SEO website structure: how to get clients from Google without ads
How to collect search queries, distribute them across pages and build a structure that’s easy to rank. With specific queries, an example site map and AI prompts.

Contents
Imagine two dental clinic websites in the same city. Same design, similar prices, both “optimised”. The first gets enquiries every day, even with ads switched off. The second only gets them while it pays for clicks. The difference isn’t the copy or the number of keywords. The difference is the structure: the first has a page for every client query, the second has one “Services” page for everything.
In this article, I’ll show the approach I use to build SEO structures for websites: how to find the queries clients use to search for you, distribute them across pages and what to write on each page. With specific example queries, a ready-made site map and prompts that let AI do half the routine work for you. And at the end — the mistake that makes a website’s pages compete with each other so that none of them reaches the top.
Why structure matters more than “SEO copy”
Google doesn’t rank a website as a whole but specific pages for specific queries. When someone searches for “dental implants dublin price”, Google looks for the page that best answers exactly that. If implants are one paragraph on a general services page, you’ll lose to a website that has a separate page for them with prices, stages, guarantees and answers to questions.
So keywords in the copy are only the tip of the iceberg. The foundation is the right number of pages, their hierarchy, URLs and the links between them. Google’s own SEO Starter Guide recommends exactly this: group pages logically, use descriptive URLs and help people and the search engine find their way around the website.
No page for a query — no chance of reaching the top for that query. However good the copy is.
How people search: 5 types of queries
Before building a structure, it’s important to understand what a person wants to get when they type a query. That determines which page should answer them.
Commercial and price queries bring in enquiries the fastest, so we start with them. Informational queries bring traffic and trust and lead the person to the commercial page through internal links.
The approach: 7 steps to an SEO structure
Step 1. Collect your clients’ queries
Don’t guess how people search for you — look. Free sources that are enough to start with:
- Google autocomplete — start typing the name of a service and write down the suggestions from the drop-down list.
- The “People also ask” and “Related searches” blocks at the bottom of the results — ready-made questions for your FAQ and articles.
- Google Keyword Planner (in Google Ads, no need to run ads) — approximate search volumes.
- Google Search Console — if you already have a website: the real queries you’re shown for.
- Top-ranking competitors’ websites — which pages they have and what they’re called.
- Your clients — how they phrase questions in messages and on the phone.
Step 2. Group the queries by intent
This is the most important step, and it’s where most mistakes happen. The rule is simple: if queries mean the same thing to the client, it’s one page. If they mean different things, it’s different pages.
Step 3. Build a site map
Each group of queries becomes a page, and the pages form a tree. Here’s what the structure might look like for a hypothetical dental clinic in Dublin:
- No deeper than 3 clicks from the home page to any important page.
- Clear, readable URLs: /services/dental-implants, not /page?id=17.
- A page for every key service, not one page for everything.
Step 4. Fill the page with answers, not keywords
A service page in the top results is a page after which the client doesn’t need to go back to search. Here’s a brief for a “Dental implants in Dublin” page:
Step 5. Connect the pages with links
Internal links show Google which pages are important and how they’re related, and show people where to go next. The article “Implant or bridge: which is better” links to the implant and prosthetics pages. The service page links to prices and the dentist. Breadcrumbs help make the hierarchy clear.
Step 6. A blog for informational queries
Someone searching “implant or bridge which is better” isn’t ready to buy yet, but they will be in a week. A helpful article that honestly answers the question and leads to the service page is the cheapest way to meet a client before your competitors do.
Step 7. Local SEO and the technical foundation
- Google Business Profile — for local queries, it’s often the main source of calls. According to Google’s explanation, local results take into account relevance, distance and the business’s prominence, including reviews.
- The same name, address and phone number on the website, in your Google profile and in directories.
- Speed and a mobile version — most local searches come from smartphones.
- Indexing — sitemap.xml, Google Search Console connected, no important pages blocked from search.
- Structured data (Schema.org) — breadcrumbs, FAQ, organisation: helps Google understand the content.
Specific queries and answers: examples for different niches
The most useful exercise is to define, for each query, what exactly the person should see on the page. Here’s what that looks like in different niches:
Notice the pattern? A commercial query leads to a service page with a price, timeline, proof and a button. “How / what / why” questions lead to an article that answers them and leads to the service.
The mistake that makes pages compete with each other
I promised to tell you at the end — here it is. When several pages of a website are optimised for the same query (for example, the home page, the services page and a blog article are all targeting “dental implants dublin”), Google can’t tell which one to show. As a result, none of them ranks high. This is called keyword cannibalisation.
