The Google algorithm is constantly changing.In 2018 alone, Google ran 15,096 Live traffic experiments, and launched 3,234 updates to its search algorithm.
Before we get into the timeline of individual google updates, it’s going to be helpful to define a handful of things upfront for any SEO newbies out there:
SEO experts, writers, and audiences will often refer to “Google’s Core Algorithm” as though it is a single item. In reality, Google’s Core Algorithm is made up of millions of smaller algorithms that all work together to surface the best possible search results to users. What we mean when we say “Google’s Core Algorithm” is the set of algorithms that are applied to every single search, which are no longer considered experimental, and which are stable enough to run consistently without requiring significant changes.
The Panda algorithm focused on removing low quality content from search by reviewing on-page content itself. This algorithm focused on thin content, content dominated by ads, poor quality content (spelling/grammar mistakes), and rewarded unique content. Google Panda was updated 29 times before finally being incorporated into the core algorithm in January of 2016.
The Penguin algorithm focused on removing sites engaging in spammy tactics from the search results. Penguin primarily filtered sites engaging in keyword stuffing and link schemes out of the search results. Google Penguin was updated 10 times before being integrated into Google’s core algorithm in September of 2016.
This machine-learning based AI helps Google process and understand the meaning behind new search queries. RankBrain works by being able to infer the meaning of new words or terms based on context and related terms. RankBrain began rolling out across all of Google search in early 2015 and was fully live and global by mid-2016. Within three months of full deployment RankBrain was already the 3rd most important signal contributing to the results selected for a search query.
One of the first 100 employees at Google, Matt Cutts was the head of Google’s Web Spam team for many many years, and interacted heavily with the webmaster community. He spent a lot of time answering questions about algorithm changes and providing webmasters high-level advice and direction.
Originally a Founding Editor, Advisor, and Writer for Search Engine Land (among others), Danny Sullivan now communicates with the SEO community as Google’s Public Search Liaison. Mr. Sullivan frequently finds himself reminding the community that the best way to rank is to create quality content that provides value to users.
Google Webmaster Trends Analyst who often responds to the SEO community when they have questions about Google algorithm updates and changes. Gary is known for his candid (and entertaining) responses, which usually have a heavy element of sarcasm.
Frequently referenced whenever people speak about Google algorithm updates, webmasterworld.com is one of the most popular forums for webmasters to discuss changes to Google’s search results. A popular community since the early 2000’s webmasters still flock to the space whenever major fluctuations are noticed to discuss theories.
In November of 2019 Google rolled out an update to how local search results are formulated (ex: map pack results). This update improved Google’s understanding of the context of a search, by improving its understanding of synonyms. In essence, local businesses may find they are showing up in more searches.
In early November, we began making use of neural matching as part of the process of generating local search results. Neural matching allows us to better understand how words are related to concepts, as explained more here: https://t.co/ShQm7g9CvN
— Google SearchLiaison (@searchliaison) December 2, 2019
In October Google introduced BERT a deep-learning algorithm focused on helping Google understand the intent behind search queries. BERT (Bidirectional Encoder Representations from Transformers) gives context to each word within a search query. The “bidirectional” in BERT refers to how the algorithm looks at the words that come before and after each term before assessing the meaning of the term itself.
Here’s an example of bi-directional context from Google’s Blog:
In the sentence “I accessed the bank account,” a unidirectional contextual model would represent “bank” based on “I accessed the” but not “account.” However, BERT represents “bank” using both its previous and next context — “I accessed the… account” — starting from the very bottom of a deep neural network, making it deeply bidirectional.
The introduction of BERT marked the most significant change to Google search in half a decade, impacting 1 in 10 searches — 10% of all search queries.
Additional Reading:
If you place reviews on your own site (even through a third party widget), and use schema markup on those reviews – the review stars will no longer show up in the Google results. Google applied this change to entities considered to be Local Businesses or Organizations.
The reasoning? Google considers these types of reviews to be self-serving. The logic is that if a site is placing a third party review widget on their own domain, they probably have some control over the reviews or review process.
Our recommendation? If you’re a local business or organization, claim your Google My Business listing and focus on encouraging users to leave reviews with Google directly.
