--- title: "Good designers, bad websites: the problem isn't talent, it's memory" locale: "en" url: "https://irz.fr/en/articles/designers-bad-websites-en" markdown_url: "https://irz.fr/en/articles/designers-bad-websites-en.md" category: "creative" tags: ["design", "accessibility", "ux", "web"] published_at: "2026-08-21T10:10:00.000Z" author: "Clara Besson" translation: "https://irz.fr/fr/articles/designers-bad-websites-fr.md" --- # Good designers, bad websites: the problem isn't talent, it's memory An accessibility consultant proposes flipping a Nielsen heuristic back onto designers: recognize accessibility problems while designing, instead of trying to recall a thousand rules. The paradox is as old as the web: some of the best designers in the trade have genuinely bad personal sites. Unreadable typography, temperamental navigation, contrast reserved for young eyes — while those same people ship polished, tested, documented products for their clients. It isn't bad faith. It's a memory problem, argues Alan Dalton on A List Apart, and there is a remedy.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/) ## Two premises that are hard to dispute Dalton, who spent eight and a half years as Accessibility Development Advisor at Ireland's National Disability Authority, starts from two statements.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/) The first: designers are good people. He says he has never heard a designer declare they don't care whether text is readable or a service is understandable. The second: some designs exclude people anyway. Everyone has seen it — text impossible to read, a device impossible to hold, a service that confuses more than it helps. For the W3C, accessibility isn't optional: it's the condition for websites and tools to be usable by everyone, whatever their abilities.[4](https://www.w3.org/WAI/fundamentals/accessibility-intro/) The gap between those premises is where things get interesting. If nobody wants to exclude anybody, why do so many products end up doing exactly that? Dalton's answer fits in four words: there is too much to recall. A designer is expected to carry interface best practices, the entire accessibility guidance corpus, every platform's conventions and performance constraints in their head. The cognitive load exceeds what a brain can retrieve at decision time. The flaw isn't moral; it's mnemonic.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/) ## Turning the heuristic against its reader To make the case, Dalton reaches for a 1990s classic: Jakob Nielsen's ten usability heuristics. Number six says information needed to use an interface should be visible or easily retrievable — recognition rather than recall.[2](https://www.nngroup.com/articles/ten-usability-heuristics/) His move is elegant: apply that rule not to the end user but to the designer. The information required to produce the design should be visible or easily retrievable while designing. In other words, stop asking designers to memorize a thousand rules, and give them tools that make problems recognizable at the moment they arise.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/) > **Recognize, don't memorize** > - User types to keep in sight.: Personas > - Someone tasked with asking the awkward questions.: Designated dissenter > After Alan Dalton, A List Apart ## Free personas, not another framework The proposed tool already exists: the eight accessibility personas created by Sarah Horton and Whitney Quesenbery for *A Web for Everyone*, now available free on the companion site of Horton's later book with David Sloan.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/)[3](https://knowaboutaccessibility.org/tag/personas/) These personas have a distinctive trait: they speak. Carol, a grandmother with macular degeneration, asks why the text can't be just a little bit larger. Lea, an editor living with fatigue and chronic pain, begs not to be made to work so hard. Maria, a bilingual community health worker, puts it flatly: when a site is confusing, she just leaves.[1](#ref-1] These aren't abstract archetypes but sentences that stay with you when choosing a font size or a dropdown menu. Dalton adds two known safeguards: Cennydd Bowles' “Personas Non Grata” — also think about what a bad actor could do with your product — and Eric Meyer and Sara Wachter-Boettcher's “Designated Dissenter”: on every project, someone is responsible for asking “will this work for Vishnu?”.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/) ## Where does the personal site fit? This is where IRZ takes some liberties with the original proposal, because the cognitive-load thesis explains one case poorly: the portfolio. On their own site, the designer isn't carrying a thousand projects in mind. Just one. Time is rarely short, no client applies pressure — yet the result often comes out worse than commissioned work. Our reading: personal sites fail less from overload than from the absence of every discipline designers impose on their clients. No brief — you design for yourself without a statement of work. No observed users — you never watch anyone visit your portfolio. No designated dissenter — no colleague to say the custom cursor is unbearable. And above all no client to challenge decisions: the self-client is the worst client there is, because it approves every whim. On top of that sits an accumulation of contradictory roles. A personal site must be business card, skill demo, experiment playground, CV and archive at once. Each role pulls the design in a different direction, and successive trends — light gray on white, tiny typefaces, scroll-hijacking animations — end up regressing accessibility with every redesign. ## A mission, not a trial Dalton closes with homework rather than an indictment: take the personas on knowaboutaccessibility.org and use them in every project, so accessibility issues become recognizable during design instead of repairable after the fact.[1](https://alistapart.com/article/good-designers-bad-websites-a-proposal/)[3](https://knowaboutaccessibility.org/tag/personas/) Transposed to the portfolio, this yields a simple method: write yourself the brief a client would have written, hand the site to someone who has never navigated it, and appoint a dissenter among your peers. Talent was never the problem. What's missing is exactly what you'd provide any client: a frame, an outside eye, and memory made visible. ## References 1. [Alan Dalton, “Good designers, bad websites: a proposal”, A List Apart, April 23, 2026](https://alistapart.com/article/good-designers-bad-websites-a-proposal/) 2. [Jakob Nielsen, “10 Usability Heuristics for User Interface Design”, Nielsen Norman Group](https://www.nngroup.com/articles/ten-usability-heuristics/) 3. [Free-to-use accessibility personas, knowaboutaccessibility.org](https://knowaboutaccessibility.org/tag/personas/) 4. [W3C Web Accessibility Initiative, “Introduction to Web Accessibility”](https://www.w3.org/WAI/fundamentals/accessibility-intro/)