SEO stands for search engine optimization: the work of making a website easier for search engines to find, understand and show to people who are looking for what it offers. Google describes Search as working in three stages, crawling, indexing and serving, and states plainly that “not all pages make it through each stage.” SEO is the practice of improving your odds at each one. It is not a payment to Google, not a trick, and not a guarantee. Google’s wording is unambiguous: “Google doesn’t guarantee that it will crawl, index, or serve your page, even if your page follows the Google Search Essentials.”
Key Takeaways
– SEO stands for search engine optimization. You will also see it written as search engine optimisation, and both mean the same thing
– Google documents three stages: crawling, indexing and serving. Work that helps at one stage does nothing at another
– Google guarantees nothing. It says in writing that meeting every requirement still does not mean a page will be crawled, indexed or served
– You cannot pay for organic placement. Google’s own words: “Google doesn’t accept payment to crawl a site more frequently, or rank it higher. If anyone tells you otherwise, they’re wrong”
– Relevance is determined by “hundreds of factors,” in Google’s phrasing. There is no published list of them, and the familiar “200 ranking factors” is not a Google figure
– E-E-A-T is not itself a ranking factor. Google says so directly, while also saying its systems use signals that identify content showing it
– Google has no preferred word count. Asked directly whether writing to a target length helps, its documentation answers: “No, we don’t”
– Timelines are not fixed. Google says some changes take hours, others several months, and some produce no visible effect at all
– Google’s position on AI search is that optimizing for it “is optimizing for the search experience, and thus still SEO”
Prepared by the Search Engine Info editorial team. Statements about Google’s behavior are drawn from Google’s own documentation and Search Console help. Direct quotations appear in quotation marks and were checked word for word against the source. Text outside quotation marks is our summary of that documentation, and passages labeled as practitioner method are our judgment rather than anything Google has published. Last editorial review: 4 November. Next scheduled review: February.
Most definitions of SEO are written for people who already know what it means. This one is not. It starts with the acronym, explains what the work actually consists of, separates what Google documents from what the industry assumes, and says plainly where the limits are.
It is long, because the question has more parts than it looks like it has. If you want one paragraph, the first section is the whole answer. If you want the field, keep going.
What Is SEO?
SEO is the practice of making a website easier for search engines to find, understand and show to people searching for what it offers. The letters stand for search engine optimization.
That is the whole concept. Everything else is detail about how search engines find pages, how they decide what a page is about, and what you can reasonably do to help.
A worked version of the same idea: a bakery in Leeds publishes a page about gluten-free sourdough. SEO is the work that makes that page reachable by a search engine’s crawler, understandable as a page about gluten-free sourdough in Leeds, and credible enough to be worth showing to someone who searched for exactly that. Three separate problems, three separate kinds of work.
What SEO is not is a payment, a relationship, or a submission process. There is no queue to join and no fee that moves you up. Google puts that in writing twice over. In the Search Essentials: “It doesn’t cost any money to appear in Google Search results, no matter what anyone tries to tell you.” And in the guide to how Search works: “Google doesn’t accept payment to crawl a site more frequently, or rank it higher. If anyone tells you otherwise, they’re wrong.”
Nor is it a guarantee. The same Search Essentials page says that meeting every requirement and best practice “doesn’t mean that Google will crawl, index, or serve its content.” SEO improves odds. That is the honest description of the entire discipline, and anything promising more than that is selling something.
What SEO Stands For, and the Terms People Confuse It With
SEO stands for search engine optimization. If you have seen it written as “search engine optimisation” with an s, that is British spelling of the same term, not a different practice.
People look it up as both the SEO acronym and the SEO abbreviation, and either phrasing will find the right answer. For anyone who wants the pedantic version: since the letters are pronounced individually rather than run together as a word, it is strictly an initialism rather than an acronym, and an initialism is itself a kind of abbreviation. Nothing practical rests on the distinction, which is why the rest of this guide just says what SEO stands for.
There is no established synonym for SEO. You will occasionally see “organic search optimization” or just “search optimization”, and older material sometimes says “website optimization”, but none of these is a distinct discipline, and SEO is the standard term in every market.
The vocabulary around it has grown faster than the definitions, so here is the set of terms people most often ask about, separated.
| Term | What it means | How it relates to SEO |
|---|---|---|
| SEO | Search engine optimization | The parent term |
| Search engine optimisation | The same thing, British spelling | Identical. No difference in practice |
| Organic search | Unpaid search results | What SEO works on. The opposite of paid placement |
| SEM | Search engine marketing | Usually the umbrella covering both SEO and paid search ads |
| PPC | Pay-per-click advertising | Paid placement. Separate budget, separate system, immediate |
| SERP | Search engine results page | The page of results your work is trying to appear on |
| Indexing | A search engine storing a page in its database | A prerequisite. An unindexed page cannot rank at all |
| Crawling | A search engine fetching a page | Happens before indexing. Also not guaranteed |
| GEO / AEO | Generative engine or answer engine optimization | Newer labels for visibility in AI answers. Google treats this as part of SEO |
SEO is not the same as SEM or PPC
This confusion costs people money. SEO earns unpaid placement over time. PPC buys placement immediately and stops the moment you stop paying. SEM is normally used as the umbrella term for both, though some people use it to mean paid search alone, which is why it pays to ask what someone means rather than assume.
The practical difference: a PPC campaign can start sending traffic this afternoon and sends none the day the budget runs out. SEO work takes effect slowly and keeps working after you stop, until the page falls out of date or someone does the job better.
One thing that is settled, and worth saying because it is still sold the other way: buying ads does not help organic rankings. Google’s guidance on hiring an SEO states it flatly. “Advertising with Google won’t have any effect on your site’s presence in our search results.” The two systems do not talk to each other in the direction anyone hopes they do.
What “SEO positioning” means
You will see the phrase “SEO positioning”, particularly in material translated from Spanish, where posicionamiento SEO is the ordinary name for the discipline. It means the same thing as SEO. It is not a separate technique, and nobody needs to buy a product called positioning.
A short glossary of the terms you will meet next
The rest of this guide uses these words without stopping to define them. Here they are in one place.
| Term | Definition |
|---|---|
| Organic traffic | Visits that arrive from unpaid search results |
| Keyword | A word or phrase someone types into a search engine. Also called a query |
| Search intent | What the person wants when they type a query. Often inferable from the wording, but not always: some queries carry several plausible intents at once, which is why their results look mixed |
| Title tag | The page title in the HTML. Google uses it, among other text, to generate the headline shown in results |
| Meta description | A short summary in the HTML. Not a ranking factor, but it can become the snippet under your result |
| Alt text | Text describing an image, read by screen readers and used by Google during indexing |
| Backlink | A link to your site from another site |
| Anchor text | The visible, clickable words of a link. They describe the destination to both readers and crawlers |
| Internal link | A link from one page of your site to another |
| Canonical | When several pages are near-duplicates, the one Google picks to show. Google clusters similar pages and selects the most representative one |
| robots.txt | A file telling crawlers which parts of a site not to request. It controls crawling, not indexing, and a blocked URL can still appear in results if Google finds it referenced elsewhere. To keep a page out of results, use noindex instead |
| noindex | An instruction telling search engines not to include a page in results. Different from robots.txt, and often confused with it |
| XML sitemap | A file listing your URLs, which helps Google discover pages it might not reach by following links |
| Structured data | Machine-readable markup describing what a page contains. It can make a page eligible for richer result formats. Eligible, not entitled |
| Core Web Vitals | Google’s measurements of loading, interactivity and visual stability. Confirmed as used by its ranking systems |
| Featured snippet | An answer pulled from a page and shown at the top of results |
| AI Overview | A generated summary shown above results, drawing on multiple sources |
What Is SEO in Marketing?
