Small Business Web Design, Run by an SEO Consultant

Rankings do not live in the visual design. They live in URLs, text, the headings that label it, the links between pages, and how fast the browser paints something useful. Restyling touches none of that; rebuilding touches all of it. Four causes account for most of the damage:

• URLs changed and nothing redirects. Ranking history stays behind unless a 301 maps each old URL to its specific replacement. Sweeping them all to the homepage is commonly treated as a soft 404.
• Content was cut because it looked heavy. The new service page has a headline, three icons and a photograph — calmer to look at, with nothing left to match a search against.
• Headings became images. A typeface the web cannot render gets exported as a graphic. To a crawler and to a screen reader there is now no heading there at all.
• Navigation is drawn by JavaScript. A menu that renders only after scripts run, using click handlers instead of real <a href> elements, may never be followed — leaving interior pages effectively orphaned.

None of this shows up in a design review, because on screen everything is fine. It shows up in a crawl. So crawl the old site first, inventory every URL that earns anything, and make that inventory a build requirement. An SEO audit costs less before a rebuild than after.

One page per intent. If people search for three distinct things — emergency callouts, planned installations, annual servicing — those are three pages, each with its own language and call to action. Compressed into one homepage section, you get a page competing with itself and winning none of the three. The opposite error is as common: twenty near-identical pages differing by a swapped word give a search engine twenty reasons to trust none.

URLs that survive. A URL is a promise to every link, bookmark and citation that will ever point at it. Short, lowercase, hyphenated, descriptive, free of anything that dates — no years, no campaign names, no generated numbers. Decide the pattern once so the next rebuild need not break it.

Internal linking as layout, not cleanup. Links between pages are usually added after the design is finished, which is why they end up crammed into a footer nobody reads. They belong in the layout from the start: a designed related-services block, contextual links in body copy, breadcrumbs, and service lists pointing at real pages rather than anchors on the same one.

The test is simple: view the page with JavaScript disabled, or read the raw HTML the server returns. What is there is what you can rely on.

• Real text. Words in HTML — not baked into images or trapped in a canvas element. If it matters, you should be able to select it with a cursor.
• Real headings. One <h1> saying what the page is about, then <h2> and <h3> in order. Headings are a document outline, not font sizes.
• Real links. <a> elements with a genuine href. A <div> with a click handler navigates a mouse user and nobody else.
• Server-rendered content. Primary content and navigation in the HTML response, not assembled by scripts later.
• No infinite scroll for primary navigation. If services or locations only appear as more items load on scroll, there is no URL for page two. Paginate instead.
• No leftover noindex. The most expensive line in web design is the tag that protected the staging site and shipped to production with it.

Core Web Vitals measure what a page feels like to use. They are not a magic ranking lever — a fast page with nothing to say still will not rank — but they are recorded on real visits, and nearly everything that breaks them is a design decision, not a development one.

The fixes are unglamorous and all design-side. Size the hero to the space it occupies, preload the LCP image, and lazy-load only what sits below the fold. Set a font-display value so text stays readable while the webfont arrives, and reserve space for anything that loads late. Every third-party widget that looked harmless is paid for by a visitor on a mid-range phone.

“Responsive” has become a proposal line meaning the layout does not break, which is the lowest bar there is. Under mobile-first indexing the mobile rendering is the version search engines evaluate, so anything trimmed from the small screen is trimmed from the page.

• Tap targets with room around them. Sized for a thumb, not a cursor, and spaced so the phone number and the enquiry form are not close enough to mis-tap.
• Type you can read without pinching. Small grey-on-grey reads fine on a designer’s monitor and vanishes outdoors on a phone.
• The same content, not a reduced one. Collapse long sections into accordions if you must — the text still exists in the HTML — but do not strip it from the mobile build.
• Interruptions in proportion. A popup covering a small screen on arrival is bad for the visitor, and Google has discouraged it explicitly.
• Tested on a real phone. A resized desktop browser tells you nothing about thumb reach, sticky headers overlapping content, or the keyboard covering the field being filled in.

Accessibility is usually framed as a compliance tax on the design. Most of it is the same work that makes a page machine-readable, so structure that helps one helps the other.

• Landmarks. <header>, <nav>, <main> and <footer> instead of anonymous divs. Assistive tech jumps between them, and they make the main content unambiguous.
• Headings in order. The outline that helps a search engine is what lets a screen-reader user skim. Jumping from an <h2> to an <h4> for the smaller size breaks both.
• Visible focus states. Designers delete focus outlines because they interrupt a clean look. Design a better one instead, and keep every interactive element reachable by keyboard.
• Colour contrast. Pale grey text and low-contrast buttons are the most frequent failures, and they hurt everyone reading outdoors. Check ratios when you pick the palette.
• Alt text that describes. Say what the image shows and why it is there; decorative images take an empty alt attribute so they are skipped.
• Link text that stands alone. “Read the local SEO checklist” says where you are going; “click here” tells a person and a crawler nothing.

