A good website does one job well: it gets a specific visitor to take a specific action faster than the next tab they are about to open. Everything else, the palette, the animation, the award-winning layout, is either serving that job or quietly getting in the way.
What makes a good website: the short answer
Ask ten designers and you will get ten lists. The lists are not wrong, they are just downstream of a single question nobody asks first: what did this person come here to do? A plumber's site exists to produce phone calls. A SaaS site exists to produce trial signups. A charity site exists to produce donations and volunteers. Once you fix the job, the quality bar stops being subjective. A page is good if it moves someone through that job with less friction than the alternatives, and bad if it does not, regardless of how it looks in a portfolio. The most-awarded studio website in the country is a bad website if it never generates an enquiry for the studio that built it. Our web design work starts from that premise: design serves the business outcome, and the aesthetics follow.
The six qualities of a good website
Strip away the trends and effective websites share the same six characteristics. They are not a menu to pick from. Miss one and the other five lose most of their value, because a fast site with unclear messaging still fails, and a beautifully written site nobody can navigate still fails.
Clear purpose and messaging
A visitor should know within about five seconds what you do, who you do it for, and what to do next. Vague taglines like "delivering excellence" say nothing. Specific, plain-English headlines say everything: "Roof repairs across Somerset, usually fixed within 48 hours" tells a visitor the service, the area and the speed in one line. Clarity is the highest-leverage design decision you will ever make, and it costs nothing but the discipline to be specific.
Navigation that follows customer logic
Menus should mirror how customers think, not how your business is organised internally. Group services the way a buyer searches for them rather than the way your departments are structured. Keep the main navigation under seven items, label every link in customer language rather than internal jargon, and make sure any page is reachable in three clicks or fewer from the homepage.
Speed and Core Web Vitals
Google measures Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift on real visits, and uses them as a ranking input. Mobile users abandon slow pages long before they consciously decide to. The fixes are well established: properly compressed AVIF or WebP images, lazy-loaded video, decent hosting and no plugin bloat. Google's Core Web Vitals documentation sets out the thresholds each metric has to clear.
Mobile-first responsive design
Most UK web traffic now arrives on a phone, so designing on a desktop and shrinking it down is the wrong order. Design for a 390-pixel-wide screen first and let the desktop layout breathe outwards from there. Test every form on a real handset before launch. If you cannot reach the call-to-action with your thumb in one motion, neither can your customers. Our guide to mobile friendly website design covers how to test this properly now that Google has retired its Mobile-Friendly Test.
Conversion-focused copy and calls to action
Every page needs one primary action (book, buy, ring, request) and that action should be visible without scrolling. Copy should sell the click before it asks for it: tell the visitor what happens next, how long it takes and what they get. "Submit" is the worst button label ever invented. "Get my free quote in 60 seconds" is the same button doing real work.
Trust signals and accessibility
Trust closes the deal: real testimonials with names and photographs, recognisable client or accreditation logos, a verifiable UK address, an SSL padlock and contact details that are easy to find. Accessibility belongs in the same bracket. Meeting WCAG 2.2 at AA level widens your audience, and the same structural habits that help a screen reader (real headings, descriptive link text, labelled form fields) also help Google parse the page.
What makes a good web page, as opposed to a good website
These are different questions and conflating them is why plenty of redesigns disappoint. A website is judged on whether the whole system produces outcomes. A page is judged on one visitor, arriving from one source, with one question. Most of your traffic does not land on your homepage, so most of your traffic is forming its opinion from a single page that has to work on its own.
A good web page does five things. It matches the intent of whatever brought the visitor there, so a page reached from the search "emergency plumber Taunton" opens with availability and a phone number rather than company history. It answers the core question above the fold, before any scrolling is required, because a visitor who has to hunt for confirmation they are in the right place usually does not. It carries one primary action rather than competing with itself. It stands alone, meaning it does not assume the visitor read anything else first. And it links onward sensibly, offering the obvious next step rather than a wall of related links.