Additional Reading:
This update included two components:First, it hit sites exploiting a 301 redirect trick from expired sites. In this trick users would buy either expired sites with good SEO metrics and redirect the entire domain to their site, or users would pay a 3rd party to redirect a portion of pages from an expired site to their domain.Note: Sites with relevant 301 redirects from expired sites were still fine.
Second, video content appears to have gotten a boost from this update. June’s update brought an increase in video carousels in the SERPs. Now in September, we’re seeing video content bumping down organic pages that previously ranked above them.
Additional Reading:
This is the first time that Google has pre-announced an update. Danny Sullivan, Google’s Search Liaison, stated that they chose to pre-announce the changes so webmasters would not be left “scratching their heads” about what was happening this time.
What happened?
What can sites do to respond to this broad core update? It looks like Google is leaning into video content, at least in the short-term. Consider including video as one of the types of content your team creates.
Additional Reading:
On Wednesday May 22nd Google tweeted that there were indexation bugs causing stale results to be served for certain queries, this bug was resolved early on Thursday May 23rd.
By the evening of Thursday May 23rd Google was back to tweeting – stating that they were working on a new indexing bug that was preventing capture of new pages. On May 26th Google followed up that this indexation bug had also been fixed.
Additional Reading:
In April of 2019 an indexing bug caused about 4% of stable URLs to fall off of the first page. What happened? A technical error caused a bug to de-index a massive set of webpages.
Additional Reading:
Google was specifically vague about this update, and just kept redirecting people and questions to the Google quality guidelines. However, the webmaster community noticed that the update seemed to have a heavier impact on YMYL (your money or your life) pages.
YMYL sites with low quality content took a nose-dive, and sites with heavy trust signals (well known brands, known authorities on multiple topics, etc) climbed the rankings.
Let’s take two examples:
First, Everdayhealth.com lost 50% of their SEO visibility from this update. Sample headline:Can Himalayan Salt Lamps Really Help People with Asthma?
Next, Medicinenet.com saw a 12% increase in their SEO visibility from this update. Sample headline: 4 Deaths, 141 Legionnaires’ Infections Linked to Hot Tubs.
This update also seemed to factor in user behavior more strongly. Domains where users spent longer on the site, had more pages per visit, and had lower bounce rates saw an uptick in their rankings.
Additional Reading:
For one day, on March 1st, Google displayed 19 results on the first page of SERPs for all queries, 20 if you count the featured snippet. Many hypothesize it was a glitch related to in-depth articles, a results type from 2013 that has long since been integrated into regular organic search results.
Additional Reading:
This broad core update, known by its nickname “Medic” impacted YMYL (your money or your life) sites across the web.
SEOs had many theories about what to do to improve rankings after this update, but both Google and the larger SEO community ended up at the same messaging: make content user’s are looking for, and make it helpful.
This update sparked a lot of discussion around E-A-T (Expertise, Authoritativeness, Trustworthiness) for page quality, and the importance of clear authorship and bylines on content.
Additional Reading:
Google begins marking all http sites as “not secure” and displaying warnings to users.
Looking forward, Google is planning on blocking mixed content from https sites.
What can you do? Purchase an SSL certificate and make the move from http to https as soon as possible. Double check that all of your subdomains, images, PDFs and other assets associated with your site are also being served securely.
Additional Reading:
Google rolled out the mobile page speed update, making page speed a ranking factor for mobile results.
Additional Reading:
Google introduces a dedicated video carousel on the first page of results for some queries, and moves videos out of regular results. This change also led to a significant increase in the number of search results displaying videos (+60%).
Additional Reading:
The official line from Google about this broad core update, is that it rewards quality content that was previously under-rewarded. Sites that had content that was clearly better than the content of it’s organic competitors saw a boost, sites with thin or duplicative content fell.
March’s update focused on content relevance (how well does content match the intent of the searcher) rather than content quality.
What can you do? Take a look at the pages google is listing in the top 10-20 spots for your target search term and see if you can spot any similarities that hint at how Google views the intent of the search.
Additional Reading:
After months of testing Google begins rolling out mobile-first indexing. Under this approach, Google crawls and indexes the mobile version of website pages when adding them to their index. If content is missing from mobile versions of your webpages, that content may not be indexed by Google.