In a marketing context, SEO is the channel that earns unpaid traffic from search engines. It sits alongside paid search, email, social and direct traffic as one route by which people arrive at a business.
What distinguishes it is where the demand comes from. Channels differ in how much of it they create and how much they capture, and search sits further toward capture than most: the person has already decided they want something and has typed what it is. Email and social can do that too, when someone has subscribed or is actively looking, but they more often reach people who were doing something else. That difference is why search traffic tends to convert comparatively well, and why the channel is worth the patience it demands.
What marketing teams typically own within it:
- Deciding which searches matter. Not every query is worth a page, and chasing volume alone produces traffic that never buys anything
- Producing content that answers them. The actual writing, which is the slow part
- Making sure the site can be crawled and indexed. Usually a shared job with whoever owns the site build
- Earning credibility off the site. Coverage, mentions and links, which cannot be bought safely
- Measuring it honestly. Mostly in Search Console and analytics, and mostly over quarters rather than weeks
One caution that belongs in any marketing conversation about SEO: its feedback cycle is usually much longer than a paid channel’s. A paid campaign can report in days. How long SEO takes varies a great deal, as the section below sets out, but expecting it to report on a paid campaign’s schedule is a reliable way to abandon it before it has had a chance to work.
The second caution is about where SEO sits in the organization. It touches the site build, the content calendar, the PR function and sometimes the product. A team that owns none of those things and is still asked to deliver search growth has been given a target without the levers. That is an organizational problem rather than a search one, and no amount of keyword research fixes it.
How SEO Actually Works
Google documents the mechanism, which means you do not have to guess at the basic shape of it. Google Search works in three stages, and Google states plainly that “not all pages make it through each stage.”
| Stage | What happens | What SEO does about it |
|---|---|---|
| Crawling | Google downloads text, images and video from pages it has found, using automated programs called crawlers | Make pages reachable: working links, no accidental blocks in robots.txt, a sitemap, pages that return a valid response |
| Indexing | Google analyzes what is on the page and stores the information in its index | Make the subject unambiguous: clear titles and headings, descriptive text, no duplicate versions competing |
| Serving | Google returns results it judges most relevant to a query | Make the page genuinely the best answer, and make its credibility visible |
Those three stages are the whole SEO process in outline. Everything that gets sold as a tactic is an attempt to help at one of them.
The useful consequence of separating these is diagnostic. If a page is not getting traffic, the first question is which stage it failed at, because the fixes are unrelated. A page blocked from crawling will not be helped by better writing. A thin page that is crawled and indexed perfectly well will not be helped by a sitemap.
Stage one: crawling, or how Google finds a page at all
Google’s starting point is that nothing tells it your page exists. In its own description, “There isn’t a central registry of all web pages, so Google must constantly look for new and updated pages.” It finds them three ways: it revisits pages it already knows, it extracts links from known pages to new ones, and it reads sitemaps that site owners submit.
Discovery is not the same as a visit. Google says that once it discovers a URL it “may visit (or ‘crawl’) the page,” and that Googlebot “uses an algorithmic process to determine which sites to crawl, how often, and how many pages to fetch from each site.” The rate is also responsive to your server: Google’s crawlers are built to avoid overloading a site, and it names HTTP 500 errors as a signal meaning “slow down.”
Google lists three common reasons its crawler cannot reach a site:
- Problems with the server handling the site
- Network issues
- robots.txt rules preventing Googlebot’s access to the page
All three are invisible from the browser. Your page loads perfectly for you while returning errors to Googlebot, or while being excluded by a line in a file you have never opened. This is why crawl problems are usually found late: nothing about the site looks wrong.
The practical checklist at this stage is short. Make sure important pages are linked to from somewhere, since internal links are the pathway you control, and a page with none is left depending on a sitemap or on being referenced from outside your site. Make sure links are real HTML links rather than click handlers. Keep redirects short and avoid loops. Submit a sitemap. And read your own robots.txt at least once, because an inherited block from a staging environment is a genuinely common and genuinely catastrophic mistake.
Where rendering fits, and why you will see SEO described as four stages
Many SEO guides describe four stages rather than three, splitting rendering out as its own step between crawling and indexing. That is a reasonable teaching device, but it is worth knowing what Google actually documents, because the difference changes where you look when something breaks.
Google places rendering inside crawling: “During the crawl, Google renders the page and runs any JavaScript it finds using a recent version of Chrome, similar to how your browser renders pages you visit.” It gives the reason directly: “Rendering is important because websites often rely on JavaScript to bring content to the page, and without rendering Google might not see that content.”
Two things follow from that sentence, and they matter more than the stage count.
First, JavaScript content can be seen. The old advice that Google cannot read JavaScript is out of date, and repeating it leads people to rebuild sites that did not need rebuilding.
Second, it can be seen only if the files that produce it are reachable. If your CSS or JavaScript is blocked from crawling, the render runs without them, and Google sees a page your visitors never see. Blocking resources to save crawl budget is one of the more reliable ways to make a working site invisible.
[Practitioner method, not documented behavior] The usual working rule is that content which requires a user action to appear, such as a click or a scroll handler that fetches more, should not be content you need indexed. Content that renders on load generally gets seen; content behind an interaction is a gamble.
Stage two: indexing, or how Google decides what the page is
Indexing is where Google works out the subject of the page. Its description covers “processing and analyzing the textual content and key content tags and attributes, such as <title> elements and alt attributes, images, videos, and more.”
This stage also handles duplication, and the mechanism is worth understanding because it surprises people. Google groups together pages with similar content, a process it calls clustering, and then selects one as the canonical: “the page that may be shown in search results.” The others become alternate versions. Your preference, expressed through a canonical tag, is a signal in that decision rather than an instruction that settles it.
Then the blunt sentence: “Indexing isn’t guaranteed; not every page that Google processes will be indexed.” Google names three common reasons:
- The quality of the content on page is low
- Robots meta rules disallow indexing
- The design of the website might make indexing difficult
Which brings up the status that causes more confusion than any other. “Crawled, currently not indexed” in Search Console means Google fetched the URL and has not added it to the index at this time. It does not, by itself, identify a cause. Google’s own description of the status is deliberately undramatic: “Google doesn’t index all pages across the web, so this is a common message for many sites and doesn’t necessarily indicate an issue that needs attention.”
