What Website Awards Actually Reward
Award criteria read against their own published weights and jury methods, the biases that follow, and two claims about award winners that I could not substantiate and therefore do not make.
Date
Category
Writer
Awwwards publishes its scoring weights, which is more than most awards do. They are: Design 40 per cent, Usability 30 per cent, Creativity 20 per cent, Content 10 per cent.
Read that as an instruction rather than a description, because that is what it is. Sixty per cent of the available score is aesthetic judgement. Ten per cent is what the thing says. Accessibility is not a category. Performance is not a category. If you were briefing a studio and handed them that weighting, you would get back exactly the kind of site that wins.
What the criteria actually say
I went and read the published criteria for three award schemes rather than relying on what people say about them, and the differences are larger than the similarities.
Scheme | Published criteria | Accessibility mentioned | Performance mentioned |
|---|---|---|---|
Awwwards | Design 40%, Usability 30%, Creativity 20%, Content 10% | Not in the published criteria | Not in the published criteria |
CSS Design Awards | Thresholds published, category breakdown not published on the page I read | Not stated | Not stated |
Webby Awards | Content, Structure and Navigation, Visual Design, Functionality, Interactivity, Innovation, Overall Experience | Yes, in two criteria | Yes, under Functionality |
The Webby criteria are the surprise. Under Visual Design, the scheme asks that design supports the message "while ensuring accessibility for people with disabilities and various cognitive needs". Under Functionality, it asks that a site "load quickly, work cross-platform, and consider diverse user requirements including special access needs and bandwidth limitations". Those are not perfunctory mentions. They are two of seven criteria carrying explicit accessibility and performance requirements.
I had assumed, before reading them, that the Webbys were the softer, more commercial award and that the site-of-the-day schemes were where the craft was. On the published criteria, the opposite is true. The Webbys ask about more of the things I think matter. That is a thing I got wrong and had been repeating.
Photograph by Edu Grande, CC0, via Wikimedia Commons.
How the judging works
Awwwards is the most transparent about process, and the process is better than its reputation suggests.
Eighteen independent scores with automatic trimming of the three biggest outliers is a defensible piece of method. It is more rigorous than most design-press editorial selection, which is one person with a deadline. Trimming outliers specifically guards against a juror tanking or boosting a friend's entry, which is the obvious failure mode of a peer-judged award.
CSS Design Awards publishes its thresholds: Website of the Day requires an average score above 8.00, Special Kudos certificates go to sites above 6, and its public-vote awards require more than 20 votes plus an average judge score above 6. What the page I read does not publish is what the judges are scoring on, or how many of them there are. It refers to an "international panel of expert judges" and leaves it there. A published threshold without published criteria tells you the bar is high without telling you the bar is what.
The biases that follow from the arithmetic
None of what follows requires anyone to be acting in bad faith. These are consequences of the weightings.
Motion
Design and Creativity together are 60 per cent of an Awwwards score, and they are assessed by jurors who look at a great many sites in a sitting. Motion is the cheapest way to read as designed and creative inside the first three seconds of a juror's attention. Nothing in the criteria penalises it.
The relevant standard is WCAG Success Criterion 2.3.3, Animation from Interactions, at Level AAA: "Motion animation triggered by interaction can be disabled, unless the animation is essential to the functionality or the information being conveyed." The Understanding document is specific about who this protects: "Some users experience distraction or nausea from animated content", with people who have vestibular disorders vulnerable to "dizziness, nausea and headaches", and reactions severe enough to require bed rest.
Level AAA is not required for ordinary conformance, and I am not going to pretend it is. But the implementation cost is low and adoption is now mainstream: the 2024 Web Almanac's accessibility chapter reports that over 50 per cent of mobile sites implement the prefers-reduced-motion media query, up from 34 per cent in 2022. A majority of the ordinary web now respects that preference. An award that scores 60 per cent on design and creativity has no mechanism to ask whether the winner does.