To quote Google themselves,
“Mobile-first indexing means that we’ll use the mobile version of the page for indexing and ranking, to better help our – primarily mobile – users find what they’re looking for.”
Essentially the entire index is going mobile-first. This process of migrating over to indexing the mobile version of websites is still underway. Website’s are being notified in Search Console when they’ve been migrated under Google’s mobile-first index.
Google states that a series of minor improvements are rolled out across December. Webmasters and SEO professionals see large fluctuations in the SERPs.
What were the Maccabees changes?
Webmasters noted that doorway pages took a hit. Doorway pages act as landing pages for users, but don’t contain the real content – users have to get past these initial landing pages to access content of any value. Google considers these pages barriers to a user.
A writer at Moz dissected a slew of site data from mid-december noted one key observation. When two pages ranked for the same term, the one with better user engagement saw it’s rankings improve after this update. The other page saw its rankings drop. In many instances what happened for sites that began to lose traffic, is that blog pages were being shown/ranked where product or service pages should have been displayed.
A number of official celebrity sites fall in the rankings including (notably) Channing Tatum, Charlie Sheen, Kristen Stewart, Tom Cruise, and even Barack Obama. This speaks to how Google might have rebalanced factors around authoritativeness vs. content quality. One SEO expert noted that thin celebrity sites fell while more robust celebrity sites (like Katy Perry’s) maintained their #1 position.
Multiple webmasters reporting a slew of manual actions on December 25th and 26th, and some webmasters also reported seeing jumps on the 26th for pages that had been working on site quality.
Additional Reading:
Google increases the character length of meta descriptions to 300 characters. This update was not long-lived as Google rolled back to the original 150-160 character meta descriptions on May 13, 2018.
Webmasters noted that this update targeted sites and pages with:
Additional Reading:
In early March webmasters and SEOs began to notice significant fluctuations in the SERPs, and Barry Schwartz from SEJ began tweeting Google to confirm algorithm changes.
The changes seemed to target content sites engaging in aggressive monetization at the expense of users. Basically sites filling the internet up with low-value content, meant to benefit everyone except the user. This included PBN sites, and sites created with the sole intent of generating AdSense income.
Fred got its name from Gary Illyes who suggested to an SEO expert asking if he wanted to name the update, that we should start calling all updates without names “Fred.”
Fred it is https://t.co/aEE7sfm7X5 @i_praveensharma @JohnMu
— Barry Schwartz (@rustybrick) March 9, 2017
The joke, for anyone who knows the webmaster trends analyst, is that he calls everything unnamed fred (fish, people, EVERYTHING).
What those who aren’t friends with @methode don’t realize is that he calls EVERYTHING un-named, fred. Fish, people, whatever.
— Ryan Jones (@RyanJones) March 9, 2017
The SEO community took this as a confirmation of recent algorithm changes (note: literally every day has algorithm updates). Validating them digging into the SERP Changes.
Additional Reading:
Google announces that intrusive pop ups and interstitials are going to be factored into their search algorithm moving forward.
“To improve the mobile search experience, after January 10, 2017, pages where content is not easily accessible to a user on the transition from the mobile search results may not rank as highly.”
This change caused rankings to drop for sites that forced users to get past an ad or pop up to access relevant content. Not all pop ups or interstitials were penalized, for instance the following pop ups were still okay:
Additional Reading:
The Google announcement of Penguin 4.0 had two major components:
SEOs also noted one additional change. Penguin 4.0 seemed to just remove the impact of spam links on SERPs, rather than penalizing sites with spammy links. This appeared to be an attempt for Google to mitigate the impact of negative SEO attacks on sites.
That being said, today in 2019 we still see a positive impact from running disavows for clients who have seen spammy links creep into their backlink profiles.
Additional Reading:
This update targeted duplicate and spammy results in local search (Local Pack and Google Maps). The goal being to provide more diverse results when they’re searching for a local business, product, or service.
Prior to the Possum update Google was filtering out duplicates in local results by looking for listings with matching domains or matching phone numbers. After the Possum update Google began filtering out duplicates based on their physical address.