Google’s advice for an important page with that status is to diagnose rather than assume: “use the URL Inspection tool and other tools to diagnose if there are any issues on your site that are preventing the pages from appearing, such as if you’ve blocked them from our crawlers.” So the sensible checks are indexing eligibility, canonicalization, whether the content is substantially unique, internal links, and crawl access.
If all of those come back healthy, there may be nothing to fix. Google does not index every page, and resubmitting a URL repeatedly does not make inclusion more likely.
Stage three: serving, or how results get ordered
When someone searches, Google says, “our machines search the index for matching pages and return the results we believe are the highest quality and most relevant to the user’s query.” The ordering is not fixed and it is not universal: “Relevancy is determined by hundreds of factors, which could include information such as the user’s location, language, and device.”
Google’s own example is that a search for bicycle repair shops shows different results in Paris than in Hong Kong. Which means there is no such thing as “the” ranking for a keyword. What a rank tracker reports is one result from one location on one device, and treating it as a fact about your page rather than a sample is how people end up chasing movements that are not real.
The result layout varies too. Google notes that the search features on the page change with the query: local results for one search, image results for another. Your competition for attention is not only the other ten links.
And indexed does not mean visible. Google lists three reasons a page can be indexed and still not appear:
- The content on the page is irrelevant to users’ queries
- The quality of the content is low
- Robots meta rules prevent serving
Turning the three stages into a diagnosis
The reason the stages are worth learning is that they tell you where to look. Each symptom has a different first question, and asking the wrong one wastes weeks.
| What you observe | First question to ask | Where to check |
|---|---|---|
| An important URL has not been crawled | Can Googlebot discover and access it? | URL Inspection, internal links, robots.txt, server logs |
| Crawled but not indexed | Is there an eligibility, canonicalization, duplication or content issue? And is there anything to fix at all? | URL Inspection and the Pages report |
| Indexed but almost no impressions | Does the page match queries anyone actually searches? | Search Console query data for the page |
| Impressions but few clicks | Is this position, query intent, or how the result is being presented? | URL-level query report, and the result as it appears |
| Clicks but no conversions | Does the page serve the visitor’s intent and the business goal at once? | Analytics and a read of the landing page |
Working down that table in order is more productive than applying tactics in parallel, because each row rules out the rows above it.
How Google decides what a page is about
Modern Google Search combines textual relevance signals with systems that interpret meaning and context, rather than using one instead of the other. Google’s ranking systems documentation describes neural matching as a system that helps it understand representations of concepts in queries and pages and match them to one another, and BERT as a system for understanding how combinations of words express different meanings and intent. A page can therefore be relevant to searches that do not reproduce its wording exactly, while the words on the page still matter.
That is why a page can rank for a phrase it never contains word for word, and why repeating a phrase fifteen times stopped working a long time ago. Google defines keyword stuffing as “filling a web page with keywords or numbers in an attempt to manipulate rankings,” names “repeating the same words or phrases so often that it sounds unnatural” as an example, and lists it among the spam policies rather than among the techniques.
Words still matter, though, and Google says so in its own list of key best practices: “Use words that people would use to look for your content, and place those words in prominent locations on the page, such as the title and main heading of a page, and other descriptive locations such as alt text and link text.”
So the instruction is not to avoid your keywords. It is to use the language your readers use, put it where it belongs, and then stop. For how that planning works in detail, see our guide to semantic SEO.
SEO Ranking Factors, and the “200 Factors” Claim
You will meet the claim that Google uses 200 ranking factors within about an hour of starting to read about SEO. It is worth knowing where that number stands.
Google’s current documentation says “hundreds of factors,” not 200. It has never published a list, and the factors it does name in that sentence are not the ones people expect: the user’s location, language and device. Those are properties of the searcher, not of your page.
So the honest position is this. Google confirms that many signals exist, confirms a handful of them specifically, and has never enumerated the rest. Any numbered list you read is a reconstruction built from patent filings, public statements, correlation studies and informed guesswork. Some of it is good. None of it is authoritative, and treating it as a checklist produces work that is busy rather than useful.
What Google does confirm
A short list is more useful than a long one, so here is what sits on documented ground.
| Signal | What Google says | Evidence tier |
|---|---|---|
| Relevance to the query | Results returned are those Google believes are “the highest quality and most relevant” | Documented |
| Location, language, device | Named explicitly as things relevancy “could include” | Documented |
| Core Web Vitals | “Core Web Vitals are used by our ranking systems” | Documented |
| Other page experience aspects | They “don’t directly help your website rank higher in search results” | Documented, and commonly misreported |
| A single page experience signal | “There is no single signal” | Documented as not existing |
| E-E-A-T | “While E-E-A-T itself isn’t a specific ranking factor, using a mix of factors that can identify content with good E-E-A-T is useful” | Documented, widely misstated |
| Keywords meta tag | Google “does not use the keywords meta tag in web ranking” | Documented as not used, announced 2009 |
| Third-party authority scores | Google “doesn’t evaluate or endorse third-party SEO tools, and these tools don’t have access to Google’s internal ranking data” | Documented as not used |
Two rows there deserve the attention. Google is explicit that aspects of page experience beyond Core Web Vitals do not directly improve rankings, while adding that they make a site more satisfying to use and are therefore still worth improving. And on Core Web Vitals themselves it adds a warning that very little vendor material repeats: good scores do not guarantee top rankings, and “trying to get a perfect score just for SEO reasons may not be the best use of your time.”
That last sentence is Google telling you to stop optimizing a number. It is the most ignored piece of guidance in the field.
What E-E-A-T is, and what it is not
E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness. It comes from Google’s quality rater guidelines, which are the instructions given to the people who assess whether search results look good.
Google’s framing is careful, and the care is the point. Its systems “identify a mix of factors that can help determine which content demonstrates aspects” of E-E-A-T. But “E-E-A-T itself isn’t a specific ranking factor.” There is no E-E-A-T score. Quality raters do assess pages and results, and they are trained to judge E-E-A-T, but their assessments are feedback on whether Google’s systems are working rather than scores applied to your site, so no individual rating moves your rankings.
Within the four, Google says trust is most important, and that the others contribute to trust rather than standing alone: “content doesn’t necessarily have to demonstrate all of them.” It also says the weighting is not even across topics. For subjects that “could significantly impact the health, financial stability, or safety of people, or the welfare or well-being of society”, which it calls Your Money or Your Life topics, its systems “give even more weight to content that aligns with strong E-E-A-T.”
The practical reading: who wrote this, why they would know, and whether the page is honest about its sources are the things worth fixing. A byline, an author page, visible sourcing and a date are cheap. Our E-E-A-T guide covers how this is assessed in detail.