Weight and load
Here is the fair version of the complaint, with the baseline attached, because the baseline is worse than people think.
The 2024 Web Almanac's page weight chapter, by Dave Smart and Jamie Indigo, puts the median home page at 2,652 KB on desktop and 2,311 KB on mobile as of October 2024, up 8.6 per cent and 6.4 per cent respectively on the year before. The median mobile page carries 900 KB of images and 558 KB of JavaScript. At the 90th percentile, mobile pages reach 7,680 KB. Over a decade, median mobile page weight has grown 357 per cent.
On performance, the 2024 Almanac's performance chapter, by Jevgenija Zigisova and Ines Akrap, reports that 43 per cent of websites had good Core Web Vitals on mobile against 54 per cent on desktop, and that 59 per cent of mobile origins had good Largest Contentful Paint.
So the ordinary web is heavy and slow. The honest claim about award winners is not that they are worse than the web, it is that the awards contain no term that would notice if they were. Awwwards' published criteria have no performance component at all. The Webbys do, under Functionality.
Accessibility
The WebAIM Million's 2026 report, on data collected in February 2026 from the home pages of the top one million sites, found detected WCAG 2 failures on 95.9 per cent of them, at an average of 56.1 errors per page. Six categories account for 96 per cent of all detected errors: low contrast text (83.9 per cent of pages), missing alternative text for images (53.1 per cent), missing form input labels (51 per cent), empty links (46.3 per cent), empty buttons (30.6 per cent) and missing document language (13.5 per cent).
Low contrast text is the headline, and it is the exact place where an award weighted 40 per cent on design points the wrong way. Grey-on-grey is a house style of contemporary web design. It reads as restraint. It fails the most common automated check on the web.
The Almanac's accessibility chapter, written by Mike Gifford, puts the same problem from the other direction: only 29 per cent of mobile sites and 28 per cent of desktop sites have sufficient text colour contrast, and 53 per cent of sites apply :focus {outline: 0}, removing the browser's default focus indicator, against 12 per cent using :focus-visible.
That focus-outline number deserves a pause. Removing the focus ring is a purely aesthetic act. It makes a site look cleaner in a screenshot and makes it unusable by keyboard. It is precisely the trade that a design-weighted award rewards and has no term to penalise.
Text density and lifespan, which I cannot verify
Two claims are made constantly about award galleries, including by me, and I have not been able to substantiate either.
The first is that award-winning sites carry markedly less text than comparable non-winning sites. This is plausible from the criteria: Content is 10 per cent of an Awwwards score, and long copy is hard to make photogenic. I have not found a published study measuring word counts across award winners against a control set, and I am not going to assert a number I have counted myself on a sample I chose.
The second is that award-winning sites have unusually short lifespans, that the elaborate ones are replaced or quietly simplified within a year or two. I believe this from watching links rot. I have no dataset, and the obvious confound is that the kinds of companies who commission award-bait sites are also the kinds who rebrand often. Treat it as a hypothesis.
Anyone can test both cheaply, and the tooling is public. That is worth saying because the usual state of this argument is two people exchanging impressions.
What the awards genuinely do well
Three things, and I do not think the critique above touches any of them.
They are a live archive of technique. There is no other place where you can see what is currently possible in a browser, made by people with a budget, at a rate of roughly one a day. Whatever you think of the winners, the galleries are a faster read on the state of the medium than any conference.
The judging method is better than the alternatives. Eighteen scorers with outlier trimming is a real procedure. Compare it with how most work gets celebrated, which is one editor, one afternoon, one set of friends.
They pay attention to the thing itself. Award juries look at the built site, in a browser, in a session. Most professional evaluation of digital work happens through a deck. There is something clarifying about a system in which the only submission is the live URL.
An award is a measuring instrument. Instruments are not right or wrong, they are calibrated for something. The mistake is not trusting the reading. The mistake is not asking what it measures.
