⚡ Quick answer: People decide whether they like a site in 17-50 milliseconds — faster than a blink. Two things win that test: low visual complexity and familiar structure. Which means a site that looks simple and is arranged conventionally almost always beats one that’s original and busy. Choose a style not by the owner’s taste but by how well it fits those two constraints and your audience’s expectations.
Conversations about style usually start with a moodboard and end in an argument about a shade. Let’s start from the other end — with what actually happens in someone’s head during the first seconds on a site.
Joint research by Google and the University of Basel produced two findings that still shape web design practice.
First: visual complexity. Sites with low complexity are judged more appealing as early as 50 milliseconds in. If a screen carries many elements, colours and accents at once, people rate it as less attractive even when the design is technically good.
Second: prototypicality — how closely a site resembles other sites of its kind. Unfamiliar structure creates resistance even when it’s objectively convenient. Logo top left, menu across the top, cart on the right: these conventions aren’t constraints, they’re saved effort for the user.
The worst combination is complex and unfamiliar at once. The best is simple and recognisable.
The prototypicality effect actually strengthens over time: the longer someone looks at a page, the more their judgement rests on familiar structure rather than the initial visual hit. An original layout can win the first second and lose the next thirty.
Which leads somewhere uncomfortable for designers: originality isn’t a value in itself. It works only where the audience is willing to spend attention decoding — a creative studio’s portfolio, say. On a dental clinic’s site it does damage.
Plenty of whitespace, a restricted palette, large typography, minimal decoration. The style that fits the research above best — by definition it carries low visual complexity.
Fits: technology companies, consulting, premium services, any business with one clear proposition.
Breaks when: there’s actually a lot of content. Minimalism across a thousand-product catalogue turns into empty screens and endless scrolling. Another trap: minimalism doesn’t forgive weak content — when a page holds three elements, each has to be flawless, and a poor photograph is instantly obvious.
Symmetry, serif typefaces, restrained palette, tidy block borders. Signals stability and experience.
Fits: legal and financial services, medicine, education, long-established companies, B2B with a conservative audience.
Breaks when: the audience skews young or the sector expects speed. Classic easily reads as “this site hasn’t been updated in years” if you overdo the decorative elements.
Asymmetry, gradients, large imagery, scroll animation, unconventional grids. Looks current and expensive when done well.
Fits: creative agencies, startups, products for younger audiences, portfolios.
Breaks when: performance matters. Animation and heavy imagery drag on speed, which affects both conversion and rankings — we covered that in the piece on speeding up a site. The second risk is low prototypicality: unconventional navigation confuses people.
Geometry, high-contrast pairings, gold and deep colours, ornamental detail.
Fits: restaurants, hotels, beauty, jewellery brands, events — sectors where emotion matters more than decision speed.
Breaks when: almost anywhere else. This is a high-complexity style by definition, and it should be chosen deliberately, understanding that you’re trading convenience for character.
| Style | Complexity | Who for | Main risk |
|---|---|---|---|
| Minimalism | low | services, tech, premium | looks empty with a large catalogue |
| Classic | medium | legal, medical, finance | reads as dated |
| Modern | medium-high | creative, startups | speed and unfamiliar navigation |
| Decorative | high | hospitality, beauty, events | complexity hurts conversion |
Who is your customer and what have they seen before you? People judge your site against the ones they use daily. If your audience is accountants who spend the day in banking dashboards, bold modern design will unsettle them rather than impress.
How much content do you have? The most underrated factor. A style has to survive the real volume: not three perfect blocks from a mockup, but forty service pages and a blog that grows for years.
Who will update the site? A complex bespoke design needs a designer every time a page is added. If a manager with no design skills maintains the content, you need a system of simple repeatable blocks, not a unique layout per page.
The working scheme is simpler than expected: one primary colour, one accent, and a neutral base of two or three greys. That’s it. The accent gets used sparingly and deliberately — on buttons and key elements — and that’s exactly why it works: make half the page vivid and there’s no accent left.
The most common palette mistake is insufficient text contrast. Light grey on white looks delicate in a mockup and simply doesn’t read on a real screen in daylight. The minimum check: open the page on a phone outdoors. If you’re squinting, contrast is too low, and no aesthetic justifies that.
On industry colour associations: they exist, but they’re overrated. Blue does dominate finance and medicine, green in eco-adjacent sectors — but that’s statistics, not a rule. Standing out against competitors with a different colour is often worth more than blending in.
One typeface for headings, one for body text — that covers almost everything. A third makes sense only as a display face for isolated elements.
What genuinely affects readability: body text size (16 pixels is the working minimum on mobile, not 14), line height about one and a half times the type size, and line length in the 50-75 character range. Long lines on wide monitors are hard to read — the eye loses the start of the next one.
Serifs suggest tradition and solidity, sans-serifs modernity and technology. But the perceptual difference is smaller than the difference in typesetting quality: a carefully set sans beats a carelessly set serif.
The cheapest way not to get it wrong is not to commission every page at once. The working order: design the home page and one internal page — a service page, say. That’s enough to see the style at work on both a dense screen and a text-heavy one.
Then do what owners rarely do: show the layout to several people from your target audience. Not colleagues, not family — people resembling your customer. Ask specific questions rather than “do you like it”: what does this company do, where would you click, what’s the main thing here.
At Netloria we recommend this step even to clients who feel certain about their choice. It takes a few days, costs nothing, and prevents rebuilding an entire site six months later. The gap between “the style didn’t land” discovered across two mockups versus forty finished pages is a difference of many multiples in budget.
Style most often gets chosen on “the owner likes it” — and that’s the leading reason sites need rebuilding within a year. The owner looks at a layout as someone who knows the product inside out, holds all the context, and has already concluded the company is good. Their customer sees the page for the first time, without context, often on a phone, between other tasks, and decides within seconds whether to spend any time on it. Those two viewpoints almost never coincide, which is why “do I like this” is worth replacing with “would a stranger understand what we offer in five seconds.” At Netloria, during design approval, we often ask clients to show the layout to someone with no connection to the business and ask one question: what does this company do? The answer is frequently a surprise.
The second common mistake is copying a competitor’s site wholesale. You don’t know whether it works: possibly their enquiries come from advertising and personal contacts, and the site simply exists.
Trends move fast: gradients one year, brutalist type the next, then flat minimalism again. But the things that determine results stay the same for years.
Treat a trend as seasoning — one or two elements, not the foundation. Then the site won’t need redrawing when fashion moves on.
A realistic horizon for a site’s design is three to five years. What changes after that isn’t so much fashion as technical expectations: screen sizes, connection speeds, user habits.
The signals that it’s time are almost always practical rather than aesthetic: the site looks wrong on newer devices, adding a new page type without a designer is impossible, conversion drops with no visible cause. If the style has merely “grown stale for the owner,” that’s the weakest of all arguments for a redesign, and the money is usually better spent on content or advertising.
If you have a brand book, the style question is largely settled: palette, typefaces and character are defined, and the site follows them. Without one, the site effectively becomes your first brand asset — in which case it’s sensible to settle the basic elements first and draw pages afterwards.
Doing it the other way round is possible but costlier: when a presentation, signage and social profiles are needed six months later, you end up reverse-engineering logic out of a finished site instead of having it written down. We covered what a logo and brand book cost and when they’re worth doing before the site.
If you’re at the style-selection stage and would rather not guess — tell us about your audience and content volume. We’re a web studio based in Ukraine, and we’ll propose a direction that survives both your real page count and your customers’ expectations.
