SEO web design, built so search engines can read each page from launch day

Who is SEO web design for?

SEO web design suits a business that wins work from people searching: for a service, for a place, or for the answer to a question it can give. The search work goes in while the site is planned and built, and it looks different at each size:

• Established local businesses: a single office or service area, competing with nearby firms for the same few searches. The site needs a page for each main service and for each area the searches back, with the business name, address and service area stated the same way on every page.
• Multi-location and regional businesses: several offices, or a service area across several counties or states. Each location gets a page with its own details, and the plan has to keep many similar pages from reading as copies of each other.
• National businesses: large sites, often hundreds of pages with several teams publishing. At that size, getting every page crawled becomes a planning question. Google’s sitemap guide says a site of about 500 pages or fewer, linked well internally and without many media files or news pages, may not need a sitemap, and that on large sites it’s harder to make sure every page is linked from another.

If your site has a handful of pages and few competitors, you may not need us. Google’s guide to hiring search help says someone running a local business on that scale can probably do much of the work themselves, starting from its SEO starter guide. If that’s you, we’ll say so on the call.

What does Google read when it visits a page?

Google sets three minimum requirements before a page can be indexed: Googlebot isn’t blocked, the page returns a working HTTP 200 status, and it has content Google can index. Meeting them makes a page eligible, nothing more. Google is plain that indexing isn’t promised. After that, a few build choices decide what Google actually sees:

• The phone version: Google indexes and ranks the mobile version of a page, crawled with its smartphone agent. Its mobile-first guidance asks for the same content on phone and desktop, and says Google won’t load content that only appears after a visitor swipes, clicks or types.
• Links: Google can generally crawl a link only when it’s an <a> element with an href attribute, and most links in other formats aren’t parsed. The link’s visible words tell people and Google what the page at the other end is about.
• Images: Google finds images in the src of an <img> element and doesn’t index CSS images. It works out what an image shows from its alt text, computer vision and the page around it, and takes very light clues from the file name.

Why does it matter how a page is rendered?

Google processes a JavaScript page in three phases: crawling, rendering and indexing. A page can sit in the rendering queue for a few seconds, or longer, before Google runs its scripts and sees the result. Google’s own advice still favors server-side rendering or pre-rendering, since pages arrive faster for visitors and crawlers alike, and not every bot can run JavaScript.

That’s how we build. Our recent sites run on Astro, set to produce every page as finished HTML when the site is built, so the words, links and markup are all in the first response a crawler gets. Scripts are kept to the jobs that need one, such as forms, booking and measurement.

A site that assembles its pages in the visitor’s browser can still rank, as long as Google renders it correctly. It adds a step that can go wrong, though, and some bots won’t run the scripts at all.

How do pages on one site avoid competing for the same search?

When two pages show much the same content, search engines have to choose between them. Google’s starter guide says they pick a single canonical URL to show for each piece of content, and asks site owners to make each piece of content reachable through one URL. Its canonical guide adds that grouping duplicates under one preferred address lets Google combine the signals, such as links, that each copy collects. On a site we build, that starts in the page plan:

• One search, one page: each page in the plan has its own primary search, and no two pages share one. A smaller question gets a section on the page that owns its topic instead of a thin page of its own.
• One question, one answer: each FAQ question is marked up on one page only, so the site never offers search engines the same answer from two addresses.
• One address per page: every page open to search carries a canonical tag naming its own address, with no trailing slash. Google treats redirects and canonical tags as strong signals of the preferred address, and a sitemap entry as a weak one.
• Links that say where they go: pages link to each other with words that describe the next page, never “click here”.

What should each page’s title, description and headings do?

Google says a page’s title link is often the main thing people use to decide which result to click. Its guidance asks for a <title> on every page that is descriptive and concise, avoids keyword stuffing and boilerplate repeated across pages, and keeps the brand short. Google builds the title link it shows from several sources, including the page’s headings, and asks that the main title stand out as the most prominent text on the page.

The meta description shapes the snippet only some of the time: Google writes snippets mainly from the page’s own content and uses the meta description when it describes the page better. That’s still a reason to write one for every page. The starter guide describes a good one as short, unique to one page and holding the page’s most relevant points.

Headings are for readers. The starter guide says heading order doesn’t matter from Google Search’s point of view, and that there’s no ideal number of headings. We write them as the questions buyers ask, in the order they’d ask them. Our build checks fail a page with more than one H1, and any two pages that share a title or a description.

How fast does a page need to load?

Google’s Core Web Vitals set the marks for a good experience: Largest Contentful Paint within 2.5 seconds of the page starting to load, Interaction to Next Paint under 200 milliseconds, so the page answers a tap or click quickly, and Cumulative Layout Shift under 0.1, so the layout holds still while it loads.

