Google's search algorithm is still shrouded in mystery, but some of the secrets of how it works have been revealed. Learn how Google selects search results.

Your ecommerce website is open for business, but how do customers find it? In addition to paid advertising and word-of-mouth, online visitors tend to use search engines like Google to discover your business. The higher you rank on the search engine results page (SERP), the more likely you are to get new business.

Google processes billions of searches every month and ranks results using complex algorithms based on hundreds of criteria. So how exactly does this algorithm work? Take a peek inside the algorithm to better understand search queries, key ranking factors, and how to improve your site’s visibility.

What is the Google algorithm?

Google's search algorithm is a constantly evolving system that retrieves content from Google's index in response to user queries. Google's index contains over 100 trillion web pages, and that number continues to grow. Google's ranking algorithm predicts which pages are most relevant to a search query, organizes them into a list, and places the most relevant results at the top. Although Google's algorithm is often referred to as a single entity, in reality there are multiple algorithms working simultaneously.

How does the Google algorithm select search results?

The technology that powers Google’s ranking system is a secret, but Google has revealed five main categories that its algorithm considers when ranking search results: meaning, relevance, quality, usability, and context. Here’s a more detailed explanation of each:

meaning

Google's algorithms use language models (AI systems trained on human language) to infer the intent of a search query. This includes correcting spelling and applying synonyms related to similar content.

The algorithm also measures a user’s preferred language and media type (e.g., images or videos), and considers locality to prioritize local businesses in search results. If a query includes trending keywords, Google’s algorithm will rank pages that focus on current events and breaking news higher.

relevance

When Google's algorithm evaluates the relevance of an indexed site, it essentially asks the question, "Does the web page use the words in the user's search query?" If the query keywords match the words on the page, it is more likely to be relevant.

The algorithm goes further, using aggregated, anonymous interaction data fed into the machine learning system to predict other relevant content. Factors such as page relevance, reliance on expert sources, and location proximity are weighted based on the algorithm’s interpretation of the meaning and intent of the query.

quality

Google's algorithm prioritizes results that pass quality tests, defining quality as what is most useful to users. To determine quality, Google looks for experience, expertise, authority, and trustworthiness (EEAT). If a well-known expert writes an article on an authoritative website, the page is more likely to receive a high EEAT score.

Backlink

When other trusted websites link to your web page, the algorithm concludes that the page is more trustworthy and ranks it higher. Creating unique and useful content increases your chances of building backlinks.

Usability

After content passes the meaning, relevance, and quality tests, the algorithm measures usability. This means that web content must be accessible, mobile-friendly, fast-loading, and provide a positive user experience.

Context

Context includes geographic location, search history, and search preferences (such as preferred language and “Turn on Safe Search”).

For example, a user in Philadelphia who searches for "hockey" will likely see results related to the Flyers, a professional hockey team in Philadelphia. A search made in Spanish will return results in Spanish. Google's algorithms also consider recent search activity as part of the context.

Factors that affect how Google's algorithm ranks your site

The inner workings of the Google algorithm are still shrouded in mystery, but we do know that five major ranking factors play a role in search rankings: backlinks, freshness, keywords, page experience, and subject matter authority. Let’s take a look at how these factors affect search rankings:

authority

If a website with authority on the topic links to your webpage, your site may rank higher in the SERPs. For example, a backlink may appear in a blog post or a high-quality product review on an authoritative site. Increase your chances of building backlinks by creating useful and engaging content that other trustworthy websites will want to link to.

Up-to-date

The algorithm prefers pages that are freshly updated. However, this factor varies, and some blog posts do not need to be up-to-date. For example, if you are looking for a definition of a term such as "What is Magstripe?", an older page may rank higher than a newer page because it scores higher on other metrics. Regularly check the accuracy of your content and update popular posts as needed to keep them up-to-date.

