Department 01
Awwwards publishes its evaluation system in full, which makes it the only major award on the web you can reverse-engineer. Here is what the number is made of.
The four criteria
Every submission is scored out of 10 across four weighted criteria. The weighting is the single most useful fact in competitive web design, because it tells you exactly where effort converts into score.
The visual case: layout, colour, typography, imagery, and whether the whole thing coheres. Nearly half the score, which is why so many winners look like they were art-directed before they were specified.
Jurors click through without instructions. A site that impresses and then strands you loses more here than it gains in creativity — the two categories are deliberately weighted so that spectacle cannot buy its way past confusion.
Risk is rewarded, but conditionally: the bold choice has to serve the work. This is the smallest of the three visible criteria, which surprises people who assume the awards are a novelty contest.
Real copy, evaluated as copy. Lorem ipsum on a submission reads as an unfinished project, and the criterion exists mainly to punish sites that treated writing as a content-fill step after the design was signed off.
Usability and content together outweigh creativity two to one.
The jury
An accepted submission goes out to a minimum of eighteen jurors — a rotating international panel of designers, developers, creative directors and agencies. Voting runs for five days.
The mechanism worth knowing is the outlier rule: the three jury scores furthest from the average are automatically discarded. One enthusiast cannot carry a site, and one detractor cannot sink it. What survives is the consensus of the middle, which explains a great deal about the house style of award-winning work — it is designed to be un-hated by a large panel, not adored by a small one.
There is one shortcut through the five-day clock. A site scoring highly with the jury and collecting votes from at least ten professional members can be called for Site of the Day before voting closes.
The six highest marks in the record. Every number on this page is made the way these were.
Before the jury sees it
Submission is not automatic. Every entry is checked against a minimum standard by a person, not a script, before it is passed to the jury — and the award body is candid that this takes time, up to about a week, longer when a weekend or a rush of entries backs the queue up.Awwwards
The public vote is filtered as well. Only professional accounts that have been through validation count towards a result, which is a plain defence against a studio conjuring up supporters — or detractors — for a nominee.Awwwards
The developer assessment runs against a published set of guidelines anyone can read — useful if you are building toward that award rather than just reading about it.
The honors
The awards are not a single prize. They are a graded series, and each level is decided by different people.
The entry tier. Clearing 6.5 with the jury means the work is good; it does not mean it won anything scarce. Most competent professional sites that get accepted land here. One catch, in Awwwards' own words: "you will only be able to see your jury score if you win SOTD." Everyone below that is told they placed, but never told by how much.
Scarcity does the work: there are 365 slots in a year, so a day is won against everything else submitted around it. This is the award most studios are actually chasing.
A second, separate assessment against developer guidelines, applied to sites that won a day. It is marked across six axes — search legibility, motion, accessibility, speed, responsive behaviour and markup — and every daily winner gets the marks published. Across 393 winners with published developer marks, the six axes separate sharply: motion averages 7.80 and accessibility 6.82, the lowest of the six by a clear margin. Speed and responsive behaviour sit level at 7.47. So this is the part of the judging where a good site is expected to clear the bar rather than stand out by clearing it — and where the same weakness shows up again and again.
The eight highest-scoring sites of each month are nominated and go back to the jury for a second review, with user votes feeding the outcome. Being judged twice is what makes this number hard to accumulate.
Every Site of the Month winner is automatically nominated, plus selected favourites. The winner is announced the following February, which is why the field's canon always runs a year behind the work.
The bodies
An award is only meaningful relative to what that jury values. The three that matter most reward materially different things, and the strongest work satisfies all of them at once.
| Awwwards | The FWA | CSS Design Awards | |
|---|---|---|---|
| Founded | 2009 | 2000 | — |
| Ladder | HM → Day → Month → Year | Day → Month → Year → Hall of Fame | Day → Month → Year |
| Judging | Hybrid: voting jury community plus an internal editorial panel | International jury panel; monthly and yearly honors derive from daily picks | Jury-scored on UI, UX and innovation |
| Rewards | Craft at scale — visual-first agencies, motion-heavy work, brand storytelling | Experiment — WebGL, immersive and interactive launches | Interface quality — a broader net than either |
| Reputation | The largest awards community on the web; the industry benchmark | One of the oldest digital design awards; the experimental standard | Prolific, faster-moving, less selective |
The practical reading: if the work is a brand experience with real motion craft, Awwwards is the target. If it is experimental — a game mechanic, a rendering trick nobody has shipped — the FWA jury is the one that will understand it. A site that places with all three in one run is not three times as good as its peers; it is a site that refused to specialise.
Winners strong on interaction design — one of the four axes the jury scores separately.
Reading the system
Take the four weights seriously and some conclusions follow that most redesign briefs get backwards.
A beautiful site with confusing navigation cannot win. Design's 40% is the largest slice but not a majority; usability's 30% is enough to hold a spectacular site below the line on its own. This is the single most common way ambitious work fails here.
Creativity is not the point of the creative award. At 20% it is worth half of design. Invention is what wins Site of the Year; execution is what wins the day.
Writing counts, slightly. Content's 10% looks negligible until you notice it is scored on real copy — the criterion is less a reward for good writing than a penalty for pretending the words don't exist yet.
The middle decides. With outliers discarded, a site that provokes strong disagreement scores like a site nobody felt much about. Divisive work needs to be divisive in ways a broad panel still scores highly — which is a much narrower target than "be bold."