The answer? No one knows. Google uses over 200 ranking factors in their search engine algorithm, and the majority of those remain unknown to prevent people exploiting the system. It’s likely that even those working within Google don’t know exactly.
The good news is that we don’t need to know every single factor for effective SEO. By considering what Google chooses to display when a search is entered, we can narrow the ranking factors down to several core areas.
What is Google looking for?
Google’s main aim is to give searchers what they are looking for. To do this, it must work out what the user means by a search query.
The query itself forms a big clue, but interpreting the words effectively has been a problem it has taken years to solve. Google’s engine uses a new technique called natural language processing (NLP) to understand a search as just a human would. If you just searched ‘Apple’, for example, Google might think you were interested in the technology company. However, if you instead search ‘eating apple’, the results are completely different: the page is dominated by results about the fruit. Instead of just searching for ‘apple’, Google uses the word eating to narrow down what it thinks the user might be interested in.
Google also uses clues from other information, like the time, your location, and previous searches. If, for example, your recent search history consisted of football results and games, Google would be more likely to show you results about Barcelona football club if you simply searched for ‘Barcelona’. Conversely, if you had been looking at other Catalan cities, Google might be more likely to show you tourist information about Barcelona.
This change signals in increasing ability to go beyond the words themselves to show users what they want as effectively as possible. Search engines are increasingly focused on user satisfaction as the core aim. The ranking factors are therefore concentrated on which webpage is most likely to fulfil the users needs.
What is used to rank pages?
For page ranking, this means that everything Google uses to rank the pages feeds into this goal of satisfying the user’s search. The intention behind all the ranking factors is to work out whether your page is the one that will best answer the user’s search.
We’ve said it before, and we’ll say it again: content is key for SEO. Your page can’t answer the user’s search if there’s no information there.
Despite being a crucial factor, measuring content is difficult. Unless Google has a secret super-advanced AI, it isn’t grading your content like a school essay. Instead, it’s likely that Google uses a combination of contributing metrics that together give a broad picture of how users rate your content.
Links have always been a large part of Google’s algorithm. The original innovation that launched Google search was to treat the web as a network, where more central websites - those linked to most often - were taken to be more important and therefore put higher in search rankings. PageRank, the algorithm behind this, is still a part of Google’s ranking toolkit and influences results. The more inbound links your page has, the higher quality Google will assume the content is.
Bounce rate is another metric that may well be used, which measures the proportion of single page sessions - i.e., sessions in which the user only looked at one page on the website and then left. This could indicate that the content was not engaging enough to make visitors explore further, so a high bounce rate usually counts against your website. Similarly, dwell time and time on page could be used (we’ve covered the difference before here, but in a nutshell: dwell time always ends back at the search engine, time on page can end anywhere). The longer visitors spend on a page, the more informative they probably found it. Pages with lots of high-quality content necessarily take longer to read than short pages with low-level content; the former is also more likely to contain the answers to future searchers’ questions.
The text itself is probably analysed to some degree. Length is a factor, with longer pages tending to rank higher over very short ones due to being more comprehensive. It’s possible that the reading level of text is also analysed as this metric used to be available in Google Analytics. A simple reading level could appeal to a wider audience, but it shouldn’t come at the cost of information. Finally, images and videos may give your page a slight boost by acting as another content quality signal.
Of course, your content will only help you rank higher if Google thinks it will answer the query. The second set of factors therefore focus on how relevant your site is to a given search.
Keywords have always been important in this regard and continue to be so. The Hummingbird update, introduced in 2013, improved the ability of Google to understand the natural language content of a page. It allowed the crawler to understand the meaning of a page without having to rely on keywords. While this means keywords aren’t as vital as they once were, they can still help to effectively signal the content of a webpage. Forcing keywords at the expense of natural writing is likely to hinder rather than help your ranking, but judicious placement in your content will help search engines discover your content and display it to the right visitors. Keywords in the metadata can also help to signal your meaning.
Some of the other relevancy metrics are likely the same as are used to gauge content quality. If a site has a high dwell time after a particular search, it’s a safe bet that the site is relevant to that query. Google is likely to rank the page higher for similar searches in the future. Links probably also play a role in determining relevancy; Google may use the outbound links on your site to gauge what field your content is in and who to show it to.
Although not new, user experience seems to be becoming a more important factor for Google and other search engines. This year, Google introduced Core Web Vitals as part of a new package of metrics focusing on this aspect. Unlike so many other potential factors, there is no guess work - Google announced that these would be incorporated directly into the ranking algorithm. We’ve talked about Web Vitals before in depth here. The three metrics that form the set of Core Web Vitals focus on three key areas of user experience - load times, interactivity, and visual stability. While a site with high usability isn’t going to rank if it doesn’t have content or if it isn’t relevant, it can give the page an edge over other, less optimised sites.
Amongst the other 200 factors are many technical signals that may give your site a small boost. Security is a known one - sites with an SSL certificate (denoted by an https:// at the start of the url) rank slightly higher than those without encryption. Site architecture also helps. Google prefers well organised and easily navigable sites, although it isn’t clear whether this is a factor in itself or whether it just allows Google’s web crawlers to explore the site and discover the content more easily.
Improving your ranking
What’s the best way to improve your ranking then? The answer is simple: build a better website. An investigation into Google’s ranking factors can only come to one conclusion - there is no pre-eminent factor or SEO trick that can shoot you to the top of the results. With so many factors, most of which are still unknown, any attempt to game the system will only end up harming your site’s chances.
When asked on Twitter what the most important factor is for ranking at the top of the results page, Google search advocate John Mueller tweeted back to simply say ‘awesomeness.’ This statement isn’t as glib as it seems. Building an informative website that fulfils visitors’ needs and is easy to use is the only way to rise in the rankings.
Google’s main concern is to direct users to the website that will serve them best. The higher the chance that your website does that, the higher its rank will be. If you want to talk to us about how to make sure your website does this, get in touch with us.