How Google ranks websites: the ranking factors that really matter

Google ranks websites in three stages: it crawls the page, adds it to its index, and, when someone searches, compares it with all the other relevant pages using hundreds of signals. Those signals boil down to four questions: does Google see the page at all, does the page answer what people searched for, do other sites recommend it, and is it easy to use. All serious SEO work is really work on these four questions. In this guide we break down each one, show how to check it on your site, and explain what to fix first.

What is organic SEO on Google, really?

Organic SEO on Google is the set of actions that gets your pages to appear high in the regular search results, without paying for each click. When someone searches for אינסטלטור בתל אביב ("plumber in Tel Aviv") or ייעוץ מס לעצמאים ("tax advice for the self-employed"), Google picks the pages from its index that best fit the question and ranks them. The goal is for your page to be one of them, in a spot where people see it.

The difference from paid advertising isn't only the price. With an ad, you pay for the position, and it disappears the moment the budget runs out. In the organic results you can't buy a spot: Google gives it to the page it judges to be the best answer, so everything in this guide is about how Google reaches that judgment. A page that gets there keeps bringing traffic even without ongoing payment, which is why organic SEO is considered an asset rather than an expense.

How Google works: crawling, indexing and ranking

Google describes the process in detail on its How Google Search works page. In short, there are three stages, and each one can block an entire site from the results:

1. Crawling: Google's robot, Googlebot, discovers pages through links and the sitemap. A page that nothing links to and that isn't in the sitemap may never be discovered at all.
2. Indexing: Google analyzes the content, works out what the page is about and stores it in the index. This is where headings, content structure, images and structured data have their effect. A page that looks like a duplicate of another page may not get in.
3. Ranking: At the moment of the search, Google pulls the relevant pages from the index and puts them in order. This is the stage where all the signals covered later in this guide come into play.

The four engines that move rankings

Google uses hundreds of signals and describes some of its systems in its guide to ranking systems. For practical work, it helps to group them into four engines. Almost every site that's stuck is stuck because of one of them, and finding the weak engine is most of the work:

• Crawling and indexing: If Google doesn't see the page, nothing else matters.
• Search intent match: The page is crawled, but it isn't the kind of page Google shows for that keyword.
• Authority: The page fits, but there aren't enough outside signals to justify preferring it over the pages already there.
• Experience: The page rises and then falls, because visitors don't find what they were looking for.

Above the last three sits content quality, which affects all of them. Let's go through them one by one.

Search intent: what decides which page can rank at all

Google tries to understand not only what people searched for but why. The same topic can come with four different intents:

• Informational intent: איך מחליפים גלגל ("how to change a tire"). The searcher wants an explanation or a guide.
• Comparison intent: איזה מזגן הכי חסכוני בחשמל ("which air conditioner uses the least electricity"). The searcher is comparing before buying.
• Navigational intent: כניסה לחשבון בנק ("bank account login"). The searcher knows where they want to go.
• Purchase intent: תיקון מזגן בחולון מחיר ("AC repair Holon price"). The searcher is ready to act.

Each intent calls for a different kind of page. For purchase intent you build a service or product page, and for informational intent you write a guide. If the top pages for a keyword are guides, a sales page will almost never rank there, no matter how many links it has. The check takes a minute: search for the keyword and see what kind of pages come up. That's what decides what to build, not the other way around.

Proper keyword research runs this check for every keyword on the map and assigns each keyword the type of page that fits it. There's more on this in our search intent (in Hebrew) guide.

Content quality and E-E-A-T: why Google trusts certain sites

Google rewards content written for people: content that answers the question fully and adds something that doesn't already exist on 10 other sites. It spells this out in its guidance on creating helpful content. An example: an article on "how to choose an air conditioner" that settles for the manufacturer's spec sheet is thin content. An article that explains how much cooling power a room of a given size needs, compares actual power consumption and points out common mistakes is the content Google wants to show.

To judge whether content can be trusted, Google uses a framework called E-E-A-T:

• Experience: Has the writer personally done what they describe?
• Expertise: Was the content written by someone who knows the field?
• Authoritativeness: Are the site and the writer recognized as a source in the field, through links, mentions and reputation?
• Trust: The heart of the framework. Is the site transparent and secure, with real contact details and accurate information?

E-E-A-T itself isn't a single ranking factor you can switch on, but the signals that make it up do have an effect. What to do in practice: an author page with a real name and area of expertise, clear contact details (address, phone, email), reviews on Google and on outside sites, a last-updated date on every article and real examples from your work. In health, money and legal topics, the effect of these signals is especially large. There's more in our article E-E-A-T: how Google measures trustworthiness.

