What Is SEO? Where to Start With Website SEO

Summarize with AI:

Link copied

SEO is the work of making a site findable and understandable to a search engine, and of making sure the person who arrives from a search finds what they came for. It does not buy a position and it does not guarantee one. Knowing which task solves which problem keeps the list short.

What SEO does and what it does not

Google's starter guide describes SEO in two directions. It helps a search engine understand your content, and it helps the person looking at a result decide whether to visit. The work is not written for a robot alone. The title in the result, the description under it and the page itself speak to the same person.

What it does not do is also clear. Google's documentation says it does not accept payment to crawl a site more often or to rank it higher. No rule and no person can guarantee that a page will be crawled, indexed or shown. An offer that promises a fixed position or a fixed timeline does not match the documentation either.

A change can show in results within hours or take months. Google suggests waiting a few weeks before judging whether something worked. Not every change produces a visible effect.

Three stages: crawling, indexing, serving

A page missing from search is not one problem. Google Search works in three stages and not every page passes each one. Finding the stage where it stopped saves time on the wrong fix.

Being indexed is not the same as being shown. Search Console can say a page is indexed while the page never appears in results. Google's documentation lists irrelevant content, low quality and robots meta rules among the reasons.

The three stages of Google Search and where to look when a page stops.
StageWhat happensWhere to look when it stops
CrawlingGoogle discovers the address through links or a submitted sitemap, downloads the page and runs it.A robots.txt rule, a server error, a login requirement. Search Console URL Inspection.
IndexingGoogle analyzes the text, the title element and the images. It decides whether the page is a duplicate or the canonical and stores it. Not every page is stored.A noindex rule, the page being a copy of another address, content quality. The Page indexing report.
Serving resultsGoogle ranks the pages it judges most relevant to the query, along with factors such as the person's location, language and device.The Performance report. Whether the page actually answers that question.

The technical base: does Google see the page like a user

However good the content is, the work stalls if Google cannot reach the page or sees something other than what the user sees. Treat the technical base as a pre-check, not as a ranking trick.

  • Is the page open to everyone

    An address that asks for a login, is closed by robots.txt or returns a server error cannot be crawled. After a launch or a move, it is worth checking whether a section was left closed by mistake.

  • Are CSS and JavaScript blocked

    For Google to see the page like a user, it needs access to the same resources as the browser. Hiding important components can lead to the page being misunderstood.

  • Does the content arrive through JavaScript

    Google runs the page with a recent version of Chrome when it crawls. Even so, it is easier to be sure things work when the main text and links are present in the first HTML. You can check this on your own page with URL Inspection in Search Console.

  • Content that changes by location

    If the page shows different content depending on the visitor's location, remember that Google's crawler mostly comes from the US. You should be satisfied with the version seen there.

  • Speed and interaction

    Page experience measurements are a separate subject. For LCP, INP and CLS, see the Core Web Vitals article.

Site structure: address, folder and duplicate page

Google's guide says a logical structure helps both users and search engines see how pages relate. You do not need to drop everything and reorganize the site for that. A search engine largely understands your pages as they are.

The address gives the user a hint. An address with meaningful words reads better than one made of random numbers. Google itself says that keywords in the address have very little effect on ranking beyond appearing in breadcrumbs. So shape the address for readability, not for ranking.

If the same content opens at several addresses, the search engine picks one. That is not a penalty, but it can spend crawling effort on addresses you do not care about and leave visitors unsure which page is right. The fix is for each piece of content to be reachable at one address. If you cannot move it, the canonical element is used. The details are in the canonical article.

If you run a multilingual site, the relationship between language versions is declared with hreflang. That is a separate subject: What is hreflang?

Search intent: the same topic, different questions

Google suggests writing with the words users will search for in mind. Someone who knows the subject well and someone new to it may search for the same thing in different words. You do not have to predict every variation. The documentation says Google's language-matching systems can understand how a page relates to many queries even when the exact words are absent.

So instead of pasting a keyword list into a page, decide which question the page answers. One page does one job. Answering the same question halfway on two pages leaves both the reader and the search engine unsure which one is the main page.