- Assign one main page to each important query — you’ll see this immediately in the structure table.
- Other pages can mention the topic but link to the main page rather than compete with it.
- If two pages answer the same thing, merge them.
Two weak pages targeting one query lose to one strong page. Always.
How AI helps build an SEO structure
AI doesn’t know real search volumes and can’t see current Google results — get that data from Keyword Planner, Search Console and search itself. But it handles the routine brilliantly: it expands the list of queries, groups them, suggests a structure and writes page briefs. Here are prompts you can use today.
You’re an SEO specialist. My business: [business description, services, city]. Make a list of search queries potential clients might use to look for my services on Google in [language]. Group them by type: 1. Commercial (want to order). 2. Price (asking about cost). 3. Local (with a city, neighbourhood, “near me”). 4. Informational (“how”, “what”, “why”, “which is better” questions). For each service — 10–15 queries. Write them the way people actually type into search: lowercase, no punctuation, with possible abbreviations. Don’t include search volumes — I’ll check them separately.
Check the resulting list in Google Keyword Planner: remove queries with zero volume and add the ones you found in Google’s suggestions.
Here’s a list of search queries for my website: """ [paste the verified queries — one per line] """ Group them by user intent: put queries that one page can answer into one group. Rules: - different services — different groups; - the price of a service — in that service’s group; - informational questions — separate groups for blog articles; - a query can’t be in two groups. For each group, specify: - the page type (service / category / article / contact); - the page name; - the main query; - the page URL (e.g. /services/service-name). Present the result as a table. Separately flag the queries you’re unsure about — I’ll check them in Google’s results.
Write a brief for the “[page name]” page of a website for [business description, city]. Main query: [query] Additional queries: [3–6 queries] Questions from the “People also ask” block: [paste from Google] In the brief: 1. Title (up to 60 characters) and meta description (140–160 characters). 2. H1 and an H2 structure for all page blocks. 3. What each block must contain so the client doesn’t have to look for answers elsewhere: prices, timelines, stages, proof, guarantees. 4. 6–8 FAQ questions with short answer outlines. 5. Which other pages of the website to link to internally. Don’t make up facts about the business — where my data is needed, write [ADD: what exactly].
Here are the pages of my website with their titles and H1s: """ [URL — title — H1, one page per line] """ Find pages that compete for the same search queries. For each conflict: - which pages conflict and for which query; - which page should be the main one; - what to change on the others: title, H1, content focus, or merge the pages. Answer as a table.
How long to wait for results
SEO isn’t a button but an investment. According to Google, in its video about hiring an SEO, it usually takes from four months to a year to implement improvements and see results. Local and narrow queries with low competition often bring the first enquiries sooner, competitive ones later.
The key point: the right structure is laid down once and works for years. Rebuilding a finished website for SEO costs much more than doing it right from the start.
Common mistakes
- One “Services” page for everything — no single service reaches the top.
- “Keyword” pages without substance — 200 words and the keyword 15 times. Google stopped ranking this long ago.
- No prices — for price queries, Google chooses pages that have prices.
- A made-up structure without analysing queries and competitors.
- SEO “later” — when the website is already finished and the structure has to be broken.
- A forgotten Google Business Profile for a local business.
How I build SEO into every website
In my work, SEO isn’t a separate “after launch” service but part of development from day one:
- Query and competitor analysis — what clients search for when looking for you.
- A page map — a page for each group of queries, already at the wireframe stage.
- Copy for queries and people — answers to clients’ questions, not a set of keywords.
- A technical foundation — speed, responsiveness, meta tags, sitemap, structured data, Google Search Console.
- A structure with room to grow — new services and articles are added without rebuilding the website.
This website is built exactly that way: a separate page for each type of website, a blog for clients’ questions and internal links between them. Details on cost are on the pricing page.
SEO structure checklist
- Queries collected from Google autocomplete, Keyword Planner, Search Console and clients.
- Queries grouped by intent, each in only one group.
- Every key service has its own page.
- Every important page is no deeper than 3 clicks from the home page.
- URLs are clear and readable.
- Unique titles, descriptions and H1s on every page.
- Service pages have prices, timelines, proof and an FAQ.
- Blog articles answer informational queries and link to services.
- No two pages target the same query.
- Google Search Console, a sitemap and Google Business Profile are connected.
Conclusion
Clients from Google without ads aren’t magic or “secret keywords”. It’s a website with a separate page and a precise, helpful answer for every client question. Collect the queries, group them by intent, build a map, fill the pages with answers and connect them with links. AI will save you hours of routine, but the decisions about structure and facts are yours.