Google doesn't rank words, it ranks trust. The more the site is seen as a reliable source, the more stable its rankings and the less sensitive they are to updates.

Authority: links, mentions and three layers of signals

Backlinks are still one of the strongest signals. In Google's eyes, every link from another site is a recommendation, and the weight of the recommendation depends on who's recommending: a site relevant to your field is worth much more than a random site, and links from spam sites can even do harm. In its spam policies, Google spells out which link patterns it considers manipulation.

You can think of the signals in three layers. The first is on the page: whether the content answers the question, in a readable structure and at a depth that suits the intent. The second is off the page: how many independent sources refer to you, and in what context. The third is what happens after the click: whether the visitor found what they were looking for. Many sites invest everything in the first layer, a little in the second and nothing in the third, and that's exactly why they rise for two weeks and then drop.

At VeloLinx we build the second layer with original articles on niche sites built to order for the client's field, rated DR/DA 40 to 90, as dofollow links with a report for every link and a free replacement if a link drops. You can order them on our buy backlinks page, and they're a fixed part of every monthly package. For more on building a profile Google trusts, see our guide to link strategy for a healthy profile.

Topical authority and content clusters

A site that covers one topic in depth, with pages logically linked to each other, ranks better than a site that touches on 10 topics superficially. Google understands natural language and whole topics, not just single words, so the old method of a separate article for every keyword, with no connection between the articles, is losing its power.

The structure that works is called a topic cluster: a central page that covers a broad topic, such as your service page, with articles alongside it that go deeper into subtopics and link to it and to each other. The internal links pass strength from the articles to the central page, and Google gets a clear picture of a whole field. Five articles around one topic beat 15 scattered articles, and for most businesses two to four good articles a month are enough, as long as they cover real search intents.

Topical authority builds slowly. The first sign that it's starting to work is showing up for long-tail keywords within the topic, and it comes long before the main keyword moves.

User experience, speed and Core Web Vitals

Google measures the loading experience with three metrics called Core Web Vitals, and sets clear thresholds for them on web.dev:

• LCP: How fast the main content loads. The target: up to 2.5 seconds.
• INP: How fast the page responds to a visitor's action. The target: up to 200 milliseconds.
• CLS: How much the layout shifts while loading. The target: up to 0.1.

Speed alone won't lift a weak page to the top of the results, but a slow site loses visitors and loses out when its content and a competitor's are similar. A clean structure matters too: readable URLs, correct canonical tags, full mobile compatibility and a secure connection. You can check the values for free in PageSpeed Insights, and it's best to start with the pages that bring the most traffic. There's more in our guide to site speed and Core Web Vitals. A site built right from the start saves months of fixes, which is why our business websites have this foundation built in from day one.

What happens after the click: behavior signals

Google states explicitly that it uses aggregated, anonymized interaction data to assess whether the results it shows are relevant to the search. It doesn't publish a formula for dwell time or bounce rate, so there's no point chasing a single metric. What is clear: a page that visitors open and then immediately leave, going back to the results to try another one, didn't solve their problem.

The practical meaning is simple: give the answer fast, at the top of the page, and don't bury it under three paragraphs of background. A visitor looking for a price wants a price, or at least a range, not a theoretical article. A page that leads the visitor to the next step (another page, a form or a call) shows that it did its job.

AI Overviews and answers from AI engines: what has changed

For some searches, Google shows an AI-generated overview above the results (AI Overviews), with links to sources. At the same time, more people are asking ChatGPT, Gemini and Perplexity. Shallow content that only summarizes existing information is losing traffic to these summaries. In-depth content with real experience, accurate information and a clear structure stays relevant, and can also appear as a source inside the answers themselves.

In its guidance on AI features in Search, Google makes clear that there are no additional requirements to appear in them, and that established SEO practices are also the way to get there. We wrote about optimizing for AI engines in our guide to SEO for AI search engines.

How to find the weak engine on your site

The check is quick, and you can do most of it in Search Console:

Each diagnosis has a different treatment, which is why general advice like "add more content" only helps by chance. The free SEO audit covers the technical part of this map, such as robots.txt, noindex, the title and the description. You check the other rows in Search Console and against the pages already on page one.

The order of fixes: what to do first

The order that yields the most in the shortest time:

1. Make sure all your important pages are crawled and in the index, and that there are no duplicates among them competing with each other.
2. Improve the pages that already get impressions but few clicks. This is almost always the fastest win, because the work is done on traffic that already exists.
3. Fix the intent match on pages that show up for the wrong keywords.
4. Build authority for pages that match the intent but are stuck behind stronger competitors.
5. Only then expand into new content for intents that don't have a page on the site yet.