The practical test is to open your top five landing pages in GA4, look only at pages where traffic arrives from search or ads, and ask of each one: if this were the only page of mine anyone ever saw, would it produce an enquiry? Pages that fail that test are usually not badly designed. They are usually written as though the visitor had already read the homepage.
What makes a website appealing without making it slow
Visual appeal is real and it is worth being honest about it. Visitors form an impression of credibility in well under a second, long before they have read anything, and that impression sets how generously they interpret everything after it. The mistake is not caring about appearance. The mistake is assuming appeal comes from decoration.
What actually reads as appealing is mostly restraint and consistency. Generous white space so the eye knows where to land. A type scale with three or four sizes used consistently rather than a dozen used arbitrarily. Two or three brand colours, with one reserved exclusively for actions so a button is always recognisably a button. Real photography of real people, premises and work, because customers recognise stock imagery instantly and quietly discount everything near it. Alignment to a visible grid. Loading behaviour that does not shift the layout under a reader's thumb.
None of that costs load time. What does cost load time is the stuff that usually gets added in pursuit of appeal: autoplay video headers, multi-megabyte hero images, animation libraries loaded for one effect, three webfont families at nine weights, a carousel almost nobody scrolls past the first slide of. The trade is worth measuring rather than assuming. If an element cannot survive the question "what does this add, and what does it cost in Largest Contentful Paint?", it is decoration.
The same logic applies to studying examples. Looking at sites you admire is useful, but copy the structure rather than the surface. Note where the primary action sits, how many words appear before it, how the navigation is grouped and what proof appears near the point of decision. Those choices transfer between industries. A colour palette borrowed from a business with a different customer usually does not.
What good website content looks like
Design gets the credit and content does the work. Most pages that convert badly have a content problem wearing a design costume, which is why a rebuild that ports the old words across so often changes nothing.
- It is written to the customer, not about the company. Count the instances of "we" against "you" on your homepage. If "we" wins comfortably, the page is a brochure rather than a sales tool.
- It is specific enough to be falsifiable. "Fast turnaround" is unverifiable and therefore ignored. "Most repairs completed within 48 hours" is a claim a customer can hold you to, which is exactly why it is persuasive.
- It answers the objection rather than avoiding it. Price, lead time, minimum order, coverage area. Pages that dodge these lose the visitor to a competitor who did not.
- It is structured for scanning. Descriptive headings, short paragraphs, lists where lists are genuinely the right shape. Most people read a page in an F pattern before deciding whether to read it properly.
- It is current. Dated case studies, recent posts and accurate prices signal an operating business. Nothing undermines a polished design faster than a 2021 copyright line.
- It is unique to you. Content assembled from the same sources as everyone else gives a visitor no reason to choose you and gives Google no reason to prefer your page.
What makes a website bad
It is often easier to diagnose the inverse. Bad websites rarely fail because of one catastrophic error. They fail because five small frictions stack up and the visitor leaves without ever articulating why. These are the patterns that show up again and again in UK site audits:
- The homepage explains the company, not the offer. "Founded in 2009, we are a family-run business passionate about quality" answers a question nobody asked.
- Contact details are treated as a secret. A phone number buried on a separate contact page, behind a form, is a deliberate obstacle for a customer who is ready to buy.
- Mobile converts far worse than desktop. If desktop sessions become enquiries and mobile sessions do not, the mobile experience is broken, usually the layout, the tap targets or the form length.
- Nothing can be updated without a developer. If a price change or a new service page means a quote and a two-week wait, the platform is wrong for the business.
- The page is over four seconds on a 4G connection. Run it through PageSpeed Insights. If Largest Contentful Paint is red, a refresh-and-tidy will not save it, the foundation needs work.
- It still carries 2018 design tropes. Rotating hero sliders, a stock photograph of a smiling team in a meeting room, "we are passionate about delivering" copy. Customers read these as a signal that nothing has been maintained since.
- High bounce, low scroll depth. Visitors arrive and leave without engaging. Usually a messaging or speed problem, occasionally layout, almost never a shortage of animation.
Our longer list of common web design mistakes covers the ones that cost the most to fix after launch.
