Detailed History of Google's Algorithm & Updates
What’s the Science behind the SEO mayhem?
Since 2000, Google has been making changes to its algorithm constantly. The algorithm is changed several hundred times a year, mostly comprised of small changes that don’t have a significant influence on search results. Fine tuning, if you will, to keep up with user complaints and requests and glitches in functionality. Every now and again Google will roll out a major algorithm change that hits SEO news sites across the web and sends SEOs into a frenzy, trying to compensate for the changes to keep their clients at the top of Google searches. Today I’m running through a thorough history of algorithm updates, large and small.
The Google Toolbar was introduced
This was the announcement that started it all. In 2000 Google announced the launch of their browser toolbar and the Toolbar “PageRank.” The toolbar is a plug-in that is compatible with users’ web browsers and brought Google to the fore-front of search engines. It enabled web surfers to use a Google search from any webpage on the internet, whether that page had it’s own search engine or not. The seamless integration of the toolbar also ranked pages in terms of relevance to the topic searched, creating the start of SEO and the quest to provide top ranked search results so searchers could find business in the jungle of the internet. It was a revolutionary plugin and made Google visible all across the internet, helping it skyrocket to the leading position its had in the search industry ever since.
Fall 2002 Update changes everything – SEOs in a Rage
After the Toolbar, the updates cooled down for a year or two until fall of 2002. This, though an unnamed update, was the first of the major game changers. The Fall 2002 update sent SEOs into a rage, some calling for a movement from Google all together. There was talk about top-ranked pages falling all the way to page 20 without any changes in the site. Nobody could figure out what was going on, how search was changed, and what the point of it all was. People were calling for Google’s head. The update was certainly not a popular one among SEOs. Aside from the unpopularity, the update was also done all at one time. This differs from the rolling updates of current times. The algorithm was changed in a major way without warning, unlike the constant small updates and occasional major updates that we experience now. Some theories hypothesize it was the first major step in Google’s crusade to provide nothing but the best and most relevant pages. A way to separate the quality sites from other sites that may have been spam and were simply manipulating Google to get a high Page Rank. Whatever the actual reason, it served almost as a stock market crash for many SEOs, starting the never-ending game of chess with Google.
Update Frenzy of 2003 - Little did we know this was only the beginning!
The Boston update of 2003 marked the first of Google’s updates that would be named. Keep in mind that hurricanes and winter storms are among the few events that receive names. A fitting likeness perhaps, as some updates have wreaked havoc on search results much like a natural disaster. At this point Google was planning monthly updates, whether it be major algorithm changes or simple tweaks to indexes. Updates that would follow during 2003 were Cassandra, Dominic, Esmerelda, and Fritz, spanning between April and July of that year. Cassandra focused on link quality through the evaluation of link-crossing. This brought a focus on link text, page titles and quality of links to the front of the SEO conversation.
Dominic, Esmerelda and Fritz marked the end of the monthly update, an idea that didn’t last long, and signified the implementation of incremental updates. Dominic changed the behavior of web crawling bots that rank information. Esmerelda and Fritz definitively marked the move from what was being labeled “The Google Dance”, or significant monthly index overhauls, to “Everflux”, which was small changes to the algorithm and indexes happening daily. This helped to speed the efficacy of updates as well as not completely change the search landscape every time an update was implemented. The dance generally ranged from three to 14 days.
The Florida update came at the end of the 2003 frenzy but definitely made a splash. It was another sea-change in the Google indexes with people up in arms about severe drops in Page Rank. However, with severe drops there are also severe gains for other pages. The update put an end to many SEO tactics of the time like key-word stuffing, and brought up questions about ad purchasing. A major theory at the time was that pages purchasing ads from Google had the upper hand with the roll-out of the Florida update and that it was designed to promote Google ad sales. This however, was only a theory, and one that Google staunchly denied. Despite the controversy, SEOs were again scrambling to figure out the new algorithm and implement new search tactics.
Austin and Brandy
Austin was an update that hit in January of 2004 and continued the scourge on dishonest SEO practices like link farms and hidden text. Sites using tactics such as these were suffering or even being banned from Google indexes. It also brought the Hilltop Algorithm into play, an algorithm used to find relevant pages or sites to the keyword searched. It has a special focus on expert or authority sites, increasing the relevance and credibility of the search results to provide the highest quality sites relating to the searchers query.
