Google Search Engine Algorithm Patents

This patent describes how Google scores results differently for normal and mobile queries. Mobile queries prioritize documents that are user friendly for mobile viewing by detecting traits such as:

1. Use of Wireless Markup Language
2. Containing downloadable content for mobile device
3. Use of mobile URL

Sometimes, there is a need to present mobile friendly results for a normal PC/Laptop, and also normal results for mobile users. Google will detect the need based on criteria such as:

1. Whether the query is a “mobile query”
2. Whether the search results exceed the threshold score
3. The number of mobile search results
4. The properties of the mobile resources found in the search
5. The properties of the generic resources found in the search

Based on its findings, Google might blend both normal and mobile results to be displayed in various ratios depending on the need.

Based on this patent we can conclude that you can adjust the mobile-friendliness of your website to suit different query needs. While it’s good to have a balance and optimize for both normal and mobile queries, Google sometimes scores “pure” websites, such as those that are only optimized just for mobile or those that are only optimized just for normal queries, better. So depending on your target platform, in certain cases it might be beneficial to have a specialization.

This patent tells us that for some mobile search results, given the limitation in screen size of mobile devices, there could be a need to maximize use of space by omitting less important parts of a search result. For example, a title might be less important to a user who wants to know directly whether the content of the search result answers their query, so Google will not display the title and, instead, show the snippet of the content that contains the query keyword.


What can be done is that if you are targeting for mobile platforms, don’t overuse your target keywords, and only include them in important parts of your content to ensure that when Google generates snippets based on your keyword, it will display the part that most accurately answers the query of the user.

This patent describes that while Google prioritizes mobile results in mobile queries, sometimes they still need to depend on the normal (desktop) version of a mobile website to rank them. This is because some mobile websites are too simplified and lacking in certain criteria that Google needs in order to properly score and rank them.

So, even though Google is creating a system to help rank mobile websites using its non-mobile version, try not to oversimplify the mobile version of your website as it might affect the ranking (make things easier for Google).

The patent reveals that in certain queries where location is an important aspect (e.g. nearest coffee shop) Google will rank the results based on two approaches:

1. If results are within the immediate area (broad area) of the search query, the results will prioritize “local prominence” – sort of like normal search results but based around that particular area.
2. If results are outside of the broad area, results will prioritize the nearest result to the search query- distance based.

From this patent it can be said that if you are operating a brick and mortar business, it is important to feature a consistent NAP – Name, Address, and Phone – since local searches are significantly affected by the location you display on your websites and other online assets.

What can be concluded in this patent is that Google considers a website Fresh if returning visitors spend more or around the same amount of time (not less) compared to their previous visits. So in order to gain Freshness score, we need to encourage returning visitors to stay longer.

What can be understood by this patent is, whether a website is considered fresh or stale depends a lot on the rate of how the content of the website is updated. The rate of update itself is divided into a few categories:

1. How frequently a website content is updated in general
2. How frequently a website content is updated for a specific time period
3. The update rate of a specific topic/section of a website
4. The trend change in update rate overtime (increasing or decreasing)
5. The update rate compared to other similar websites
6. The rate of the number of backlinks the website receives overtime

In general, what can be done is to keep your website updated regularly. And if you are posting about any recent topics, it is a good idea to update the content in smaller, more frequent chunks rather than waiting to pile them up into fewer bigger blocks, as the update rate is what matters more in Fresh contents.

This patent reveals that Freshness score not only take into account a website’s own Freshness score, but also the Freshness score of other websites linking to that website. It’s sort of like an “Off-page” aspect of the Freshness score.

Incoming links are scored based on the time the link was created (presumably the more recent, the better) and whether the linking website is updated within a certain period relative to the current time. The websites providing the links will then be grouped based on whether it is Fresh or not. If most of the incoming link comes from Fresh sources the “bonus” Freshness score granted to the receiving website will be more compared to if the links were to mostly come from Stale sources.

In conclusion, it can be said that all incoming links will help increase your Freshness score. But Fresh links from Fresh sources carry more weight, so if possible, focus on quality by earning links from Fresh websites instead of on focusing on quantity (might even be considered as linkspam) but include only Stale sources.

This patent reveals that Freshness is also influenced by the rate of how a website receives backlinks overtime. There are several aspects to this criteria:

1. A website is considered Stale if it receives decreasing backlinks overtime.
2. A website is considered Stale if existing backlinks starts to disappear (linker website is closed/link removed etc).
3. Even among websites that have increasing backlinks overtime, a higher score will be granted to websites that have an increment rate exceeding a certain threshold.
4. The system also takes into consideration the distribution rate of the backlinks (spiking numbers or slow distribution etc) in scoring Freshness. This criteria is also used to detect link spamming – an alarming rate of increase in backlinks might mean that link spamming is in place or the topic might actually be viral, in which case will most probably result in Google going for further (could be manual) evaluation.

Among the actions that can be taken in order to rank higher in Fresh results is that you can strategize your link earning in an increasing, steady rate, instead of in bulk initially (might also trigger spam detection) but which decreases overtime. It will also help to partner up and earn links from authoritative (educational or governmental) websites to avoid disappearing links from closed websites.

This patent tells us that Google might take into consideration the inception date of a website in ranking them. Inception date can be determined based on 2 factors:

1. The first time a website is indexed.
2. The first time a link to the website is discovered.

Google may adjust the ranking based on the inception date and the intent of a query. A query that requires a Fresh answer might favor newer websites while another query that requires more solid information such as FAQs might favor older websites. The inception date can also be used to even up older and newer websites by using the rate of the backlink growth rate from its inception date instead of just a total sum of all backlinks.

Based on this patent we can conclude that inception date matters for ranking, but doing it correctly means knowing your intent firsthand. There are some intents that favor older websites while some favor newer ones, so it’s important to adjust accordingly. This patent also lets us know that it is possible to compete with older websites with more backlinks since there are other factors involved such as backlink growth over time based on inception date.

This patent describes how Google understands the context of a website based on how the information is presented. Google can determine that a group of information is related or belonging to the same category if they are discovered as being presented in the same list (based on the HTML). While understandable to a human reader, it might be harder for Google to understand if the information is explained in plain sentences instead. For example, presenting a list of Saturn Facts in a list will enable Google to pick up all the entries of the list to answer a query, whereas if the list of Saturn facts aren’t presented in a list format, Google might only pick up the first line of the facts and miss the others.

So in conclusion, using text formatting to structure your content not only helps your (human) reader easily read through your it, it also helps search engines to better understand your content and to pick up important points to help them answer queries.

The patent explains that Google does not rank websites in a real time format, but instead ranks websites through the use of a stored representation of the website in its cache. This means that your website representation that is stored by Google might not be the latest version of your website and hence, not ranked accurately based on the present version. What can be done is to get Google to recrawl/index your site after a big update in order to get Google to search the latest representation of your website.