How Search Engines Work (Basically)
Search engines used to be much simpler than they are today. In fact, they didn't even start out as "engines" but were curated lists of links that humans put together that were really "directories".
The Internet is a bit like a video game—if you want to win the game, you have to master the rules. And when it comes to search engines like Google and Bing, the rules are constantly changing. Coming out on top of the search listings requires ingenuity, dedication, and even a little bit of luck.
Over the years, Google and Bing have become increasingly sophisticated, and what was originally outlined as a relatively straightforward ranking algorithm has evolved into a complex system incorporating social media, image identification, artificial intelligence, and a whole host of individual factors that we’ll never know.
There are some mainstays to the search engine process, however.
First step: Identify relevance
The first thing search engines do is look at a body of content, examine a particular search phrase within that content, and “score” the phrase, assigning a value of relevance that determines a page’s importance, authority, and reliability. In turn, this helps decide whether a certain site will show up for a given search phrase. For Google, this complicated inner working is broken down into a 0-10 scale called PageRank.
A site’s given PageRank depends on a few factors:
- How long the web page has existed
- The locations and frequency of a given search phrase within the piece of content
- The number of other, reputable sites that are linking to the specific page the content is on
A link to a site from another page counts as a “vote” toward raising the site’s ranking. The more links to your site, the better its ranking will be. The more reputable the sites linking to yours are, the faster your score will rise.
Second step: Identify ranking
Once the relevance for a given search phrase has been determined, search engines assign rank. Rank is assigned based on priority and relevance, and relevance is based on the number of incoming links, so here too the links the come into your page—the number of them and their sources—matter greatly.
Links aren’t the only factors at play. Other elements, like the inclusion of certain search terms in the page, especially in the title tag, matter as well. A page with lots of incoming links from other relevant pages may not rank in search results well if it doesn't have on-page relevance itself. Links alone are not the answer, but they are part of it.
A new twist: Artificial intelligence
Not all websites have links coming to the page, but they may still have good, relevant content. To solve this problem, Google created RankBrain, which looks at pages much like a human would and then judges its merit. This solves the problem, helping make sure that just because a page has no links to it doesn’t mean it has to languish in search result purgatory forever. RankBrain helps ensure these types of pages see the light.
What is the difference between Google and Bing?
For most of us, when we think “search engine” we think Google, but Bing offers some unique differences in their ranking algorithms. SEO is still important, but Bing puts less priority on backlinking and more on pages that have been around for longer periods of time.
In some studies Bing's results were superior to Google's.
Bing doesn’t crawl through pages the way Google does. Instead, it focuses on ranking a fraction of the content and letting links do the rest. Where it outshines Google is its grasp of multimedia content. Bing’s algorithm is able to read and understand non-text content like Flash, videos, and images, factoring the meaning of the media content into the overall meaning of the page.
Finding the balance
Google and Bing offer differences in their algorithms, and a good web master should know how to achieve the benefits of both. Focus on producing quality content, keeping keywords relevant, and building strong linking relationships with reputable websites and you should be successful on both engines.