Brandy is the update that followed Austin. It expanded indexes and continued the quest for higher authority results. It also introduced Latent Semantic Indexing to the Google game, a technique that has the ability to identify concepts of pages and articles based on the terminology used in them. At a very base level, certain terms and vocabulary are often used together when describing concepts or ideas. LSI enabled these terms to be identified and the meaning of large bodies of text to be derived from them. It was a major progression in Google’s keyword analysis and knowledge. Just after “Stella got her grove back,” Google did as well!
Highlights of 2005
2005 was host to 9 updates, the first of which regarded the introduction of the no-follow attribute to links. This attribute was designed to improve link quality and eliminate spam. It weeded through the unverified or un-authoritative links by enabling them to be identified by site owners. This would prove to be a vital change for links in the future.
The next significant update of 2005 was the XML Sitemap update. This newly implemented protocol allowed webmasters to directly influence the crawling of their sites by Google’s web bots. A sitemap is a list of URLs within a site that are available for bots to crawl. The sitemap protocol allows its to more intelligently crawl sites, making crawling more effective. Google also took a stab at personalization in 2005 when they started using an individual’s search history to tailor search results. It didn’t have an immediate impact but the concept would go on to be implemented in other Google updates and practices.
Towards the end of 2005, the Jagger update hit. This update was rolled out in three parts and they became known as Jagger1, Jagger2, and Jagger3. This update made a bigger splash, reminiscent of updates like Florida. It again targeted links like reciprocal links and paid links, as well as content duplicates. The duplicates were mostly targeted in URLs, as an attempt to decipher if two slightly different URLs were intact similar or the same content.
2006 passed without much adieu in terms of algorithm updates but in 2007 Google unveiled their universal search. This, though not a traditional algorithm update, vastly changed the way search results would appear for users. The universal search blended results from many different categories of search into one large result, or a “universal” result. What does that mean? Before the switch to universal search there were many other vertical search varieties. A vertical search is essentially a specialized or specific search. If a user searched a topic and only wanted results of news sites, they would click the link for news above the search bar. While the user can still do this, the universal search brings up results from all categories, whether it be news, images, maps, etc. The change was designed to bring content located in those vertical searches into the view of users that previously ignored the vertical search tabs.
Two updates , called “Buffy” and “Dewey,” followed the Universal Search update. Buffy was a smaller update that didn’t have any major ramifications but Dewey seemed to be slightly more than that. Some SEOs reported better page ranking, while others reported the opposite. It was unclear as to the exact changes Dewey made but a definite change in indexing occurred. The big update of 2008 was Google Suggest.
Google Suggest is a feature that shows search suggestions in a drop down menu under the search bar while a user is entering search criteria. The suggestions are pulled from searches other users have performed using similar terms, ideas and keywords. Google Suggest had already been available on other platforms like the Google Toolbar, the Firefox search bar and the iPhone, but the 2008 roll-out made it available on Google’s homepage. The development started in 2004 but Google wanted to refine and perfect it before bringing it to the homepage.
Vince Favors Big Brands? – The Need To BRAND started in 2008!
The update dubbed “Vince” had SEOs talking about big brands in a big way. Though Matt Cutts said the update wasn’t a large scale one, search results seemed to be rendering a much higher percentage of big brands in the top rankings for huge keywords. Cutts addressed these concerns in a video in which he outlined that trust and quality were taken into higher consideration when retrieving results for generic search queries. This led many people to draw the conclusion that Google considered big brands more authoritative and more trustworthy.
Real Time Search
Also in 2009 was the unveiling of real-time search, a concept Google had been talking about for some time. Well in this update it was finally functional. What Real Time Search does is provide results for queries in a real time element. As results, updates, or sources come in the search results are automatically updated to show the most current and authoritative information. That’s not to say that Real Time indexes or provides every single new piece of information or update, but it does index reputable or trusted sources as the information comes in. This was another step toward the constant influx of updated information we experience today with live Twitter feeds, on the fly news updates and Facebook statuses.