The Parts of SEO: On-Page, Off-Page, Technical and Content
The field gets divided into four areas. The division is a convention rather than a Google classification, but it is useful because the four need different skills and often different people.
| Area | What it covers | Typical first job |
|---|---|---|
| Technical SEO | Whether search engines can reach, render and index the site at all | Confirm important pages are actually indexed, and fix what is blocking the ones that are not |
| On-page SEO | What is on an individual page: titles, headings, structure, internal links | Make the title and first paragraph say what the page answers |
| Content | Whether the page is worth showing to anyone | Answer the question the searcher actually has, completely |
| Off-page SEO | Signals from elsewhere: links, mentions, reputation | Earn coverage by being worth covering. Links bought for ranking purposes breach Google’s policies unless qualified with nofollow or sponsored |
There is a rough order to these. Technical problems make everything else pointless, because an unindexed page cannot rank no matter how good it is. Content quality is the ceiling, because no amount of technical polish makes a thin page worth showing. Off-page work is the slowest and the least directly controllable.
Technical SEO: can the machine get in
Technical SEO is the work of removing obstacles between a search engine and your content. It is unglamorous and it is first, because every other kind of effort depends on it.
The things that actually matter, roughly in order of how much damage they do when wrong:
- Indexability. Does the page return a valid response, and is anything telling search engines to skip it? A stray noindex is the single most expensive one-line mistake in the field
- Crawl access. Can Googlebot fetch the page and the CSS and JavaScript it needs to render it
- Site structure. Can every important page be reached by following links from the home page in a few steps
- Duplication control. Are there four URLs serving one page, and have you told Google which you prefer
- Core Web Vitals. Confirmed as used by Google’s ranking systems, with the caveat above about chasing perfect scores
- HTTPS. Named in Google’s page experience self-assessment
- Mobile. Google’s self-assessment asks whether content displays well on mobile devices
- Structured data. Makes a page eligible for richer results. Eligibility is not a promise of display
Google’s own framing of the technical bar is reassuring, and people rarely quote it: “There are actually very few technical things you need to do to a web page; most sites pass the technical requirements without even realizing it.” Technical SEO on a normal site is mostly checking that nothing is broken, not building something elaborate.
On-page SEO: is the page legible
On-page SEO is everything on the page itself that helps a reader and a machine work out what it covers.
- The title. Says what the page answers, in the words a reader would use. Google’s best-practice list specifically names the title and main heading as prominent locations for those words
- The first paragraph. Answers the question immediately rather than introducing the topic of answering the question
- Headings. A structure a reader can skim, with each section doing one job
- Internal links. With anchor text that describes the destination, since Google names link text as a descriptive location
- Alt text. Named by Google as a content attribute it analyzes during indexing, and the thing that makes your images usable by screen readers
- The meta description. Not a ranking factor. It can become the snippet, so it is worth writing, and it is not worth agonizing over
- URL. Readable and stable. Changing URLs has a cost and rarely a benefit
Notice what is absent from that list: a target keyword density, a word count, and the optimization score your plugin shows you. None of those is a documented Google requirement. The plugin score in particular is a vendor’s model of what it believes Google wants, and Google’s position on third-party tools is that it does not evaluate or endorse them and that they have no access to its ranking data.
Wording itself is a different matter, and it does count. Google asks you to use the words people would use to look for your content and to place them prominently. What is not required is forcing an exact phrase in repeatedly or optimizing toward a ratio. Use the phrase where it belongs, in language a reader would recognize, and then leave it alone.
Content: is the page worth showing
Content is the part that no technical work substitutes for, and Google has published its own assessment questions for it. The useful ones to ask before publishing:
- Does the content provide original information, reporting, research, or analysis?
- Does it provide a substantial, complete, or comprehensive description of the topic?
- Does it provide insightful analysis or interesting information that is beyond the obvious?
- If it draws on other sources, does it avoid simply copying or rewriting those sources, and instead provide substantial additional value and originality?
- Is this the sort of page you’d want to bookmark, share with a friend, or recommend?
- Does the content have any easily-verified factual errors?
Those are Google’s words, and the fourth one is the test most content fails. A page assembled by reading the top ten results and rewriting them has nothing in it that was not already available, which is a reasonable description of most of what gets published.
Google also publishes the inverse: warning signs that you are making content for search engines rather than people. Among them, producing lots of content on many topics hoping some performs, using extensive automation, mainly summarizing what others say without adding value, and entering a niche with no real expertise because you thought you would get traffic.
And one that answers a question this guide cannot avoid, since it is itself long: “Are you writing to a particular word count because you’ve heard or read that Google has a preferred word count? (No, we don’t.)” Length is a consequence of how much a subject genuinely requires. It is not a target, and a page padded to a number is a page that reads like it was padded to a number.
Off-page SEO: what the rest of the web says
Off-page SEO covers the signals that come from outside your site: links, mentions, coverage, reputation. It is the slowest part and the one you control least, which is precisely why it carries weight.
It is also the part with the clearest rules, because Google spells out what it will not accept. Link spam is defined as “creating links to or from a site primarily for the purpose of manipulating search rankings,” and the examples are specific:
- Exchanging money for links, or posts that contain links
- Exchanging goods or services for links
- Sending someone a product in exchange for them writing about it and including a link
- Excessive link exchanges, or partner pages existing only for cross-linking
- Using automated programs or services to create links
- Advertorials or native advertising where payment is received for articles containing links that pass ranking credit
- Low-quality directory or bookmark site links
- Widely distributed links in the footers or templates of various sites
There is a legitimate path through this, and Google states it: “Google does understand that buying and selling links is a normal part of the economy of the web for advertising and sponsorship purposes. It’s not a violation of our policies to have such links as long as they are qualified with a rel=’nofollow’ or rel=’sponsored’ attribute value.”
So sponsorship is fine. Sponsorship that passes ranking credit is not. The distinction is a single attribute, and the number of agencies selling the second thing while describing the first is the reason this section exists.
What is left, once you remove everything on that list, is genuinely narrow: publish something worth citing, tell people it exists, and be the sort of organization others reference. Google’s own key best practices put it almost that plainly, listing “Tell people about your site. Be active in communities where you can tell like-minded people about your services and products.”
Types of SEO: The Specialisms
Past the four areas, SEO divides again by the kind of site being worked on. These are not different disciplines so much as the same discipline with different dominant problems, and the distinction matters mostly when you are hiring, or deciding what to learn next.
| Specialism | What makes it different | The problem that dominates |
|---|---|---|
| Local SEO | Visibility for searches tied to a place, including map results | Business listings, reviews and consistent information, which run on partly separate signals from web results |
| Ecommerce SEO | Category and product pages at volume, with variants and filters | Near-duplicate pages and filter URLs multiplying into thousands of crawlable combinations |
| Enterprise SEO | Very large sites, many stakeholders, slow release cycles | Getting changes shipped. The hard part is rarely knowing what to fix |
| International SEO | Multiple languages or countries | Telling search engines which version serves which audience, and keeping translated pages from competing with each other |
| News SEO | Time-sensitive publishing | Speed of indexing, since a story indexed tomorrow is worthless |
| GEO / AI search | Visibility inside generated answers rather than link lists | Being retrievable and quotable at the passage level. Google treats this as part of SEO, not a separate practice |
Two notes on the last row, because it is the one changing. Google’s hiring guidance now lists “Optimizing for generative AI” among the services SEOs provide, which is an acknowledgment that the work exists. It also suggests a test for anyone selling it: ask whether their advice “is aligned with Google Search’s official guidance on optimizing for generative AI features.” That is a usefully concrete question, because a large amount of what is currently sold as GEO contradicts that guidance directly.