Keyword

Generally, the more frequently a relevant keyword appears on an indexed page, the higher it will rank for that keyword. Keywords should appear in the page title and meta title in addition to the body content. You can optimize your page to increase mentions of relevant keywords, but using too many repeated keywords (i.e. keyword stuffing) can actually be detrimental. Avoid keyword stuffing by incorporating keywords naturally into your sentences.

Page Experience

Google takes user experience (UX) very seriously. The algorithm measures page loading speed, determines site navigation, checks for mobile accessibility, and looks for meta tags that match the content of the page. Keep the user experience crisp by avoiding using large images that take a long time to load or stuffing too many ads on your web pages.

How can you stand out as an ecommerce site? Kyle Risley, SEO Lead at Shopify, suggests selling unique products that people are searching for. “These queries will be easier to rank for. This will identify a market/niche that isn’t being served effectively. On-page optimization will be the most valuable work you can do if you go this route.” Another option is to sell something in a highly competitive niche. “In this scenario, ranking will require on-page optimization and backlink acquisition,” says Kyle. “The quality and quantity of backlinks you need will depend on the backlink profiles of your competitors in that niche.” He also reminds us that on-page experience will always matter.

Important Google Algorithm Update

Google constantly adjusts its search ranking rules. Some are major algorithm updates , while others are minor updates that are barely noticeable.

On average, there are two or three major updates per year, which are called broad core algorithm updates. These updates can have a significant impact on your search results rankings. If you see a significant change in your SEO ranking performance, a core update could be the reason.

Here's a history of the most important Google algorithm updates:

Panda (Panda)

The Panda algorithm update was introduced in early 2011 and was integrated into the core algorithm in 2016. Panda evaluates web pages, assigns quality scores, and demotes pages that contain spam content such as duplicate pages, plagiarism, and keyword stuffing.

Penguin

Penguin was launched in the spring of 2012 and was created to limit negative link building, such as buying backlinks from link farms. It also analyzes irrelevant pages and spam links.

Hummingbird (Hummingbird)

This 2013 update was introduced in response to keyword stuffing and superficial content. It improved the interpretation of search queries by using natural language processing to infer the user’s intent.

Thanks to Hummingbird, even if a page doesn't exactly match a search term, if the semantic elements of the web page (synonyms and related terms) are closely related to the search meaning, the page can rank higher.

Pigeon (Pigeon)

In 2014, Google announced an algorithm that improves the accuracy of local search results and connects them to other ranking signals, most notably in Google Maps and web searches.

Mobile-first indexing (mobile-first index)

In April 2015, Google reorganized its index to prioritize mobile versions of websites over desktop versions. Sites that lacked mobile presentation or functionality were ranked lower. Officially announced in 2016 , Mobile-First Indexing, with recent updates in 2018 and 2020, has reinforced the importance of fast and usable mobile web pages. Mobile-first indexing changed which version of a page is used for scoring and ranking.

RankBrain

In October 2015, RankBrain was launched as a spinoff of the Hummingbird update, tackling three search result problems: relevance to search queries, low content quality, and poor user experience. RankBrain uses machine learning to explore the meaning behind search queries.

Google doesn’t go into detail about how RankBrain works internally, but it does mention that it’s the third most important factor used to rank results pages. Analysts speculate that RankBrain takes into account the context of a search term, which probably means semantic processing and taking into account a user’s search history.

Intrusive Interstitials Update (Update on insert ads that harm UX)

In 2016, Google began to address intrusive interstitial ads (pop-ups) that obscure web page content in its search index. This does not mean that all interstitial ads are bad or harmful to search engine rankings. Among them, those that abuse the user experience on mobile devices with small screens are particularly serious. The most egregious pages have been demoted since 2017 .

BERT

The BERT update, introduced in October 2019, makes heavy use of natural language processing to identify and downrank poorly written, unfocused, and poorly written text, while rewarding well-written content. BERT is the culmination of previous updates such as Panda, Hummingbird, and RankBrain.