The full plan, broken down month by month, is in our article the month-by-month work plan.

Checklist: your site against the ranking signals

1. Every important page is in the index, and the sitemap is up to date in Search Console.
2. Every important keyword has one page responsible for it, of the type Google shows for that keyword.
3. The main keyword appears naturally in the title tag, the H1 and the description.
4. The answer to the searcher's question appears at the top of the page.
5. The topic is covered in depth, at least as thoroughly as on the pages already on page one.
6. Articles link to their service page, and the service page links back to the articles.
7. There's an author page, full contact details and a last-updated date.
8. LCP, INP and CLS are within the targets on mobile.
9. Outside links are built at a steady pace from relevant sites, and there's a full list of them.
10. The metrics are checked once a month against the previous month.

What sets the pace

Three main factors: the level of competition for the keywords (עורך דין גירושין בתל אביב, "divorce lawyer in Tel Aviv," is much harder than עורך דין גירושין בעפולה, "divorce lawyer in Afula"), the age of the domain and the authority it has already built, and the pace of investment in content and links. A new site starts with no history, so it should start with less competitive long-tail keywords. A small business can win on local, focused keywords long before it moves on national ones. No one can set in advance a date when a page will reach a given position, and anyone who names such a date doesn't know your field. We wrote about realistic timing expectations in our article how long it takes to see results.

How we work on ranking signals at VeloLinx

Every project starts with a diagnosis: where the site stands against its competitors, which pages already appear and where the weak engine is. From there we build a keyword map and a content plan aimed at your customers' search intent. The monthly packages include links rated DR/DA 50 to 90 on niche sites built for your field, together with on-page work on the site itself: titles, structure, internal links, structured data and speed. Every month you get a report that shows what was done, where each link was placed and what changed in the metrics. You can see all the plans and prices on the organic SEO packages page.

Frequently asked questions

Why isn't my website showing up on Google at all?

The common reasons are a block left over from the development environment (a noindex tag or a block in the robots.txt file), a page that no internal link points to, or a new site that hasn't been crawled yet. The "Pages" report in Search Console shows right away whether Google knows the page, whether it's blocked or simply not crawled yet, and you can also request a crawl for a single page.

Is E-E-A-T a ranking factor?

Not as a single factor you can switch on. It's a framework Google uses to assess quality and trustworthiness, and it rests on signals that do have an effect: who wrote the content, what their experience is, who links to the site and what people say about it. In health, money and legal topics, its effect is especially noticeable.

Does dwell time affect Google rankings?

Google doesn't publish such a metric as a ranking factor, but it does say it uses aggregated interaction data to check whether results are relevant. In practice, a page that visitors leave immediately to try another result didn't answer the question. The fix is to improve the answer, not to try to stretch the time on the page artificially.

What is a content cluster, and how does it help rankings?

A cluster is a central page that covers a broad topic, with articles alongside it that go deeper into subtopics and are linked to it and to each other. The structure helps Google understand that the site covers a whole field and not just a single article, and it passes the strength of the internal links to the page that matters most to you.

Is updating an old article better than writing a new one?

Often, yes. An article that has already been crawled and has a history responds to a substantial update faster than a new article does. A new article is better when it targets a search intent that doesn't have a page on the site yet, or when the existing page is about a different topic.

What should you do with pages that get no traffic at all?

First check whether they target an intent already covered by another page, and if so, merge them and redirect the old URL to the new one. If the topic is relevant but the content is thin, deepen it. Thin pages that aren't relevant to the business are better removed, so they don't drag down the overall impression of the site.

Do Google's AI Overviews hurt organic SEO?

They change it. Shallow content that only summarizes existing information loses clicks to the overview, but in-depth content with experience and original angles stays relevant, and sometimes appears as a source inside the overview itself. Google notes that appearing in it requires no special optimization beyond regular SEO work.

Can a brand new website rank on Google?

Yes, but it takes patience. A new site starts with no history or authority, so at first you focus on less competitive long-tail keywords that bring the first appearances. At the same time, you build in-depth content and a link profile, and as the domain gains trust, you move on to more competitive keywords.

What to do next

The first step is knowing which of your engines is weak. The free SEO audit shows whether a noindex tag or a canonical pointing to another URL is keeping the page you enter out of the index, and what to fix on it first. You can see which keywords the site already appears for in the Search Console Performance report, and you can go over it with us in a free intro call. If you'd like to work on it together, the plans are listed on our organic SEO page, and you can message us on WhatsApp: 050-741-8449.