If you are choosing where to specialize, the honest answer is that it depends on your market, the role, your experience and the kind of employer you want, and we have no data that would support ranking these by demand or pay. What can be said about the work itself: local and ecommerce involve more repeatable operational work, technical SEO demands stronger diagnostic and development skills, and enterprise adds coordination, governance and implementation problems that have little to do with search.
What SEO Is Not
A good deal of what gets sold as SEO is not SEO, and the fastest way to understand the discipline is to see what sits outside it.
- It is not paying Google. Organic results cannot be bought. Google’s wording: “Google doesn’t accept payment to crawl a site more frequently, or rank it higher. If anyone tells you otherwise, they’re wrong”
- It is not advertising. Running Google Ads has no effect on organic placement, which Google states directly
- It is not keyword density. There is no published target ratio, and keyword stuffing, which Google defines as “filling a web page with keywords or numbers in an attempt to manipulate rankings”, sits in the spam policies rather than among the techniques
- It is not the keywords meta tag. A separate matter from density, and settled longer ago: Google announced in 2009 that it “does not use the keywords meta tag in web ranking”
- It is not a word count. Asked whether Google has a preferred length, its documentation answers: “No, we don’t”
- It is not refreshing dates. Google names “changing the date of pages to make them seem fresh when the content has not substantially changed” as a warning sign, and says adding or removing content to seem fresh will not help: “No, it won’t”
- It is not a one-time project. Results pages change because competitors publish, demand shifts and Google updates its systems
- It is not guaranteed. Google’s guidance is blunt: “No one can guarantee a #1 ranking on Google.” It adds that you should be wary of anyone who claims a special relationship with Google or a priority submission
- It is not separate from writing well. Google’s own framing is that “SEO can be a helpful activity when it is applied to people-first content, rather than search engine-first content”
- It is not a vendor score. Third-party authority and optimization scores are vendor models. Google neither evaluates nor endorses them, and they have no access to its ranking data
Why SEO Matters
The case for SEO is usually made with statistics, most of which come from vendors with something to sell and none of which we can verify from source. So here is the case made structurally instead, which is more durable anyway.
Search is unusually good at capturing demand that already exists. Someone typing “emergency plumber Leeds” is not being persuaded to want a plumber. They stated the need themselves, in their own words. Other channels capture existing demand too, and the useful difference is one of balance: search sits further toward capture and further from creation than most of them. That makes it valuable for reaching people with an identifiable need, though how well the traffic converts depends on the query, the audience and the page they arrive on.
The asset does not need recurring media spend. A paid campaign delivers traffic for exactly as long as the budget lasts. An established page can keep attracting organic traffic after the work stops.
That traffic is not guaranteed, though, and the decline is not always gentle. Rankings and clicks can fall gradually as a topic ages or competitors improve, or abruptly after a technical fault, a ranking systems update, a change in how results are presented, or a drop in demand for the query itself. The inertia also runs the other way: problems persist quietly, and a page that fell out of the index last March is still out of it today unless somebody looked.
The work improves the site regardless. This is the part that survives any algorithm change. Pages that load quickly, say clearly what they are about, answer the question they promised, and can be navigated by a human are better pages. If search disappeared tomorrow, most of the work would still have been worth doing. That is not true of much marketing spend.
It is cheap to start learning and expensive to catch up on. Search Console is free, Google’s documentation is free, and you can diagnose a site without spending anything. Doing the work is a separate question: content production, development time and any help you bring in all cost money whether or not the tools did. What is genuinely expensive is the years of compounding you never started, which is why “we will do SEO later” costs more than it sounds.
The counterweight, stated honestly: search is a channel with a landlord. Google can change what a results page looks like, and has. AI Overviews answer some queries in place, which means impressions without clicks for the pages that informed them. A business whose entire acquisition depends on one channel it does not control has a concentration risk, and SEO being worth doing does not make that risk go away.
Does SEO Actually Work?
Yes, with two honest qualifications attached.
The first is that it works probabilistically. Google’s wording is that sites following its guidelines “are more likely to show up,” which is a statement about improving odds, not about producing a result. Any individual page can do everything right and still not rank, because something better exists or because the query simply does not have room.
The second is that the effect is hard to attribute cleanly. Rankings move for reasons outside your control: algorithm updates, competitors improving, seasonal demand, changes in what the query means to people. A page that improves after an edit may have improved because of the edit. Without a recorded baseline, you cannot tell.
What makes it worth doing anyway is the economics described above. Search traffic is demand that already exists, arriving from people who described what they wanted. It does not stop when a budget stops. And the work that produces it, clearer pages that answer real questions, tends to be worth doing for its own sake.
How SEO Work Actually Gets Done
Knowing what SEO is does not tell you what an SEO does on a Tuesday. The work divides into four activities that repeat indefinitely: finding out, deciding, changing, and watching.
Research: finding out what is true
Research comes first because almost every expensive SEO mistake is a confident action taken without it. It covers four separate questions.
- What are people actually searching for? Keyword research, which produces the queries, roughly how often they occur and how contested they are. The volumes are vendor estimates rather than Google figures, and should be read as order of magnitude
- What do those searches mean? What currently ranks is evidence about what the search engine believes the query wants. Treat it as evidence to investigate rather than proof: if every result is a comparison table, that is a strong signal a personal essay is the wrong format, but a query can also support several intents at once, and results that look mixed usually mean exactly that
- What does the site already have? Most sites have pages nobody remembers publishing, several pages competing for one query, and a handful of pages earning impressions for things nobody intended. Search Console shows all three
- What are competitors doing? Less useful than it sounds, and mostly valuable for finding questions nobody has answered rather than copying what everyone has
The output of research is not a spreadsheet. It is a short list of things you now know that you did not know last week.
Planning: deciding what not to do
An SEO strategy is a decision about which searches you are going to try to serve, and which you are not. That is the whole of it. Everything else, the content calendar, the technical fixes, the outreach, follows from that decision.
The reason strategy matters more than tactics here is that effort is finite and search demand is not. A site with forty pages cannot compete for everything, so the useful question is never “how do we rank” but “for what, and why that rather than something else.” A strategy that cannot name what it has excluded is a wish list.