Caffeine and 2010
2010 was a busy and significant year for Google algorithm updates. Google Places became more prominent in the Local Business Center, making it separate from its previous position within Google Maps. It became more closely tied to specific local searches and local advertising. But the first big ripple of 2010 was what came to be known as the May Day update. May Day was an authentic algorithmic change that targeted long-tail search results as opposed to head search results. Just to give a quick definition for those of you new to this, a head search is a generic query. For example, what would happen if you search for “hats” or “winter hats.” You will receive a wider variety of search results that may have anything to do with hats or winter accessories. Now if that search were to be a long-tail search, it would look more like this: “Winter hats with tassels,” or “Winter hats with cookie monster on them.” A long-tail search is a more specific search query, usually longer than two words. May Day didn’t affect searches on a wide-scale basis like previous switches, but SEOs focusing on long-tail searches saw major repercussions. It was another change designed to focus on quality, altering the way Google analyzed long-tail searches so as to weed out any results that may be unrelated that rank and appear for specific searches anyway. It was not a temporary change, but a change that was designed to remain. It again forced SEOs to evaluate their sites for quality.
Caffeine was the banner update of 2010. Though not an algorithmic update, it was a major overhaul of the Google indexing infrastructure. It was designed to bring nothing but the freshest results to users. As noted by Google in their description of Caffeine, it provided 50% fresher results than previous indexes. The major change was taken on to keep up with the demand for real-time, on-the-go information. Users were expecting to see the newest results, and publishers were expecting to have their most recent work shown in search results ASAP. It’s important to remember that when a user searches on Google, they are not searching the live web. Google is more like a compilation of the web that’s made available in an enormous book with indexing, so that a surfer can find the information they are looking for quickly and efficiently. With that being said, in order to appear on Google, a webmaster’s site needs to be able to be crawled and indexed by Google bots. The infrastructure change of Caffeine made these indexing results available more quickly than previous infrastructures. After Google bots had crawled sites and added the information to the cache of info that is Google, the results could now appear for searchers in mere seconds. I had just entered the SEO world at this point, which was by complete accident (I met Patrick Coombe, who was an acquaintance at the time. He is the only reason I know SEO exists.) and boy did I stumble into the hottest industry of the next 4 years!
The rest of 2010 was comprised of smaller, more specific updates and changes. Among them were Google Instant, the taking into account of the negative review, and the use of social media in ranking. Google Instant was a step up from Google Suggestion that fell in line with the Caffeine infrastructure change, making search suggestion appear instantly as a user typed queries into the search box. The negative reviews made a slightly larger splash but mostly because the change was prompted by a New York Times article about how a specific site was ranking based upon negative reviews. Google rarely allows outside forces to influence its updates or changes in search, but they followed suite by altering the algorithm to account for and rank sites for negative reviews following the publishing of the piece.
Google also finally confirmed that they were taking social media into account when ranking pages. Though they said it, at the time, was not a major factor, they were taking social media platforms such as Facebook and Twitter into account when evaluating a page for ranking and indexing. Quality was again a part of the conversation saying that Google was attempting to verify the authority of such social media persons or pages being used in the evaluations.
Penalties and Panda
2011 seemed to be a year of sharpening up and warnings issued by Google. The year kicked off with the very public penalizing of the company Overstock for violating Google’s webmaster’s guidelines with its SEO practices. The company was found to be encouraging members of the academic community to post links to its website, giving Overstock .edu links, in return for discounts. A site having a .edu link tends to rank higher in Google search results because websites of universities or colleges are thought to have lots of authority. A link from them is considered trusted, quality and authentic. It was a clear stand by Google against below-board SEO practices, with Overstock dropping from ranking at the top for many keywords, to appearing on 5th and 6th pages. It dropped their visibility and likelihood of being found in a Google search tremendously. JC Penny also suffered similar penalties later in the year. Google would continue to make its quality crusade with the upcoming Panda update.
Panda, also known as Farmer, was a heavy hitter. It cracked down on quality in a big way, directly effecting 12% of overall searches. Some sites, like Suite 101, saw a descent in traffic of up to 94%. This update was a major overhaul designed to make sure that authoritative and credible sites were ranking higher SERPs than sites with thin content but high optimization. It also targeted sites with heavy and obnoxious ad presence that took away from the content users were trying to find. Yet another update that focused on quality for the user. Panda would continue to be rolled out and updated throughout the next few years so I’m not gonna bore you with each and every update but try to focus in on Panda updates with a lasting or significant impact.
Though Google+ is another non-algorithmic related topic, I found its introduction significant to include. Google+ is Google’s best stab at social media, and was successful. Within a couple of weeks Google+ hits users registrations reaching the 10 million mark and claimed that billions of things were being shared each day on the social platform. Its most popular feature was the ability to groups your “friends” or contacts into different circles, hence the features naming of Circles. The feature took on the idea that platforms like Facebook ignored that fact that people have friends that are closer to them than other friends.