What a studio should take from them
I used to treat these galleries as a quality benchmark, which is a category error I would now push back on if a junior designer made it. Here is the replacement.
Use it for | Do not use it for |
|---|---|
Finding out what is technically achievable this year | Deciding what your project should be |
Spotting a technique worth learning | Inferring that a technique is appropriate |
Calibrating how much visual ambition a budget buys | Calibrating quality |
Seeing what jurors reward, if you intend to enter | Seeing what clients get value from |
Studying a specific interaction in detail | Copying a whole site's structure |
And a practical routine that costs an hour a month. Take three current winners. Run each through whatever public performance and accessibility tooling you use. Write down the numbers. Then look at the site properly and decide, separately, whether you admire it. Doing both and keeping them apart is the whole discipline. Plenty of sites I admire score badly, and some of them are right to: a piece of work whose entire purpose is a five-minute immersive experience is not obliged to be a news site.
Two cautions about the tooling, because the numbers are easy to over-read. The Almanac's accessibility chapter says it directly: "automated tests can only address a portion of WCAG Success Criteria" and "a perfect score does not guarantee an accessible site". A clean automated report means the page passes the machine-checkable subset. It does not mean a screen reader user can complete the task. In the other direction, a poor score on a deliberately experimental piece is information about a trade-off, not a verdict.
The useful conclusion is narrower than the usual one. The awards are not corrupt and the winners are not bad work. The criteria are published, they are weighted heavily towards visual impression, and the schemes deliver what they weight. Awwwards gives 365 Sites of the Day a year on a scoring system where 60 per cent of the marks are design and creativity. It would be strange if the results looked like anything else.
What a studio owes its clients is a separate scorecard, written before the project starts, weighted for what that client actually needs. If it happens to overlap with a jury's, take the entry fee out of the marketing budget and enter. If it does not, the gallery is a magazine, and you can read a magazine without taking instructions from it.
Sources
Awwwards, "Evaluation", published criteria and jury process: Design 40%, Usability 30%, Creativity 20%, Content 10%; minimum 18 jurors; removal of the 3 most deviant scores; 5-day voting; Honourable Mention at 6.5; Developer Award above 7.0. awwwards.com/about-evaluation
CSS Design Awards, "About", score thresholds for Website of the Day (above 8.00), Special Kudos (above 6) and public awards (more than 20 votes and judge average above 6). cssdesignawards.com/about
The Webby Awards, "Judging Criteria": Content, Structure and Navigation, Visual Design, Functionality, Interactivity, Innovation, Overall Experience, including the accessibility and load-time language. webbyawards.com/judging/judging-criteria
WebAIM, "The WebAIM Million: The 2026 report on the accessibility of the top 1,000,000 home pages", data collected February 2026. webaim.org/projects/million
Mike Gifford, "Accessibility", chapter of the 2024 Web Almanac, HTTP Archive, 2024. Colour contrast,
prefers-reduced-motion, focus outline and Lighthouse median figures, and the caveats on automated testing. almanac.httparchive.org/en/2024/accessibilityDave Smart and Jamie Indigo, "Page Weight", chapter 16 of the 2024 Web Almanac, HTTP Archive, published 11 November 2024. Median and 90th percentile page weights, October 2024. almanac.httparchive.org/en/2024/page-weight
Jevgenija Zigisova and Ines Akrap, "Performance", chapter of the 2024 Web Almanac, HTTP Archive, published 11 November 2024. Core Web Vitals pass rates for mobile and desktop. almanac.httparchive.org/en/2024/performance
W3C Web Accessibility Initiative, "Understanding Success Criterion 2.3.3: Animation from Interactions", WCAG 2.2, Level AAA. w3.org/WAI/WCAG22/Understanding/animation-from-interactions.html
Related reading
Latest Articles.
© Moc Mallen Studio
Thoughts, ideas, and perspectives on design, simplicity, and creative process.
.jpg){kind=link}