A plan that is actually a plan has four things in it: the queries you are going after, the pages that will serve them, the order you will build them in, and what you will consider evidence that it worked. Everything else is scaffolding.
For a small site that usually resolves to a short list: the handful of searches that describe what you actually sell or know, the questions those people ask before and after, and nothing else until those are done properly.
Two planning decisions cause most of the later pain. The first is whether a topic needs one page or several, and the default should be one, because two pages competing for a single intent is a self-inflicted wound. The second is sequencing: fixing indexing before writing, because publishing into a site Google cannot crawl is pouring water into a bucket with no bottom.
Implementation: changing things
Implementation is where SEO stops being analysis. It comes down to three kinds of change.
- Creating pages that do not exist. The slowest and usually the highest value
- Improving pages that do. Faster, and frequently better value, because a page already earning impressions for a query it half-answers is the cheapest win available
- Removing or merging pages. Two thin pages on one subject are usually worth less than one good page. Consolidating them, with a redirect, is real work with real upside
One caution on the third. Deleting content because a page gets no traffic is a reasonable-sounding instinct that Google explicitly warns against doing for ranking reasons: removing a lot of older content “primarily because you believe it will help your search rankings overall by somehow making your site seem ‘fresh'” will not help. Remove pages because they are wrong or redundant, not because the site needs pruning.
Monitoring: noticing when something breaks
Competitors are not the only thing that costs you traffic, and often not the most immediate. Deployments, plugin updates, migrations and well-meaning template changes are a significant operational risk, and the reason they are worth watching for specifically is that they go unnoticed: nothing looks wrong from a browser.
The short list worth watching:
- Pages dropping out of the index, visible in Search Console’s Pages report
- Sudden impression or click drops on pages that matter
- Server errors and slow responses, which Google treats as a reason to crawl less
- Accidental noindex or robots.txt changes, especially after a launch
- Internal links breaking when URLs change
[Practitioner method, not documented behavior] A monthly check of the Pages report and the top twenty pages by clicks catches most of this. The failure mode is not insufficient tooling. It is nobody looking for six months.
How to Measure SEO
SEO measurement goes wrong in a specific way: people measure the thing that moves first rather than the thing they care about.
What to measure, in order of usefulness
| Metric | What it tells you | Why it misleads |
|---|---|---|
| Conversions from organic | Whether the channel produces business | Slow, and attribution across visits is imperfect |
| Clicks | People who actually arrived | Seasonal, and brand searches can mask a decline in everything else |
| Impressions | Whether you are being shown at all | Can rise while clicks fall, which usually means you appeared lower or in a feature that answered the question for you |
| Indexed page count | Whether your pages are eligible to appear | A high number means nothing if the wrong pages are the ones indexed |
| Average position | Roughly where you sit | Averaged across locations and devices, so it blurs exactly the variation Google says exists |
| Rank tracker positions | Position for one location, one device, one moment | Reported as fact. Google says results vary by location, language and device, so this is a sample |
| Vendor authority scores | A vendor’s model of your link profile | Not a Google metric. Google does not endorse these tools and they cannot see its ranking data |
The ordering is deliberate. The metrics at the top move slowly and matter. The ones at the bottom move constantly and mostly generate meetings.
Working out whether SEO paid for itself
ROI on SEO is genuinely harder to calculate than on paid channels, and pretending otherwise is how the channel loses budget when someone eventually checks.
One figure that is easy to calculate is the advertising equivalent: multiply the relevant organic clicks by an estimated paid cost per click for those queries. Be clear about what that produces. It is a directional estimate of what replacing the traffic through advertising might cost. It is not SEO revenue, profit, return on investment, or a guaranteed minimum value, because ad prices, organic visitor behavior and conversion value are different quantities, and treating them as one inflates the result.
To assess ROI properly you have to compare the attributed value of conversions from organic search against what the work cost to produce and maintain, and then state the limits of that comparison: attribution across multiple visits is imperfect, branded searches are credited to organic while being driven by other marketing, assisted conversions are invisible in a last-click view, and paid and organic visitors do not behave identically.
What does not work is attributing revenue to SEO as though no other channel existed. Someone who searched your brand name after seeing an ad arrived through organic search and was delivered by the ad. Any honest SEO report separates branded from non-branded queries for exactly this reason.
How Long SEO Takes, and How Hard It Is to Learn
How long does SEO take?
Google answers this directly and the answer is less satisfying than most published estimates: “Every change you make will take some time to be reflected on Google’s end. Some changes might take effect in a few hours, others could take several months.” Google goes on to recommend waiting “a few weeks to assess whether your work had beneficial effects,” and adds the part that usually gets left out: “not all changes you make to your website will result in noticeable impact in search results.”
So the honest range is hours to months, with no guarantee of a visible effect. Anyone quoting a fixed timeline is quoting their own experience, which may be useful but is not a published figure.
The variation has reasons behind it. A technical fix that unblocks crawling can show up within days, because the obstacle was mechanical. A new page competing in an established topic takes longer, because it has to be discovered, indexed, and then earn its way up against pages that have been accumulating signals for years. Those two things get averaged into a single “how long does SEO take” answer, which is why the answer is always wrong.
How hard is SEO to learn?
The fundamentals are genuinely straightforward and can be understood in an afternoon: make pages reachable, make them clear, make them worth reading. Most of what a small site needs is on this page.
What takes time is judgment. Knowing which of twenty possible changes matters for your particular site, reading a results page accurately, deciding when a page should be split and when two should be merged, telling a real ranking drop from normal variance. That is pattern recognition, and pattern recognition takes repetition.
The unusual difficulty in this field is the noise. A large volume of SEO advice describes tactics that Google has publicly said do nothing. Learning to check a claim against Google’s own documentation before acting on it is, in practice, the single most valuable skill.
Google itself frames the decision in terms of site size rather than aptitude: “If you run a small local business, you can probably do much of the work yourself.” That is a fair reading. The difficulty of SEO scales with the complexity of the site, not with the difficulty of the concepts.
SEO and AI Search
AI Overviews, AI Mode, ChatGPT Search and Perplexity have changed what a results page looks like, and raised a reasonable question about whether SEO still applies.
Google’s answer, published in May, is blunt: optimizing for generative AI search “is optimizing for the search experience, and thus still SEO.” The guide also names things you can ignore for Google Search, including llms.txt files, which it says Google ignores, and content chunking, and purchased mentions.
Two things genuinely do change.
Retrieval can operate below the level of the whole page. An AI system can draw on the relevant part of a page rather than treating the page as one indivisible answer. That makes clear headings, self-contained sections and enough context inside each section useful, for readers first and for automated retrieval possibly as well.
Two qualifications, because this is the claim most often overstated, including in earlier versions of this guide. AI systems synthesize and paraphrase as well as quote, so there is no guarantee your wording is what appears. And passage-level structure is not a documented ranking requirement on any platform. Google’s position is the opposite of the advice usually given here: it says you can ignore “chunking” content. Write sections that stand on their own because that helps a reader, not because a system is believed to require it.