If customers find you by searching for a service near where they live, the site has a second job beyond persuading them: corroborating who and where you are, in a form machines can read. This is where the build meets local SEO and your Google Business Profile. How the requirements shift by sector is worked through in local SEO for dentists and local SEO for contractors.

• Name, address and phone number as real text. Identical in every page footer, and matching character for character what your Google Business Profile and directory listings say. If the profile says “Road” and the site says “Rd”, fix it.
• LocalBusiness structured data. JSON-LD describing the business, address, telephone, hours and service area — and only ever describing what is genuinely on the page.
• Location pages that actually differ. Each needs its own reason to exist: the services offered there, neighbourhoods a local would recognise, photographs of work nearby. The same text with the town swapped is built to be ignored — the thinking behind a page like local SEO in London rather than a list of interchangeable towns.
• Click-to-call everywhere. Numbers marked up as tel: links so a tap dials — for many local services the call is the conversion.
• Hours and service area in writing. Whether you cover their postcode, when you answer, whether there is a callout fee — answered on the page, not behind a form.
• A map, with the address in text beside it. Embeds are heavy and load late, so reserve their space — and keep the text address, which is what gets read and copied.

The first screen has one job: confirm the visitor is in the right place. Within seconds someone should be able to tell what you do, who for, where, and what happens next — a specific headline, a supporting line, and a visible way to make contact. Full-screen abstract imagery with a one-word slogan answers none of that, and is usually the heaviest LCP element on the page, so it costs you twice.

Forms should ask the minimum that lets you reply. Every extra field is a reason to abandon. A name, one contact method and a description of the job is enough; qualification can happen in the reply. Keep the form on the page rather than behind a link to a contact page, and say who replies and when. Offer another channel too — plenty of people would rather send a message than fill anything in.

Place the call to action where the offer has just been explained, not at fixed intervals. It works after the description of what the service involves, after the section handling the main objection, after the pricing. Identical buttons every three hundred pixels train people to ignore them. A page that explains the offer properly also has the vocabulary to match real searches — the local SEO checklist covers the rest.

Most of the frustration I hear about previous web projects has nothing to do with design. It is that the business cannot get at its own site: the domain sits in an agency account, the analytics belong to someone else, and changing provider means starting again.

• The domain is registered to you. Your name, your registrar account, your login. If someone else holds it, moving it is worth sorting out whoever builds the site.
• Hosting is in an account you control. You grant and remove access, see the bill, and can move the site without anyone’s cooperation.
• You get the code and the content. The files that make up the site, plus text and images you can reuse. Nothing locked to a licence only I can renew.
• Analytics and Search Console are yours. Created under your account, access granted to whoever works on the site.
• You can edit what changes. Wanting to update text or add pages yourself is a requirement to design for at the start.
• Platform choice is yours. Lightweight custom code, WordPress, or a hosted platform you already use — I set out the trade-offs on cost, speed, editing and leaving, and you decide.

If this is the arrangement you want, message Searcora on WhatsApp with what your business does and what the site needs to do — you get a written scope and quote before anything is charged.

A rebuild is disruptive and risks whatever visibility you have. If your site loads reasonably, reads well on a phone, and has pages that already produce enquiries, fixing beats replacing it.

If you are unsure which row you are in, that is a diagnosis rather than a design brief. An SEO audit will tell you whether the site is the constraint — and if it is not, I will say so.

Nothing starts with a price list, because a five-page site for one location and a site with a dozen location pages are not the same job. Small business web design is quoted per project, after a conversation covering the same ground every time:

• What the business does — services, the areas you cover, the customers you want more of.
• What exists now — your current site, where it is hosted, who controls the domain, and what is not working.
• Whether you need a rebuild at all — I would rather tell you to fix what you have than sell a project you do not need.
• What the site has to do — phone calls, form enquiries, bookings, or supporting a sales conversation elsewhere.
• Scope — how many pages, what content exists, what needs writing, what you want to edit yourself.
• How it fits with search — whether the build stands alone or runs alongside local SEO and Google Business Profile optimisation.

You then get a written quote with the scope in plain terms, so you see what is included before committing. If a rebuild is not your best next step, you get that answer instead. Searcora Digital Ltd is a UK registered company in London, and web design in London is quoted the same per-project way as a build anywhere else — the work is done remotely in both cases, so where you are changes nothing about the scope or the price.

Start the conversation. Message Searcora on WhatsApp at +92 300 2462807 or email contact.amirjutt@gmail.com with what your business does, the area you serve, and your website address if you have one. You get a straight answer on whether a rebuild makes sense, and a per-project quote if it does. Searcora Digital Ltd, 78 Dulverton Road, London SE9 3RL — work handled by Muhammad Amir, Founder & Local SEO Consultant: Top Rated Plus on Upwork, 600+ projects, 99% Job Success. If you would rather hire through Upwork for its payment protection, that works too — the work and the person are the same.