Businesses who saw some of their listings removed from the local pack may have initially thought their sites were dead (killed by this update), but they weren’t – they were just being filtered (playing possum). The term was coined by Phil Rozek
SEOs also noted that businesses right outside of city limits also saw a major uptick in local rankings, as they got included in local searches for those cities.
Additional Reading:
Google boosts the effect of the mobile-friendly ranking signal in search.
Google took time to stress that sites which are not mobile friendly but which still provide high quality content will still rank.
Additional Reading:
Google Removes sidebar ads and ads a fourth ad to the top block above the organic search results.
This move reflects the search engine giant continuing to prioritize mobile-first experiences, where side-bar ads are cumbersome compared to results in the main content block.
Google Confirms core algorithm update in January, right after confirming that Panda is now part of Google’s core algorithm.
Not a lot of conclusions were able to be drawn about the update, but SEOs noticed significant fluctuations with news sites/news publishers. Longform content with multi-media got a boost, and older articles took a bit of a dive for branded terms. This shift could reflect Google tweaking current-event related results to show more recent content, but the data was not definitive.
Additional Reading:
Google starts indexing the https version of pages by default.
Pages using SSL are also seeing a slight boost. Google holds security as a core component of surfacing search results to users, and this shift becomes one of many security-related search algo changes. In fact, by the end of 2017 over 75% of the page one organic search results were https.
in testing since April 2015, Google officially introduced RankBrain on this date. RankBrain is a machine learning algorithm that filters search results to help give users a best answer to their query. Initially, RankBrain was used for about 15 percent of queries (mainly new queries Google had never seen before), but now it is involved in almost every query entered into Google. RankBrain has been called the third most important ranking signal.
Additional Reading:
Google introduces an algorithm specifically targeting spammy in the search results that were gaining search equity from hacked sites.
This change was significant, it impacted 5% of search queries. This algorithm hides sites benefiting from hacked sites in the search results.
@rustybrick yesr. That’s what we’re aiming for.
— Gary “鯨理” Illyes (@methode) October 9, 2015
The update came right after a September message from Google about cracking down on repeat spam offenders. Google’s blog post notified SEOs that sites which repeatedly received manual actions would find it harder and harder to have those manual actions reconsidered.
Additional Reading:
Google switches from displaying seven results for local search in the map pack to only three.
Why the change? Google is switching over (step-by-step) to mobile-first search results, aka prioritizing mobile users over desktop users.
On mobile, only three local results fit onto the screen before a users needs to scroll. Google seems to want users to scroll to then access organic results.
Other noticeable changes from this update:
Additional Reading:
Roll out of Panda 4.2 began on the weekend of July 18th and affected 2-3% of search queries. This was a refresh, and the first one for Panda in about 9 months.
Why does that matter? The Panda algorithm acts like a filter on search results to sort out low quality content. Panda basically gets applied to a set of data – and decides what to filter out (or down). Until the data for a site is refreshed, Panda’s ruling is static. So when a data refresh is completed, sites that have made improvements essentially get a revised ruling on how they’re filtered.
Nine months is a long time to wait for a revised ruling!
This change is an update to the quality filters integrated into Google’s core algorithm, and alters how the algorithm processes signals for content quality. This algorithm is real-time, meaning that webmasters will not need to wait for data refreshes to see positive impact from making content improvements.
What kind of pages did we see drop in the rankings?
In hindsight, this update feels like a precursor to Google’s 2017 updates for content spam and intrusive pop ups.
Additional Reading:
Google boosts mobile-friendly pages in mobile search results.
This update was termed Mobilegeddon as SEOs expected it to impact a huge number of search queries, maybe more than any other update ever had. Why? Google was already seeing more searches on mobile than on desktop in the U.S. in May 2015.
In 2018 Google takes this a step further and starts mobile-first indexing.
Additional Reading:
Google’s local algorithm, known as Pigeon, expands to international English speaking countries (UK, Canada, Australia) on December 22, 2014.
In December Google also releases updated guidelines for local businesses representing themselves on Google.
Additional Reading:
Google releases an “improved DMCA demotion signal in Search,” specifically designed to target and downrank some of the sites most notorious for piracy.