A 20-minute good website checklist
You can assess most of the six qualities yourself without any specialist tooling. Work through this in order on your own site, on a phone, with mobile data rather than office wifi:
- Open the homepage and count to five. Can a stranger say what you sell and who it is for? If not, stop here, this is the fix that matters most.
- Find the primary call to action without scrolling. Is there exactly one, and is it obvious?
- Tap the phone number. Does it dial? Tap the email. Does it open a mail client?
- Run the homepage and your top service page through Google PageSpeed Insights. Note the mobile score and which Core Web Vital fails.
- Navigate to your three most commercially important pages. Count the clicks. More than three from the homepage is a problem.
- Fill in your own enquiry form on a phone. Time it. Note every field you resented typing.
- Check the form actually arrives, and that it lands somewhere a human monitors rather than an unattended inbox.
- Look for trust: named testimonials, accreditations, a real address, a real photograph of a real person.
- Read the footer. Is the copyright year current? A stale year tells visitors the business may have folded.
- Search your own brand name on Google and click through. Does the page you land on match what the search result promised?
Anything that fails here is worth fixing before you spend money on traffic. Sending paid or organic visitors to a page with a broken conversion path is the most expensive mistake in digital marketing, and the easiest to avoid. For a deeper pass, work through our full website audit checklist.
How to measure whether your website is actually good
Everything above is a judgement until you attach numbers to it. A successful website is not the one that wins compliments, it is the one whose figures move. Five measurements tell you almost everything, and all five are available without buying anything.
- Conversion rate by landing page. The only metric that directly reflects whether the site does its job. Track it per page rather than site-wide, because a site-wide average hides both your best page and your worst.
- The mobile against desktop conversion gap. Same pages, different devices. A material gap is a user-experience failure, and it is usually the cheapest significant gain available.
- Core Web Vitals field data. Real-user LCP, INP and CLS from the Search Console report, not a one-off lab score. Lab data tells you what might be wrong, field data tells you what is.
- Organic impressions against clicks. High impressions with few clicks means you rank but your titles and descriptions are not earning the click. That is a copy fix, not a ranking problem, and it is fast.
- Enquiry quality, not just quantity. Track how many enquiries become customers. A form that triples submissions while halving close rate has made the business worse, and only tracking outcomes reveals it.
Set a baseline before you change anything. Redesigns that were never measured are the reason so many businesses cannot say whether their last one worked, and improving these numbers deliberately is what conversion rate optimisation actually is.
What makes a website credible
Credibility is the quality most often confused with polish. A slick site from an anonymous company converts worse than a plain site from an identifiable one, because buyers are not assessing your taste, they are assessing their risk. Four things do most of the work.
Specificity is first. Named clients, real project outcomes, actual numbers and dated case studies all survive scrutiny in a way that "trusted by businesses nationwide" does not. Second is evidence you did not write yourself: third-party reviews, accreditation bodies, trade memberships and press coverage carry more weight than any claim on your own page, which is why review widgets pulling live data outperform hand-typed quotes. Third is transparency about who you are, including a real address, a company number, named team members with photographs, and a straight answer on pricing even if it is only a starting figure. Fourth is currency: recently dated content, a working blog and a maintained footer all signal an operating business, and their absence signals the opposite.
None of this requires a redesign. Most of it is copy and proof, which is why credibility is usually the cheapest conversion gain available to a small business.
Why good web design is an investment, not a cost
The cheapest website on a five-year-old template is usually the most expensive one you can buy, because every paid click and every organic visit lands on a page converting at half the rate it should. A well-built site changes the arithmetic on every other channel at once. If your conversion rate doubles, your cost per lead halves and the same Google Ads budget suddenly delivers twice the pipeline. The same is true of organic: rankings you already hold start producing more work without a single extra visitor. That is why we treat the website as the central asset rather than a marketing afterthought, and why a rebuild is often the highest-return item on a marketing plan even when the existing site "looks fine". Our UK website cost guide sets out what each budget band actually buys.
For sector-specific versions of the same argument, see our breakdowns of ecommerce website redesign for online retailers, web design for tech startups and web design for non-profits, each of which weights the six qualities differently.
