modern search works. The buyer journey has split into two distinct phases.
Research Phase: Prospects use AI tools to identify potential partners. They ask questions, compare options, and build shortlists. This happens inside ChatGPT, Perplexity, Google AI Overviews, and similar tools. Visual portfolios don’t register here.
Purchase Phase: Once shortlisted, prospects visit websites directly. They review work, assess cultural fit, and make contact decisions. Your portfolio matters enormously here, but only if you made the shortlist.
The agencies winning new business are those optimising for both phases. They build Trust Architecture for AI visibility in the research phase, then deliver visual excellence for conversion in the purchase phase.
Optimising only for the purchase phase means relying entirely on referrals and direct outreach. That works, until it doesn’t. When a competitor builds AI visibility and you haven’t, their shortlist presence compounds while yours stagnates.
Here’s what makes this particularly challenging for creative agencies: the platforms that built your reputation actively work against AI visibility.
Dribbble optimises for peer engagement, not client discovery. The format, isolated shots with minimal context, is precisely what AI tools cannot parse. A hundred thousand followers means nothing to ChatGPT when a prospect asks for agency recommendations.
Behance offers slightly more context but still prioritises visual gallery browsing over searchable, citable content. Neither platform allows the semantic depth that AI recommendation engines require.
This doesn’t mean abandoning these platforms. It means understanding their role. Use Dribbble for peer networking and inspiration. Use Behance for visual discovery by human browsers. But build your client acquisition infrastructure on owned property with AI-readable content.
The Hub-and-Spoke model works here: your website is the hub with full case studies and semantic density. Platform profiles are spokes that tease work and link back. Traffic flows inward. Authority consolidates on property you control.
Transforming your digital presence doesn’t require rebuilding from scratch. It requires systematic enhancement of what exists.
Days 1-30: Foundation
Start with your three strongest projects. For each, write a comprehensive case study of 1,500-2,000 words covering: the client’s business challenge (not just the brief), your strategic approach and methodology, key decisions and why you made them, collaboration dynamics and obstacles overcome, measurable outcomes and client impact.
Attribute each case study to a named expert at your agency. Include their credentials, experience, and perspective. This builds the expertise signals AI tools look for.
Days 31-60: Amplification
Take those case studies and create derivative content for other platforms. A YouTube video walking through your process. A LinkedIn article exploring one strategic insight in depth. A Reddit contribution in a relevant community answering a question your case study addresses.
Each piece links back to your site. Each builds citation surface area. Each reinforces your expertise in formats AI tools can parse.
Days 61-90: Citation Building
Identify industry publications, directories, and platforms that AI tools reference. Pitch contributed articles. Seek award nominations. Request client testimonials on third-party review sites. Build the external validation that signals authoritativeness.
Track your AI visibility monthly. Run the same prospect searches and document when you start appearing. The gap between your current state and AI recommendation presence is your Trust Architecture deficit.
Here’s the strategic reality: most creative agencies haven’t recognised this shift. They’re still optimising for peer validation on visual platforms while AI tools reshape how clients find partners.
That creates a window. The agencies building Trust Architecture now will establish AI visibility before competitors wake up. They’ll compound their recommendation presence while others wonder why enquiries have declined.
This window won’t stay open. As more agencies recognise the shift, the barrier to AI visibility will rise. First movers build advantages that late adopters struggle to overcome.
The question isn’t whether to adapt. It’s whether to adapt now, while the competitive advantage is available, or later, when you’re playing catch-up.
Trust Architecture doesn’t just improve AI visibility. It improves conversion once prospects arrive.
Case studies with strategic depth demonstrate thinking, not just execution. E-E-A-T signals build confidence before the first conversation. Consistent messaging across platforms reinforces positioning. The same content that makes AI tools recommend you makes human prospects trust you.
This aligns with what premium clients actually want. Research into boutique agency selection shows that clients paying £8k+ seek agencies who “understand business challenges, not just execute creative briefs.” They want strategic partners who can “explain processes clearly and offer proactive advice.”
A portfolio of pretty pictures doesn’t communicate strategic capability. Detailed case studies do. Trust Architecture serves both AI visibility and client conversion.
Traditional creative metrics, followers, likes, features, measure peer approval. Business metrics measure client acquisition. These aren’t the same.
Track instead: AI recommendation appearance rate across target searches, website traffic from direct and branded search (indicates research-phase shortlisting), enquiry quality and average project value, conversion rate from first contact to signed project.
If your Behance followers are growing but these metrics aren’t, you’re building the wrong kind of visibility.
Petzold’s concept of the “Sovereign Creator” captures what’s required: designers who own their data, their audience, and their platform. Who build destinations rather than contributing to feeds. Who control context rather than surrendering it to algorithms.
For creative agencies, this means treating your website as a strategic asset, not a portfolio gallery. It means investing in content that compounds rather than content that expires with the feed refresh. It means building infrastructure that serves your business development, not just your creative reputation.
The era of the portfolio grid is ending. The era of Trust Architecture is beginning. The agencies that recognise this shift, and act on it, will capture the clients that others don’t even know they’re losing.
Want to know your agency’s actual Trust Architecture Index score? I’m offering complimentary audits for UK design agencies this quarter.
The audit includes: complete AI visibility assessment across ChatGPT, Perplexity, Claude, and Google AI, competitive citation analysis showing who’s appearing where you should be, E-E-A-T signal gap identification, and a prioritised 90-day improvement roadmap.
No pitch attached. Just the data and what it means.
Request your Trust Architecture Audit →
Stuart Watkins is Founder and Director of Devstars, a digital consultancy specialising in Generative Engine Optimisation and bespoke web development. With 35 years in digital marketing, he’s helped agencies and brands build visibility in both traditional and AI-powered search environments.
Currently scheduling strategic partnerships for Q1-Q2 2026. Limited spaces remain.
Get a free technical consultation and project roadmap. We’ll assess your requirements and provide transparent pricing for your growth-stage development needs.
Call: +44 020 8898 3993