In October Google also released an updated report on how they fight piracy, which includes changes they made to how results for media searches were displayed in search. Most of these user interface changes were geared towards helping user find legal (trusted) ways to consume the media content they were seeking.
This update impacted 1% of English search queries, and was the first update to Penguin’s algorithm in over a year. This update was both a refresh and a major algorithm update.
Panda 4.1 is the 28th update for the algorithm that targets poor quality content. This update impacted 3-5% of search queries.
To quote Google:
“Based on user (and webmaster!) feedback, we’ve been able to discover a few more signal to help Panda identify low-quality content more precisely. This results in a greater diversity of high-quality small- and medium-sized sites ranking higher, which is nice.”
Major losers were sites with deceptive ads, affiliate sites (thin on content, meant to pass traffic to other monetizing affiliates), and sites with security issues.
This change impacted search, but was not an algorithm change, data refresh, or UI update.
Starting mid-to-late September, 2014 Google de-indexed a massive amount of sites being used to boost other sites and game Google’s search rankings.
Google then followed-up on the de-indexing with manual actions for sites benefiting from the PBN. These manual actions went out on September 18, 2014.
Additional Reading:
Authors are no longer displayed (name or photo) in the search results along with the pieces that they’ve written.
Almost a year later Gary Illyes suggested that sites with authorship markup should leave the markup in place because it might be used again in the future. However, at a later date it was suggested that Google is perfectly capable of recognizing authorship from bylines.
Additional Reading:
Sites using SSL began to see a slight boost in rankings.
Google would later go on to increase this boost, and eventually provide warning to users when they were trying to access unsecure pages.
Additional Reading:
Google’s local search algorithm is updated to include more signals from traditional search (knowledge graph, spelling correction, synonyms, etc).
Additional Reading
Photos of Authors are gone from SERPs.
This was the first step towards Google decommissioning Authorship markup.
Where Payday Loans 2.0 targeted spammy sites, Payday Loans 3.0 targeted spammy queries, or more specifically the types of illegal link schemes scene disproportionately within high-spam industries (payday loans, porn, gambling, etc).
What do you mean illegal? We mean link schemes that function off of hacking other websites or infecting them with malware.
This update also included better protection against negative SEO attacks,
Additional Reading:
Payday Loan Update 2.0 was a comprehensive update to the algorithm (not just da data refresh). This update focused on devaluation of domains using spamy on-site tactics such as cloaking.
Cloaking is when the content/page that google can see for a page is different than the content/page that a human user sees when they click on that page from the SERPs.
Google had stopped announcing changes to Panda for a while, so when they announced Panda 4.0 we know it was going to be a larger change to the overall algorithm.
Panda 4.0 impacted 7.5% of English queries, and led to a drastic nose dive for a slew of prominent sites like eBay, Ask.com, and Biography.com.
@mattcutts I hope it is not too painful 😉 pic.twitter.com/VUmBJ59IEG
— Steven Broschart (@optimizingexp) May 21, 2014
This is a refresh of Google’s algorithm that devalues pages with too many above-the-fold ads, per Google’s blog:
We’ve heard complaints from users that if they click on a result and it’s difficult to find the actual content, they aren’t happy with the experience. Rather than scrolling down the page past a slew of ads, users want to see content right away.
So sites that don’t have much content “above-the-fold” can be affected by this change. If you click on a website and the part of the website you see first either doesn’t have a lot of visible content above-the-fold or dedicates a large fraction of the site’s initial screen real estate to ads, that’s not a very good user experience.
The Page Layout algorithm was originally launched on January 19, 2012, and has only had one other update in October of the same year (2012).
Authorship gets less of a boost in the search results. This is the first step Google took in beginning to phase out authorship markup.
Technically the 5th update to Google’s link-spam fighting algorithm, this minor update affects about 1% of search queries.
Penguin 2.1 launching today. Affects ~1% of searches to a noticeable degree. More info on Penguin: http://t.co/4YSh4sfZQj
— Matt Cutts (@mattcutts) October 4, 2013
Hummingbird was a full replacement of the core search algorithm, and Google’s largest update since Caffeine (Panda and Penguin had only been changes to portions of the old algorithm).