Being cited is not the same as being clicked. Your page can inform an answer the reader never clicks through from, and the reporting splits across two places. Search Console’s dedicated generative AI performance report covers AI Overviews and AI Mode but shows impressions only, defined as “how many times links to your site were shown to a user in a generative AI feature on Google Search.” It gives no clicks, no click-through rate and no average position.
Clicks from those features are counted, just not there. Google states: “Just like the rest of the search results page, sites appearing in AI features (such as AI Overviews and AI Mode) are included in the overall search traffic in Search Console. In particular, they’re reported on in the Performance report, within the ‘Web’ search type.”
The consequence is specific, and worth getting right before anyone builds a report on it: AI feature clicks are in your totals, mixed in with ordinary web search, so you can see AI impressions and you can see combined clicks, but you cannot calculate an AI-specific click-through rate from what Search Console currently gives you.
One point of continuity worth noting, because it affects how the tactics are policed: Google’s spam policies now explicitly include “attempting to manipulate generative AI responses in Google Search” within the definition of spam. The generated answer is not a separate venue with separate rules.
What does not change is that crawl access still matters. It is worth being precise about why, because the usual shorthand, that a page which cannot be crawled cannot be retrieved by anything, is too strong. Google Search, ChatGPT Search, Perplexity and the rest do not share one discovery system or one crawler. Blocking a particular crawler limits that provider’s access to your content, and what follows from it depends on the provider and the product.
So do not assume that being indexed by Google guarantees retrieval by every AI service, or that one crawler being blocked proves your content is unreachable through any other route. Check each system you care about on its own terms. We cover this in more depth in our guides to generative engine optimization, optimizing for AI Overviews and how ChatGPT Search works.
SEO as a Job and as a Service
SEO is both a thing you do and a thing you can be. The job title covers a wide range of actual work, which is why two people with the same title often cannot do each other’s jobs.
Google’s own list of what SEO professionals provide is a fair description of the scope:
- Review of your site content or structure
- Technical advice on website development, for example hosting, redirects, error pages, use of JavaScript
- Content development
- Management of online business development campaigns
- Keyword research
- SEO training
- Expertise in specific markets and geographies
- Optimizing for generative AI
In practice people cluster into technical SEO, content-led SEO, and the strategy and reporting layer that sits above both. The skill that distinguishes a good practitioner from a busy one is diagnostic: the ability to look at a site that is underperforming and work out which of the three stages is failing, rather than applying every tactic at once and hoping.
Hiring an SEO, and the warnings worth taking seriously
Google publishes guidance on hiring an SEO, which is unusual and useful, because it means the red flags come from the search engine rather than from a competing agency.
The ones that carry the most weight:
- “No one can guarantee a #1 ranking on Google.” Google adds that you should be wary of anyone claiming guaranteed rankings, a “special relationship” with Google, or a “priority submit”
- Limit access at first. On an audit, Google’s advice is to “only grant read access to Search Console (at this stage, don’t grant them write access)”
- Check the sourcing. Ask whether they “cite official Google documentation as supporting evidence for their recommendations”
- Be skeptical of tools. Google notes that it “doesn’t evaluate or endorse third-party SEO tools” and warns against tools claiming to be approved by Google
- “You should never have to link to an SEO.” A required link in the footer is payment extracted in a currency you did not agree to
- Ignore cold outreach. Google’s phrasing is memorable: reserve for unsolicited search-engine email “the same skepticism as you do for ‘burn fat at night’ diet pills”
- You carry the risk. “Ultimately, you are responsible for the actions of any companies you hire.” A penalty earned by a contractor lands on your site
Google also supplies the interview questions, and two of them are better than most of what gets asked: what results do you expect to see and in what timeframe, and will you share all the changes you make along with the reasoning behind them. An SEO who cannot answer the second one comfortably is not someone to give write access to.
How to Learn SEO
The field has no formal qualification, no licensing body and no agreed curriculum, which makes learning it a matter of choosing sources carefully rather than finding a course.
Start with the primary sources. They are free, and they are the only material that is authoritative rather than interpretive:
- Google’s SEO Starter Guide, which is the practical introduction and, in Google’s words, “describes much of what your SEO will do for you”
- Google Search Essentials, which defines the technical requirements, spam policies and key best practices
- The in-depth guide to how Google Search works, which is the mechanism this whole page is built on
- The search quality rater guidelines, which show how Google’s own raters are asked to assess quality
- Search Console, which gives you Google’s own record of how your site performs in Search and what it has indexed, rather than a third-party estimate of it
Then build something. Reading about search is a poor substitute for watching a page you published get crawled, indexed and found. A small site on a subject you actually know is a far lower-risk place to make mistakes than a client’s site, and it teaches the thing no article teaches, which is how slow and uneven the feedback is.
Develop one habit above all others. When you read an SEO claim, find out whether its source is Google, a controlled test, or someone’s impression. Those three are often written in identical confident prose, and distinguishing them is most of what expertise consists of in this field.
A final warning that Google issues and most training does not. Tactics that work by gaming a system rather than helping a reader tend to stop working, and the spam policies describe what happens when they are caught: sites “may rank lower in results or not appear in results at all.” Learning a tactic that is already on that list is worse than learning nothing.
How to Start
If you own a small site and want the shortest useful path, in order:
- Verify the site in Google Search Console. Free, and the one place you get Google’s own record of how your site performs in Search and what it has indexed
- Check what is indexed. The Pages report tells you which URLs Google has stored and why the others were left out
- Fix anything blocking crawling or indexing. An accidental noindex tag or robots.txt rule is common and costs everything downstream
- Write a clear title and opening paragraph for each important page. Say what the page answers in the words a reader would use
- Give each page one clear primary intent, covering closely related questions together rather than splitting them across pages. Two pages competing for a single intent is worse than one page doing it properly
- Link your pages to each other with descriptive link text, so a reader can follow their next question
- Record a baseline and wait. Weeks, not days
That list is deliberately short. For the step-by-step version with the first-month detail, see our complete beginner’s guide to SEO.
Frequently Asked Questions
What does SEO stand for?
SEO stands for search engine optimization. The British spelling, search engine optimisation, means exactly the same thing. It refers to the practice of making a website easier for search engines to find, understand and show to people searching for what it offers.
What is SEO in marketing?
In marketing, SEO is the channel that earns unpaid traffic from search engines, sitting alongside paid search, email and social. Search often reaches people expressing a need they have already identified, which is what makes it commercially valuable. Its feedback cycle is usually longer than paid search, since you are earning placement rather than buying it, though both the time required and the conversion performance vary by market and by page.
What is the difference between SEO and SEM?
SEM, search engine marketing, is normally used as the umbrella term covering both SEO and paid search advertising. SEO earns unpaid placement over time; paid search buys placement immediately and stops when the budget does. Some people use SEM to mean paid search alone, so it is worth asking which definition someone is using rather than assuming. One thing is settled either way: Google states that advertising with Google has no effect on your presence in organic search results.
