An SEO example is a specific change to a website (a category page, the structured data on a recipe, one page per city, an online calculator) whose effect you can see on a real Google search. Below are nine: seven read on Google UK on a smartphone on 10 October 2026 with DataForSEO, and two from our own projects on Google Italy. For each search we give the monthly volume, who wins and why.
Finding SEO examples sounds easy.
Then you search "seo examples" on Google UK and the first page gives you an AI Overview built on Nike, NYT Cooking, NerdWallet and a bakery in Florida, a list of landing pages from Zendesk and HubSpot, and a gallery of website designs.
Real examples, with a website, a search and a number next to them? Hardly any.
When a business looks for a model to copy, the mistakes we see most often are three:
- taking the American giants as the example (Nike, Apple, NerdWallet), with budgets and brands no UK small business can replicate;
- confusing an "optimised website" with a "website that ranks": tidy titles and meta descriptions don't bring traffic on their own;
- copying the tactic without looking at the search, which means doing what a competitor did on a query that's worth nothing to you.
Of the nine organic results on that page, the only UK organisation is Shelter, with a guide to writing SEO content: no UK business shows what it did and what it got.
We take our examples from two places: the projects we run and the results pages we measure.
Macropix makes LED walls in Italy: on the word "ledwall" its page collected 67,560 impressions and 1,495 clicks in twelve months. Sunscape looks for land for solar farms in Italy: in six months it went from 1 to 38 keywords on page one.
And the other seven?
They're websites we don't work on, read from the outside: BBC Good Food, Taylors Estate Agents, Decathlon, Asda Money and others. For each one you'll find:
- the search and how many times a month it's made in the UK;
- what the winning page does (and what the losers don't);
- the lesson you can apply to a business website, even a small one.
Let's start.
What counts as an SEO example (and why an "optimised" website isn't one)
An SEO example only holds up if it has three parts: an action on the website, a specific search and a measurable result on that search. If one of the three is missing, it's good practice, not an example.
The reason is in the numbers.
In Ahrefs' study of around 14 billion pages (December 2023, global sample, not broken down by country) 96.55% of pages get no traffic at all from Google.
Almost all of them, we'd imagine, have a title and a meta description.
So an "example of an optimised website" shown with a screenshot of its tags tells you nothing: that's the invisible 96.55%. A useful example starts from the search and works back to the page that wins it.
That's how our cards are built. (Method at the end of the article, with its limits.)
The 9 examples in one table (mobile results, October 2026)
Nine searches, nine different levers: the table sums up the test, the sections below explain it one row at a time.
| Search | Searches a month | Who wins | SEO lever |
|---|---|---|---|
| ledwall (Google Italy) | 8,100 (Italy) | Amazon, then AGMultivision and Macropix | A category page that answers the question |
| vendo terreno per fotovoltaico ("selling land for solar", Google Italy) | 320 (Italy) | Millenergie, Terrenifotovoltaico.com | One page per intent and region |
| tiramisu recipe | 60,500 | BBC Good Food (first recipe card), Olive (first organic result) | Recipe structured data and reviews |
| vat calculator | 165,000 | Enerpize, an ERP software company | A calculator next to what you sell |
| houses for sale bristol | 14,800 | Greenwoods Property Centre, Rightmove, Taylors | Pages generated from a database for each city |
| decathlon manchester | 4,400 | Decathlon's map panel, then its Liverpool store page and Apple Maps | Store pages for a chain |
| sofa bed | 246,000 | Furniture Direct Online, then Bed Kingdom and an Ideal Home guide | Ecommerce category versus guide |
| how long does a passport last | 2,400 | Sources cited: Asda Money, Skyscanner, Post Office | Source of the AI answer |
| take home pay calculator | 301,000 | calculatemysalary.co.uk and eight more calculators | The format the results page asks for |
Volumes: Google Ads monthly average from September 2025 to August 2026, read with DataForSEO, for the United Kingdom (the first two rows are Google Italy, our clients' market). Google Ads rounds large volumes into bands, so read the biggest figures as orders of magnitude.
Example 1: the category page that answers the question (Macropix on "ledwall")
On "ledwall", 8,100 searches a month in Italy, the Macropix page is the third organic result on a smartphone on Google Italy, behind Amazon and a specialist online shop. It isn't a product page: it's a page that explains what a LED wall is, how big it is, how it's installed.
Google notices.
The snippet shown in the results is the definition, word for word: "Un ledwall, o LED wall, è uno schermo modulare composto da pannelli LED…" ("A ledwall, or LED wall, is a modular screen made of LED panels…")
And it's a crowded page: there's Amazon (twice), a block of products with prices up to €44,518, and a "Places" box with nine companies in the sector. Macropix appears both in the organic results and in the box.
The numbers from the Macropix case study, from September 2025 to August 2026:
- "ledwall": 67,560 impressions, 1,495 clicks, average position 5.5;
- "led wall": 27,868 impressions, 746 clicks;
- of the 135 tracked keywords, 79 are on page one and 28 in the top three.
The figure that explains the most is another one, though. On "monitor pubblicitario" (advertising monitor) the website went from position 88 to position 2, and it took five years (we tell the story in our guide to SEO for manufacturing).
The lesson: on category keywords a manufacturer doesn't beat Amazon on price. It beats it on the answer.
Example 2: one page per region and per intent (Sunscape)
Sunscape develops solar farms in Italy and looks for land: its customer is a landowner, who searches Google with the name of their region and the verb they know (sell, lease). A generic page caught none of those searches.
So we built the architecture on two intents, selling and leasing, with dedicated pages for each region.
The results, from the Sunscape case study:
- organic impressions from 329 in March to 5,809 in August 2026;
- monthly organic clicks from 46 to 195;
- keywords on page one from 1 to 38 out of 143 between July and September.
Job done, then?
No. On "vendo terreno per fotovoltaico" (selling land for solar, 320 searches a month) the average position between June and August was 6.6. On the smartphone results page of 8 October the page wasn't among the first nine organic results, and the AI Overview (the summary Google generates at the top of the page) cited Biosolar and Epico.
An average position is an average: on a given day, on a given phone, it moves. None of Sunscape's 143 keywords is in the top three yet, and that's the goal of the 2026-2027 link building plan.
The lesson: if demand changes from region to region, you need different pages. And you need months: the real timescales are in our article on how long it takes to see SEO results.
Example 3: recipe structured data (BBC Good Food on tiramisu)
"Tiramisu recipe" gets 60,500 searches a month in the UK, and BBC Good Food takes three places on the first page: the first card in the recipe carousel, an organic result with its tiramisu collection, and a citation in the AI Overview.
In the carousel the card shows 4 stars, 349 ratings and 30 minutes of preparation.
That information doesn't turn up by chance. Google reads it from the Recipe markup the page declares in its code, as explained in Google Search Central's documentation on recipe structured data.
The others in the carousel: Broma Bakery (4.9 stars, but on 38 ratings), RecipeTin Eats (4.9 on 215), Sally's Baking Addiction (4.8 on 191, 12 hours with chilling). BBC Good Food has the lowest score and the most ratings, and it comes first. The first organic result, Olive magazine, shows 5 stars on 5 ratings.
And the links?
Few, for a page like that. According to DataForSEO, BBC Good Food's best ever tiramisu recipe gets links from 30 domains. Here the difference comes from ratings collected over years and the markup that carries them into the results.
The lesson: if your content has a structured data type that Google displays (recipes, products, events, reviews), declaring it properly is worth more than a link. How to do it without duplicates is in our guide to structured data on WordPress, and what it actually does today in the one on schema markup.
Example 4: the calculator on a software company's website (VAT calculator)
On "vat calculator", 165,000 searches a month, the first organic result on a smartphone is a calculator by Enerpize, a company that sells cloud accounting and ERP software. According to DataForSEO the page was first seen in August 2026 and has a link from a single domain.
We expected the opposite.
In the Italian version of this test, the winner on the equivalent search was an old one-purpose calculator with links from over a thousand domains. In the UK the most linked calculator, vatcalculators.co.uk (4,825 backlinks from 150 domains according to DataForSEO), isn't among the organic results on page one: it's the first source cited in the AI Overview.
Look at who else the AI Overview cites, and two variants stand out:
- Xero, KashFlow and Clear Books, three accounting software brands, each with its own VAT calculator: a B2B company using a tool to catch a search right next to its market (invoices and VAT returns);
- Blue Arrow, a recruitment agency, and Architecture Social, a jobs and careers platform for architects, are both cited in the AI Overview for "take home pay calculator" (301,000 searches a month) with their own salary calculators. Someone weighing a job offer is exactly who they want to reach.
The lesson: a calculator works if the problem it solves is close to what you sell. Otherwise it brings links and visits that will never become customers. How much links still matter, with data, is in our guide to link building.
Example 5: pages generated for each city (estate agents and Rightmove in Bristol)
"Houses for sale bristol" gets 14,800 searches a month, and the first page is shared between portals and agents: Greenwoods Property Centre first, then the AI Overview citing Rightmove, Zoopla and OnTheMarket, then Rightmove, Taylors Estate Agents, House + Co, Davies & Way and Zoopla. They're pages generated from a listings database, one for each city and area.
Look at the snippets:
- Taylors Estate Agents: "402 properties for sale in Bristol";
- Rightmove: "The average price for a property in Bristol is £392025 over the last year";
- Greenwoods: "1 Bed Flat For Sale. £240,000".
The number in the snippet updates itself, and it tells the searcher the page is alive. It's programmatic SEO: one page template multiplied across thousands of places, which holds up because every page has different data (listings, prices, neighbourhoods). On Google UK it isn't only the portals doing it: two estate agent chains, Greenwoods and Taylors, sit above Zoopla.
Can a small website do it too?
Yes, but with a clear limit. Google treats scaled content abuse as spam when pages exist only to manipulate rankings. A thousand "plumber + town" pages with the same text and the town name swapped are exactly that.
The lesson: city pages work when there's different data behind every city. How an estate agent beats the portals in its own area is covered in our guide to real estate SEO.
Example 6: a chain's store pages (Decathlon on "decathlon manchester")
Anyone searching "decathlon manchester" (4,400 times a month) sees a panel at the top for the Manchester store, with 4.5 stars from 850 Google reviews. So far, so good.
Then come the organic results.
First is Decathlon's store page for Liverpool. Then Apple Maps, a customer service article on finding your nearest store, and the store page for Ealing Broadway, in London.
Further down: an Instagram account for Decathlon Stockport, a Manchester Evening News article from 2013, and Yell with its "Decathlon Manchester-Eastlands" listing.
Decathlon ranks with stores 30 and 200 miles from Manchester.
A Decathlon store page for Manchester doesn't appear among the organic results; the Stockport one turns up only in the images.
The lesson: even a huge brand loses its own local search if the store page isn't the most relevant page on the website for that city. For a business with several locations the work starts here, and you'll find it in our guide to local SEO.
Does your brand search end up on a directory?
If people searching your name with their city find Yell before they find you, you're paying for awareness and collecting only half of it. In 30 minutes we look together at the searches for your brand in Search Console and the pages that win them today.
Book a strategy consultationExample 7: the ecommerce category versus the guide (sofa bed)
"Sofa bed" gets 246,000 searches a month in the UK at an average cost per click of £1.01 on Google Ads (converted from US dollars at the Bank of England rate for 8 October 2026, $1.3226 to the pound): it's an ecommerce search, and the first organic result is a category page from Furniture Direct Online, with the price in the title ("From £379") and 4.7 stars in the snippet.
And the big names?
ScS, one of the largest UK sofa retailers, is on page one with an internal search results page whose snippet reads "No information is available for this page". It wins on its domain, not on relevance.
The most instructive result is a guide: Ideal Home's "The 9 best sofa beds out of the 100 I've tried and tested", updated in March 2026. The snippet gives a concrete fact: over 100 sofa beds tested.
Above the organic results, a Popular products block shows prices from £211.50 (Argos) to £1,299 (Dreams, 4.6 stars on 951 ratings).
Three ways to be on the same first page: the strongest domain, the most useful answer, the snippet with the most data.
The lesson: if you can't beat the big retailers on the category, you can beat them on the question that sits under the category. How to build categories and product pages that sell is explained in our guide to ecommerce SEO.
Example 8: the source the AI cites without being on page one (passport)
On "how long does a passport last" (2,400 searches a month) the AI Overview answers with two numbers, 10 years for an adult passport and 5 for a child's, and cites Asda Money, Skyscanner, the Post Office, AllClear and Citizens Advice.
Now look at the organic results: London Southend Airport, a GOV.UK news story from 2017, GoCompare, Oasis Insurance, the3million, CIBT Visas, UK Passport Checker, TravelSupermarket and TUI Ireland.
None of the five sources the AI cites is among the nine organic results on page one.
Yet the answer is theirs.
The Post Office page writes the figure in plain words: "Passports issued after 2018 are valid for exactly 10 years", followed by the exception for older passports and the two rules for travel to Europe. It's the easiest format to extract for a system that has to summarise.
The lesson: on searches with a short answer, being the cited source matters as much as ranking. What changes in clicks and how to measure it in your own data is in our article on the impact of AI Overviews on SEO; it's the work we do with SEO for AI.
Example 9: when the results page asks for a calculator (take home pay calculator)
"Take home pay calculator" gets 301,000 searches a month at a cost per click of £2.04, twice that of "sofa bed". There are dozens of calculators competing.
The first organic result on a smartphone is calculatemysalary.co.uk, with "HMRC's confirmed 2026/27 tax bands" in the snippet. All nine organic results on page one are calculators, and one still advertises 2025-26 rates.
So is a calculator always the answer?
No. On the Italian equivalent, "calcolo stipendio netto" (net salary calculation), the first two results were a Reddit thread and a tax forum, because people working out their net pay often have a case of their own (a job offer, a contract, a bonus) that a generic calculator doesn't solve. In the UK the same need is met by tools: they ask for tax code, pension and student loan plan.
The AI Overview points first to the official GOV.UK Income Tax and National Insurance estimator, then to The Salary Calculator and Listen to Taxman, which are not among the organic results.
The lesson: before you build a tool, look at what wins the results page. The same need can be answered by a forum in one country and a calculator in another. Reading a results page like this is the basis of any keyword research.
What no longer counts as an SEO example (FAQs with the questions in the results)
Many SEO guides still show, as an SEO example, the FAQs that open up under a Google result: since August 2023 that rich result hadn't been shown for most websites, and since 7 May 2026 Google doesn't show it at all.
Google announced it on the Search Central blog in 2023: FAQ rich results limited to "well-known, authoritative government and health websites", and step-by-step (HowTo) results removed. In May 2026 it went further: the FAQ documentation changelog says the feature "will no longer appear in Google Search starting May 7, 2026".
FAQs on the page are still useful, but for other reasons: they answer real questions and give AI summaries material to work with. We cover this in our article on what FAQs are for after the end of rich results.
The same goes for screenshots of an "optimised website" with a plugin's green light. A score of 100 out of 100 isn't an example: it's proof that the plugin is happy.
How to read an example before you copy it (three questions)
Before you replicate any of these examples, ask yourself three questions: is the search worth anything to your business, does the result depend on the brand, and is the figure measured or just seen?
- Does the search bring customers to you? Xero's VAT calculator reaches people who send invoices, Blue Arrow's salary calculator reaches people looking at job offers. The same tool on an unrelated website would bring visits and no customers. Volume only counts if the search is close to what you sell.
- How much does the domain weigh? ScS ranks for "sofa bed" with an internal search page Google can't even describe. If the example only works because of the brand, you can't replicate it.
- Is it measured or is it a screenshot? Macropix and Sunscape have Search Console behind them. The other seven are a snapshot of one day, on one device.
On method, for transparency: we read the UK results pages with DataForSEO (Google UK, English, smartphone) on 10 October 2026, depth 10-20 results; the two Italian examples were read on Google.it on 8 October 2026. Volumes are the Google Ads average for the last twelve months, links are DataForSEO's count on the same day. A results page read from another city or on another day may look different.
If you want to turn an example into a plan with the right priorities for your website, an SEO strategy starts right here. Or we can do it together with our SEO consultancy.
SEO examples: frequently asked questions
It depends on the type of website. For a manufacturer it's a category page that explains the product, like the Macropix page on "ledwall" (67,560 impressions and 1,495 clicks in twelve months). For a recipe website it's Recipe structured data, which brings stars, ratings and cooking times straight into Google's results.
The lever does, the numbers don't. Calculators win searches in the UK too ("vat calculator" gets 165,000 searches a month), but volumes, competitors and page format have to be read on the UK results page, not the US one: on Google UK the top VAT calculator was a software company's page with a link from a single domain.
No. From 2023 Google showed FAQ rich results only for well-known, authoritative government and health websites, and from 7 May 2026 it no longer shows them at all. FAQs on the page are still useful for answering real questions and for AI-generated summaries.
Yes, in its structure: which searches it covers, with which pages and which data. Not the text, and you need to check how much its result depends on the domain: ScS ranks for "sofa bed" with an internal search page, something a small website can't replicate.
It depends on how competitive the search is. Sunscape went from 1 to 38 keywords on page one in about six months of work, while Macropix took five years to climb from position 88 to 2 for "monitor pubblicitario" (advertising monitor).
In short (and the example that really matters)
An SEO example is action + search + result.
On categories you win with the answer, not the price.
Different demand per region, different page per region.
Structured data brings into the results what you already have.
A calculator brings customers if it sits next to what you sell.
Even Decathlon loses its own local search.
The source the AI cites may not be on page one.
One thing the industry rarely says: the most useful SEO example for your business isn't on any of these nine searches. It's the page of your direct competitor that today wins the search you want. Open it, read it the way we read these, and ask yourself which of the three parts you're missing: the action, the search or the measurement.
Speak soon!