Humminbird helped most with conversational search for results outside of the knowledge graph — where conversational search was already running. Hummingbird was a significant improvement to how google interpreted the way text and queries are typed into search.
This algorithm was named Hummingbird by Google because it’s “precise and fast.”
Knowledge Graph Expands to nearly 25% of all searches, displaying information-rich cards right above or next to the organic search results.
Additional Reading:
Panda begins going through monthly refreshes, also known as the “Panda Dance,” which caused monthly shifts in search rankings.
The next time Google would acknowledge a formal Panda update outside of these refreshes would be almost a year later in May of 2014.
Google rolled out an anti-link-spam algorithm in June of 2013 targeting sites grossly violating webmaster guidelines with egregious unnatural link building.
Matt Cutts even acknowledged one target – ‘Citadel Insurance’ which built 28,000 links from 1,000 low ranking domains within a single day, June 14th, and managed to reach position #2 for car insurance with the tactic.
By the end of June sites were finding it much harder to exploit the system with similar tactics.
@screamingfrog very natural…. @mattcutts
— Jamie Knop (@JamieKnop) June 21, 2013
@JamieKnop @screamingfrog yup, we saw that. Multi-week rollout going on now, from next week all all the way to the week after July 4th.
— Matt Cutts (@mattcutts) June 22, 2013
This update impacted 0.3% of queries in the U.S., and as much as 4% of queries in Turkey.
This algorithm targets queries that have abnormally high incidents of SEO spam (payday loans, adult searches, drugs, pharmaceuticals) and applies an extra filters to these types of queries specifically.
We just started a new ranking update today for some spammy queries. See 2:30-3:10 of this video: http://t.co/KOMozoc3lo #smx
— Matt Cutts (@mattcutts) June 11, 2013
Penguin 2.0 was an update to the Penguin algorithm (as opposed to just a data refresh), it impacted 2.3% of english queries.
What changed?
One of the biggest shifts with Penguin 2.0 is it also analyzed linkspam for internal site pages, whereas Penguin 1.0 had looked at spammy links specifically pointing to domain home pages.
This marked the first time in 6 months that the Penguin algorithm had been updated, and the 4th update to Penguin that we’ve seen:
This update reduced the amount of times a user saw the same domain in the search results. According to Matt Cutts, once you’ve seen a cluster of +/- 4 results from the same domain, the subsequent search pages are going to be significantly less likely to show you results from that domain.
On May 8th, 2013 SEOs over at Webmaster World noticed intense fluctuation in the SERPs.
Lots of people dove into the data – some commenting that sites who had taken a dive were previously hit by Panda, but there were no conclusive takeaways. With no confirmation of major changes from Google, and nothing conclusive in the data – this anomaly came to be known as the “Phantom” update.
This is the 25th update for Panda, the algorithm that devalues low quality content in the SERPs. Matt Cutts confirmed that moving forward the Panda algorithm was going to be part of a regular algorithm updates, meaning it will be a rolling update instead of a pushed update process.
The 24th Panda update was announced on January 22, 2013 and impacted 1.2% of English search queries.
New Panda data refresh rolling out today: 1.2% of English queries affected. Background: http://t.co/EMBEJqGZ
— Google (@Google) January 22, 2013
The 23rd Panda update hit on December 21, 2012 and impacted 1.3% of English search queries.
On December 4, 2012 Google announced a foriegn language expansion of the Knowledge Graph, their project to “map out real-world things as diverse as movies, bridgets and planets.”
In November 2012 Panda had two updates in the same month – one on November 5, 2012 (1.1% of English queries impacted in the US) and one on November 22, 2012 (0.8% of Enlish queries impacted in the US).
On October 9, 2012 Google rolled up an update to their Page Layout filter (also known as “Top Heavy”) impacting 0.7% of English-language search queries. This update rolled the Page Layout algorithm out globally.
Sites that made fixes after Google’s initial Page Layout Filter hit back in January of 2012 saw their rankings recover in the SERPs.
This was just a data refresh affecting 0.3% of English queries in the US.
Weather report: Penguin data refresh coming today. 0.3% of English queries noticeably affected. Details: http://t.co/Esbi2ilX
— