SEO Website Design Services

I start with an audit even on a new build if the client has an existing site. For a recent 200-page B2B site redesign, I found 30% of pages had thin content and 12% had broken internal links. That data shaped the page retention and rewrite decisions. Without the audit, we would have rebuilt bad data.

Next I create a keyword-driven sitemap. I list every target page, its primary keyword, and the user intent — informational, navigational, or transactional. For a SaaS client that meant 15 landing pages for transactional keywords, 40 blog posts for informational queries, and 5 core product pages. That sitemap becomes the blueprint for IA and URL structure. I insist on a heading hierarchy that mirrors that sitemap: one H1 per page matching the primary keyword, H2s for subtopics, and H3s for supporting details.

During wireframing I set performance budgets: Largest Contentful Paint (LCP) under 2.5 seconds, Total Blocking Time under 50 milliseconds. I test every template against those budgets using Lighthouse before any developer writes CSS. One client wanted a fullscreen video hero; I proved it pushed LCP to 6 seconds. We compromised on a static hero image with a video on a secondary page. That saved the redesign.

I also deploy structured data early. For the SaaS site I added [BreadcrumbList](/schema-markup/) and FAQPage schema to the templates. That meant no retrofitting after launch. The result: organic traffic grew 35% in four months on a £25,000 budget. Would I do it differently? I would push harder for a content-first design: write the H1, meta description, and page copy before the designer picks a colour palette. That alignment saves rounds of revision.

The biggest watch-out is treating SEO as a post-launch add-on. I have done it. You end up rewriting templates, fixing broken internal links, and adding schema after the fact. That retrofitting costs roughly twice the time and money of doing it upfront. For one e-commerce site, fixing crawlability after launch took three months and £15,000 — the original redesign was £30,000.

Second watch-out: designing for aesthetics over speed. A visually stunning site that takes 8 seconds to load will lose rankings and conversions. But I have seen the counter-argument: luxury brands need immersive experiences. The decision rule I use is: if your primary traffic source is organic search, prioritise [Core Web Vitals](/core-web-vitals/) and crawlability. If brand recognition drives most of your traffic, you can bend performance but not break it. For example, use server-side rendering for the hero section and lazy load decorative assets.

Third: assuming responsive design alone equals mobile SEO. Google uses mobile-first indexing, meaning your mobile version must have equivalent content, structured data, and internal linking. I have seen sites where the mobile version stripped out the schema markup and hid long-form content behind accordions. That killed visibility. Test your mobile version separately with Google's mobile-friendly test and a crawl of the mobile DOM.

Fourth: trusting every developer understands SEO. I worked with a team that built a JavaScript-heavy navigation using client-side rendering. The crawler saw only a blank shell. I now require [JavaScript SEO](/technical-seo/) knowledge in the dev brief or specify server-side rendering for critical content. Also, ensure your [website structure](/website-structure/) uses clear, crawlable internal links — no buttons or dropdowns that only fire on user interaction.

Fifth: ignoring canonicalisation during design. A common mistake is having the same content accessible under multiple URLs (e.g., www and non-www, with and without trailing slashes). Set a [canonical tag](/canonical-tag/) strategy from template one. I saw a site lose 20% of indexed pages because the CMS generated duplicate URLs for category filters. A canonical tag per template prevented that.

Early in my career I worked on a single-page application (SPA) for a news site. The design team insisted on a rich client-side rendered interface with complex JavaScript transitions. I did not involve SEO until the site was live. The crawlers saw an empty shell. Organic traffic dropped by 80% within a month. We spent three months and £30,000 migrating to a hybrid renderer using Next.js. The mistake was not setting a technical requirement for server-side rendering during the RFP. I now put SSR or dynamic rendering in the first draft of any project brief.

I also got mobile SEO wrong. I used to believe that if a site was responsive, it was automatically ready for mobile-first indexing. I learned that responsive is a necessary condition but not sufficient. Google indexes the mobile version first, so if your mobile site hides content behind tabs, removes structured data, or has different internal links, you lose visibility. I now crawl the mobile DOM separately using a mobile user-agent in Screaming Frog before launch. One client's mobile site had no schema markup and omitted the breadcrumb trail. That explained why their site dropped from position 3 to 12 after the mobile-first indexing rollout.

Finally, I overestimated what web design agencies know about SEO. I assumed any reputable agency would include SEO best practices in their standard workflow. They do not. Most focus on aesthetics and user experience for humans, not for crawlers. I now write a clear scope of work that lists specific deliverables: a keyword-driven sitemap, a template-level heading hierarchy, technical SEO requirements including meta tags and schema, and a performance budget. That document saves arguments later. One agency pushed back, saying SEO stifled creativity. I showed them a competitor who used a simple, fast design and outranked them. That changed the conversation.