Illustrative example

Picture a small furniture workshop. This is not a client record, only a situation made up to show the distinction. A page about caring for a wooden table speaks to someone caring for a table. A person who wants a table made to measure is looking for something else: how to order, what information is needed, how delivery works. Pressing both into one page weakens both.

Once the question is separated, the page title, the first paragraph and the internal links separate on their own. Adding more keywords does not replace that.

Content: for the reader, not for the search box

Google says interesting and useful content will likely influence your site's presence in search more than most of the other suggestions in its guide. It lists the traits: text that is easy to read and well organized, content that is unique, up to date, and helpful and reliable for the reader.

Unique means not rehashing what someone else has written. Up to date means reviewing older pages from time to time, updating them, or deleting them if they are no longer relevant. Padding a page or repeating the same word adds nothing. Google says that repeating the same words excessively, even in variations, tires users and is against its spam policies.

In an article that makes a technical claim, accuracy comes before length. On a page where the reader will make a decision, a claim with no traceable source costs more than a ranking.

Internal links: how Google finds pages

According to Google's documentation, the vast majority of new pages are found through links. If a page is not linked from the menu or from any other page, discovery stays weak even when a sitemap exists.

  • Give important pages a path

    A person or a crawler following links from the home page should be able to reach every page you consider important.

  • Let link text describe the page

    Link text tells users and Google something about the target page. Instead of "click here", write what the target is.

  • Link to the same address

    When linking from inside the site, use the address you treat as canonical, not a duplicate. Consistent linking shows Google your preference.

  • Trust what you link out to

    If you link to a site outside your control, make sure you trust it. If you do not but still want to link, use nofollow or a similar annotation. Sections that accept user content, such as comments and forums, should add it automatically.

How you look in results: title and description

The title in the result and the description under it influence whether a person clicks. The title element is one source of the title link, and a good title is unique to the page, short and accurate about its content. Google can also build the title from other text on the page. Why it changes, and why a change may not show, is covered in detail in the title link article.

The snippet comes from the page's own content. Sometimes it can come from the meta description tag. A short description that is unique to the page and carries its most important points works well. It is also worth knowing that Google does not use the keywords meta tag.

A starter checklist

If you are starting from zero, this order removes the most uncertainty for the least effort. Visibility first, structure second, detail last.

  • Search for your site on Google with the site: operator. If you see results, you are in the index. If not, look for a technical blocker first.
  • Add your property to Search Console and verify ownership. The Search Console article walks through the steps.
  • Check whether robots.txt, noindex or a login requirement keeps an important page closed.
  • Inspect the home page and one service or article page with URL Inspection. Does the page Google sees resemble what the user sees.
  • Make sure each piece of content is reachable at a single address. Redirect duplicate addresses or mark them with canonical.
  • Link every important page from the menu or from another page.
  • Write a title and a short description specific to each page.
  • Write in one sentence which question each page answers. If two pages answer the same question, consider merging or separating them.
  • A small site does not need a sitemap. If the site is large, new or media-heavy, an XML sitemap helps.
  • Wait a few weeks, then look at the Performance report to see which queries show your page.

What you do not need to focus on at this stage

Many ideas that circulate about SEO are out of date or were never true. Google's starter guide separates some of them from its own point of view. The list below is taken from Google's document. The general message is: do what suits your business best.

Topics Google's starter guide suggests not prioritizing.
TopicWhat Google says
Keywords meta tagGoogle Search does not use this tag.
Keyword stuffingRepeating the same words excessively tires users and is against the spam policies.
Keywords in the domain or URL pathLittle effect on ranking beyond appearing in breadcrumbs. Choose the domain name for your business.
Content lengthDoes not matter for ranking on its own. There is no magic word count.
Number and order of headingsGood practice for screen readers, but the order does not matter for Google Search.
Duplicate content "penalty"The same content at several addresses does not cause a manual action. It is inefficient. Copying others' content is a different matter.
E-E-A-TThe document states it is not a ranking factor.

Sources