Speed isn’t the whole story. Google says there is no single page experience signal, that its systems always seek to show the most relevant content even when the page experience is sub-par, and that chasing a perfect score just for SEO reasons may not be the best use of your time. Where a search has lots of helpful pages to choose from, Google says a great page experience can contribute to success.

What we build for speed: JavaScript only where a page needs it, fonts served from the site itself so text doesn’t jump as it loads, and images turned into WebP files when the site is built. Before launch, we run Google’s Lighthouse checks on a phone setting for each type of page.

What does structured data do for a page?

Structured data is markup, in a standard vocabulary, that tells search engines what a page is about: the business, the service, the questions the page answers. Google calls it explicit clues about a page’s meaning. It can make a page eligible for rich results, the listings that show extra detail, but only when every required property is present.

Two rules keep it honest. Google says markup has to describe content the visitor can see on the page, never information that isn’t visible, even when it’s accurate. And markup can break after launch through a template or serving problem, so Google suggests the Rich Results Test during development and Search Console’s reports afterwards.

Our build checks that every page’s markup parses, that a page showing questions and answers carries FAQ markup, and that no page carries FAQ markup without the questions on it. At launch, we run the Rich Results Test on the live pages.

Which SEO advice can you safely ignore?

Some common SEO advice is out of date or mistaken. Google’s starter guide lists topics it believes you shouldn’t focus on:

• The keywords meta tag: Google Search doesn’t use it.
• Keyword stuffing: repeating a phrase to rank is against Google’s spam policies.
• Keywords in the domain or address: on their own, they have hardly any effect beyond appearing in breadcrumbs.
• A word count: length alone doesn’t matter for ranking. Google’s helpful content guide asks whether you’re writing to a word count because you heard Google prefers one, and answers that it doesn’t.
• Heading order: from Google Search’s point of view, headings out of order don’t matter.
• A duplicate content penalty: Google says the same content at more than one address on your site is inefficient, but won’t cause a manual action. Copying another site’s content is a different matter.
• E-E-A-T as a ranking factor: Google says it isn’t one.

Two more from Google’s guide to hiring search help: advertising with Google has no effect on a site’s presence in its search results, and it costs nothing to appear in its organic results.

How can you check a web designer’s SEO claims?

You don’t need to read code. Ask for these on a site they’ve built:

• The words in the page source: open a page, view its source, and look for the main text. If it isn’t there, the page depends on scripts to show it.
• A distinct title on every page: compare the browser tabs of a few pages. Titles that repeat mean a template is setting them.
• How the preview stays out of search: an unfinished copy that gets indexed can compete with the real site. Ours is marked noindex until launch.
• Whether they follow Google’s rules: one of the questions Google’s guide to hiring search help suggests asking is whether they follow the Google Search Essentials. The same guide warns against anyone who claims a privileged tie to Google.

Then hold one line firm. Google’s guide says plainly that nobody can promise a site the first result in its search, so anyone who does is offering what Google says isn’t theirs to give. We do not guarantee search rankings, traffic, or uninterrupted uptime.

What happens to search traffic when a site is rebuilt?

Some movement is normal. Google’s guide to site moves says to expect temporary ranking fluctuation while it recrawls and reindexes, and that on a medium-sized site the switch to the new addresses can take a few weeks or more to show. Permanent redirects from each old address to its new page keep the old page’s value: Google says 301 redirects don’t cause a loss in PageRank. It also suggests making one change at a time: a new domain, a new platform and a new layout, one after another.

The details live on the pages for each case: website redesign for a new look on the same business, and website migration for moving a site off its current platform into our system.

Where does SEO web design stop and ongoing SEO begin?

The build controls what’s on the site: which pages exist, what they say, how they’re coded and how they link. It can’t control the rest. Search Console’s help says not to expect every page to be indexed, that a page left out can be fine for the right reasons, and that an indexed page still won’t show in every search, or at the same position, because results vary with each person’s history, location and more.

After launch, the work moves to what happens off the site and over time: pages for the searches the plan left for later, links from other sites, local listings and the questions AI answers draw on. Our SEO services cover that, with technical SEO for crawl and index problems and on-page SEO for pages already live. Generative engine optimization is the work of making a business easier for AI answers to find, read and cite.

From launch on, our monthly plan looks after the site, and its scope is written into your proposal. Our page on website maintenance covers the upkeep side.

What does SEO web design look like on a live site?

Two examples. The first is a client’s: the Alvy Logistics website, a 70-page site we rebuilt in our own system on Astro and run with Alvy today. Alvy is owned by our co-founder Noah Alvarez, and DSI is an Alvy company.

The second is the page you’re reading. View its source and you’ll find the words in the HTML, one main heading, its own title and description, a canonical tag naming its address, and structured data for the service, its questions and its place in the site’s breadcrumb trail. The rest of this site comes out of the same build and the same checks.

If you want a site search engines can read from its first day, book a 15-minute call. Name the searches you most want to be found for, and we’ll start from those.