Encryption and Freshness
After four more Panda updates Google jumped on the privacy tip and released an update that encrypted search queries to improve web privacy for users. Though nice for users, the encryption negatively affected organic search components and displayed error type messages for certain search criteria. They also took another step toward being as current as possible with search results in another update designed to improve the freshness of results. The update was another improvement on the fluidity of fresh information.
2012 was another profession-altering event
The end of 2011 brought the new event of Google releasing what began as 10 packs of updates with small explanations. They would release two of these at the end of the year and continue the trend through the new year with a much larger scale “update information packet”. Many felt this was an attempt to be more available and transparent for both SEOs and the public. From 2012 to current there have been such a flurry of updates that I couldn’t hope to retain your attention through all of them. So from here out we’re gonna bring you the big boys.
Penguin was yet another update targeting web spam in a major way. The algorithm change focused on separating out sites that use techniques like keyword stuffing, link schemes, intentionally duplicated content, over-inundation of advertising and other black-hat SEO practices. Google expected it to have a 3% effect on all searches. It was another large step in Google’s quest to have the webmaster guidelines be adhered to and penalize violators through dropped search rankings.
Pirate was another update in the summer of 2012 that targeted illegal P2P sharing operations. The copyright battle has been raging in the news media and Google finally took steps that delighted the music and film industry by cracking down on the listings of sites that had multiple copyright violations. These sites would be penalized by Google not only in indexing, but also through DMCA “take-down” requests. DMCA stands for Digital Millennium Copyright Act, and was implemented to protect copyrighted material circulating on the web. Since Napster, copyright infringements have been a hot-button topic in the media and entertainment industries and this was Google’s first major step to help protect copyrights and punish web pirates.
Due to the immense amount of Panda updates that were taking place the naming switched to a numeric value as opposed to names like Panda 2.1. This was one of the more major Panda updates, and not just a refresh. The new update was an actually algorithm change that effected approximately 2.4% of queries in English. It rolled out amidst an exact-match-domain update that was targeting low-quality sites ranking based upon exact matches of keywords that were within domain names.
Penguin 2.0 had an impact on a similar scale of Panda 20 but focused again on web-spam. It was originally thought to have had a massive impact on the industry but ended up having only a moderate one.
New News Results
Google added an additional type of search results catered to in-depth articles on any given subject. The addition was in response to research that pointed to a need for a deeper understanding of topics. Your typical Google user is only looking for a quick answer, but there is a small percentage of users out there looking for further exploration and education on certain subject. The in-depth article feature was designed to sate that need.
Hummingbird is a roll-out that has been likened to Caffeine and is thought to aid with the Knowledge Graph and change semantic search. Knowledge Graph was an update designed to improve the specificity of search results by understanding the linguistics of queries put into the search box. It was a step to better analyze meanings in context so as to decipher the difference between words with multiple meanings and which was the intended use. Knowledge Graph, in particular, provides structured and in-depth information about the topic searched so as to cut down on the amount of user effort needed to surf multiple sites to derive the information they need.
Panda 4.0 was both a data refresh and algorithm update that affected approximately 7.5% of sites in English. Yet another update on attempting to limit low-quality sites from indexing well.
Pigeon was an update aimed at local results and shook up ranking on a local level. It was designed to take into higher account where the web signal of the searcher was coming from so as to deliver geographically accurate search responses.
HTTPS Security Update
With online security raging in the news through 2014 Google made a move to give sites with higher levels of security preferential treatment. Originally, the advantage gained by secure sites was to be very small. The update put a definite stress on the importance of site and web security.
Another Panda update. Its goal was to increase the ranking of smaller and medium sized sites with high quality content. It also fine tuned the identification of low quality sites.
In the most recent Pirate update Google re-targeted file-sharing sites, particularly those with torrent downloading. The update had a major affect on search queries like “free download”. Some file sharing sites were majorly affected as the war on copyright infringement continues.
Well there you have it! As I outlined, this is NOT a complete history. Google makes so many changes to its algorithms and platforms that you could write a book on the entire history of algorithm updates. I’ve been thorough and included many of them that had significant impacts or affected the future of Google search the way you see it today. The overall arc of Google algorithm changes have been designed to improve the quality of search results and eliminate web spam. Almost every change Google has made has an end game of the best user experience. Best of luck keeping up with the updates as they roll in, because they are not going to stop now.