This week’s web design landscape delivered practical lessons across platform strategy, interactive front-end techniques, and product testing methodology. From Framer’s marketplace controversy to hands-on tutorials pushing WebGPU and Lottie animations, our team has pulled out the insights that matter most for client projects right now.

Key Takeaways

• Framer’s marketplace policy changes are forcing template creators to rethink distribution — agencies relying on third-party templates need a contingency plan.
• Tactile, animation-driven UX built with Lottie is raising the bar for interactive web experiences that feel physical and intentional.
• Three.js paired with WebGPU and TSL opens new territory for immersive 3D web elements without heavy framework overhead.
• AI’s real-world behaviour diverges sharply from cultural expectations — design teams must ground their AI integration work in actual capabilities, not science fiction.
• Dogfooding catches bugs but cannot replace structured user research — internal testing creates blind spots that only real users expose.

Framer’s Marketplace Shake-Up Demands Agency Contingency Planning

Framer has made a string of controversial changes to its template marketplace, and creators are divided on whether the platform is undermining its own ecosystem. As detailed in this breakdown of 10 Framer marketplace controversies, policy shifts around revenue splits, template visibility, and curation have left independent sellers uncertain about their future on the platform.

For our agency, the takeaway is straightforward. We never build client delivery pipelines on top of a single third-party marketplace. Any team leaning heavily on Framer templates for rapid site builds should diversify now — maintain your own template library, version-control your design systems, and treat marketplace templates as accelerators, not foundations.

Lottie Animations Are Setting a New Standard for Tactile Web Experiences

Alexey Kopytin’s case study on building tactile UX with Lottie animations demonstrates how to create web interactions that feel genuinely physical. The project — a digital stress-relief squeeze toy — used DOM events and distance-based interaction logic to make animations respond to user input with satisfying precision.

We have started incorporating Lottie-driven micro-interactions into client projects where engagement metrics matter most: landing pages, product showcases, and onboarding flows. The key principle here is that architecture must serve art direction. Animation isn’t decoration. When the interaction model is designed first and the code serves it, the result feels intentional rather than gimmicky.

Three.js and WebGPU Push 3D Web Design Into Practical Territory

A step-by-step tutorial on creating an interactive 3D cluster with Three.js, TSL, and WebGPU shows how noise functions, post-processing, and real-time interaction can transform a simple icosahedron into a dynamic visual element. WebGPU support is maturing, and the performance gains over WebGL are measurable.

For client sites where visual differentiation drives conversions — think luxury brands, creative portfolios, property developments — these techniques are no longer experimental. We treat them as production-ready tools, provided the fallback experience for older browsers remains solid.

AI Design Expectations Need Grounding in Reality, Not Fiction

A thoughtful analysis of what Star Wars got wrong about AI highlights a critical gap between cultural expectations and actual AI behaviour. The saga imagined thinking machines as immortal individuals. Real AI is copyable software with no persistent identity.

This matters for design teams. When we integrate AI features into client products, user expectations are shaped by decades of fiction. Our job is to design interfaces that communicate what AI actually does — pattern matching, content generation, data retrieval — without implying sentience or infallibility.

Dogfooding Cannot Replace Proper User Research

Nielsen Norman Group’s latest piece on dogfooding versus QA versus user research makes a clean distinction. Internal product use catches functional bugs. It does not catch usability problems because your team knows too much to represent real users.

We enforce this separation on every project. Our QA process handles technical validation. Usability testing happens with recruited participants who match the client’s actual audience. Skipping that step is how teams ship products that work perfectly — for people who already understand them.

Each of these developments reinforces a consistent principle: build on what you control, design for how things actually work, and test with the people who will actually use the product. Agencies that treat these as non-negotiable habits ship better work, full stop.

Frequently Asked Questions

What is tactile UX in web design?

Tactile UX refers to web interactions designed to feel physical and responsive, often using animation libraries like Lottie to create feedback that mimics real-world touch. It prioritises intentional motion design over static interfaces to increase user engagement.

How do web designers use Three.js for client projects?

Web designers use Three.js to build interactive 3D elements — product visualisations, immersive landing pages, animated backgrounds — directly in the browser. With WebGPU support improving performance, these elements are now viable for production sites, not just experiments.

Why can’t dogfooding replace user research in website design?

Your internal team has expert knowledge of the product, which means they navigate it differently from real users. Structured user research with external participants is the only reliable way to uncover genuine usability issues before launch.

What should agencies do about Framer marketplace changes?

Agencies should maintain their own internal template and component libraries rather than depending on any single marketplace. Treat third-party templates as starting points, and always retain full ownership of your design system assets.