What are the main types of SEO?
SEO is usually divided four ways by activity: technical SEO, which is whether search engines can reach and index the site; on-page SEO, which is what appears on an individual page; content, which is whether the page is worth showing; and off-page SEO, which is links and reputation from elsewhere. It also divides by site type into specialisms such as local, ecommerce, enterprise, international and news SEO, plus the newer work of optimizing for AI answers, which Google treats as part of SEO rather than separate from it.
How many Google ranking factors are there?
Google has never published a list. Its documentation says relevancy “is determined by hundreds of factors,” and the ones it names in that sentence are the user’s location, language and device rather than properties of your page. The widely repeated figure of 200 ranking factors is not a current Google figure. Any numbered list you encounter is a reconstruction from patents, public statements and correlation studies, which can be informative but is not authoritative.
Is E-E-A-T a ranking factor?
No, and Google says so directly: “While E-E-A-T itself isn’t a specific ranking factor, using a mix of factors that can identify content with good E-E-A-T is useful.” There is no E-E-A-T score, and search quality raters do not grade individual sites in a way that affects rankings. Google does say trust is the most important of the four aspects, and that its systems weight E-E-A-T signals more heavily for topics affecting health, financial stability or safety.
How long does SEO take to work?
Google states that some changes take effect in a few hours and others several months, recommends waiting a few weeks before assessing whether a change helped, and notes that not all changes produce a noticeable impact at all. There is no published fixed timeline. Any specific figure you see is someone’s experience rather than a documented rule.
Does SEO actually work?
It works in the sense that it improves your odds, which is also how Google describes it: sites following its guidelines are “more likely” to appear, with no guarantee for any particular site. It does not work in the sense of producing a specific position on demand. The effect is also hard to attribute cleanly, because rankings move for reasons outside your control, which is why recording a baseline before you change anything matters.
Can I pay Google to rank higher in organic results?
No. Google’s documentation states that it “doesn’t accept payment to crawl a site more frequently, or rank it higher. If anyone tells you otherwise, they’re wrong.” It also confirms that appearing in search results costs nothing, and that running Google Ads has no effect on organic placement. Anyone offering to pay Google on your behalf for organic rankings is describing something that does not exist.
Does Google prefer longer content?
No. Google’s guidance on helpful content lists writing to a target length among its warning signs of search-engine-first content, and answers the question in parentheses: “Are you writing to a particular word count because you’ve heard or read that Google has a preferred word count? (No, we don’t.)” Length should follow from how much the subject requires. A page padded to hit a number reads like one.
Why is my page indexed but not showing in search results?
Google names three reasons this happens: the content is irrelevant to users’ queries, the quality of the content is low, or robots meta rules are preventing serving. The first two are judgments about the page rather than technical faults, and the third is a setting worth ruling out first because it is quick to check. Being in the index means being eligible to appear, not being shown.
Is buying links against Google’s rules?
Buying links for ranking purposes is, yes. Google’s spam policies define link spam as creating links primarily to manipulate rankings and list “exchanging money for links” and “exchanging goods or services for links” as examples, including sending a product in exchange for a review containing a link. Paid links are acceptable when they are qualified with a rel=”nofollow” or rel=”sponsored” attribute, which tells search engines not to pass ranking credit. Sponsorship is fine; sponsorship that passes credit is not.
How hard is SEO to learn?
The fundamentals are learnable in a few hours. The difficulty is judgment: knowing which change matters for your site, reading a results page accurately, and telling a real drop from normal variance. The field also carries an unusual amount of advice for tactics Google has publicly said do nothing, so the most useful early skill is checking claims against Google’s own documentation. Google’s own view is that if you run a small local business, you can probably do much of the work yourself.
Do I need to hire an SEO?
Google’s position is that a small local business can probably do much of the work itself, and that the best times to hire are before a redesign or a new site launch. If you do hire, its published warnings are worth taking literally: no one can guarantee a number one ranking, grant only read access to Search Console during an audit, ask whether recommendations cite official Google documentation, and remember that you are responsible for the actions of anyone you hire.
Is SEO still relevant with AI search?
Google’s published position is that optimizing for generative AI search “is optimizing for the search experience, and thus still SEO,” and that no special markup or separate process is required. Two things do shift. AI systems can draw on the relevant part of a page rather than the page as a whole, which rewards self-contained sections, though they paraphrase as well as quote and Google advises against artificially chunking content. And being cited is not the same as being clicked: Search Console’s dedicated generative AI report shows impressions only, while clicks from AI Overviews and AI Mode are counted in the standard Performance report under the “Web” search type, mixed in with ordinary search, so an AI-specific click-through rate cannot be calculated from it. Crawl access still matters, but systems differ, so blocking one crawler does not mean your content is unreachable everywhere.
Sources and References
Every quotation on this page was taken from the source listed below on 4 November and checked word for word against it.
- Google: In-depth guide to how Google Search works, for the three stages, the absence of any crawl, index or serve guarantee, rendering during the crawl, the “hundreds of factors” wording, and the statement that Google accepts no payment to rank pages higher
- Google: SEO Starter Guide, for the timeline wording and the note that not all changes produce noticeable impact
- Google: Search Essentials, for the three components, the key best practices list, the statement that appearing costs nothing, and the note that meeting the requirements still does not guarantee crawling, indexing or serving
- Google: Spam policies for Google web search, for the definitions of link spam, keyword stuffing and cloaking, the nofollow and sponsored qualification, and the inclusion of generative AI manipulation
- Google: Creating helpful, reliable, people-first content, for the self-assessment questions, the E-E-A-T statement, YMYL, and the answer on preferred word count
- Google: Understanding page experience in Google Search results, for the absence of a single page experience signal and the position of Core Web Vitals
- Search Console Help: Page indexing report, for what “Crawled, currently not indexed” does and does not mean, and Google’s advice to diagnose rather than assume
- Search Console Help: Generative AI performance report (Search), for its coverage of AI Overviews and AI Mode and the fact that it reports impressions only
- Google: AI features and your website, for the statement that AI feature traffic is reported in the standard Performance report under the “Web” search type
- Google: Do you need an SEO?, for the list of SEO services, the hiring warnings, the ranking guarantee statement, and the position on third-party tools
- Google: A guide to Google Search ranking systems, for neural matching and BERT
- Google Search Central Blog: Google does not use the keywords meta tag in web ranking, the original 2009 announcement, still published
- Google: Guide to optimizing for generative AI features on Google Search, for the position that this is still SEO and the list of tactics to ignore
- Google: Overview of Google crawlers and fetchers
- Google: Introduction to structured data markup
SEO stands for search engine optimization: the practice of making a website easier for search engines to find, understand, and show to people searching for what it offers. This guide separates what Google documents from what the industry assumes — including what it confirms, what it denies, and what it has never actually said.