The UK-specific requirements most designers skip
Three things catch out UK businesses commissioning a website, and most freelancers will not raise them unless you ask:
- GDPR and cookie consent. A compliant consent banner is not optional, and the wording of your consent flow directly affects the quality of your analytics data. Sloppy implementations either break tracking entirely or leave a compliance exposure. Use a recognised consent management platform rather than a free plugin bolted on at launch.
- Accessibility law. The Public Sector Bodies Accessibility Regulations apply to public-sector sites, and private-sector businesses are increasingly held to a comparable standard under the Equality Act 2010. WCAG 2.2 AA is the practical floor, and building to it is far cheaper than retrofitting it.
- Company details on the site. A UK limited company must display its registered company name, company number, place of registration and registered office address on its website. It is a small detail that is trivial to add at build time and awkward to be caught without.
- Hosting jurisdiction. If you handle UK customer data, hosting in a UK or EU data centre keeps the data-transfer conversation simple. US-only hosting, still common with cheaper builders, introduces questions you do not need to answer.
None of these change how the site looks. All of them determine whether it is fit for purpose. A competent partner raises them in the first conversation. A poor one mentions them a fortnight before launch, when fixing them costs real money.
How to choose a web design partner in the UK
Pick a partner who asks about your business before they talk about colours. The right questions before quoting are: what does a qualified customer look like, where do they currently come from, what happens after the form is filled in, and how does the website fit the rest of your marketing? If a designer goes straight to a moodboard, they are decorating rather than solving a commercial problem. Ask to see live UK projects with measurable outcomes, and confirm who actually does the work as opposed to who runs the discovery call. It is also worth asking what happens after launch, because a site nobody maintains starts decaying within months, and ongoing SEO is what turns a good build into compounding traffic.
Frequently asked questions
What are the 7 C's of website design?
The 7 Cs come from Rayport and Jaworski's e-commerce customer interface framework: context (layout and design), content (what is actually on the page), community (user-to-user interaction), customisation (personalisation to the visitor), communication (site-to-user messaging), connection (links to other sites) and commerce (transaction capability). It dates from the early 2000s and is more useful as an audit checklist than a design method, since it says nothing about speed or mobile.
What are the five golden rules of a website?
There is no official standard behind the phrase, but five rules survive scrutiny across every industry. Make the purpose of the page obvious within five seconds. Give each page exactly one primary action. Design for the phone first, because that is where most visitors and Google both arrive. Load fast enough to clear Core Web Vitals on a mid-range handset. And prove you are real with named testimonials, a genuine address and current content. Everything else is detail on top of those five.
What are the 5 criteria for evaluating websites?
The classic five come from information-literacy teaching and assess a site as a source: accuracy (is it correct and verifiable), authority (who wrote it and what is their standing), objectivity (is there a bias or sales agenda), currency (how recently was it updated) and coverage (is the treatment complete). For evaluating a commercial website as a business asset the criteria differ: clarity of purpose, usability, mobile performance, credibility and conversion. Be clear which of the two questions you are asking, because they reward different things.
What makes a good web page rather than a good website?
A website is judged on whether the whole system produces outcomes; a page is judged on one visitor arriving with one question. Since most traffic lands somewhere other than the homepage, a good page has to match the intent that brought the visitor, answer the core question before any scrolling, carry a single primary action, make sense without anything read beforehand, and offer an obvious next step. Test it by asking whether that page alone would produce an enquiry.
How much does a good business website cost in the UK?
A bespoke small-business website typically lands between £3,000 and £10,000, with content-heavy or e-commerce builds running £8,000 to £25,000 and beyond. Anything materially cheaper is usually a template with light customisation, which can be the right answer for a very early-stage business.
Does web design affect SEO?
Directly and substantially. Site speed, mobile usability, heading structure, internal linking, crawlable navigation and accessible markup are all design decisions that Google reads. A redesign that ignores them can lose rankings a business spent years building, which is why the SEO work belongs at the wireframe stage rather than after launch.
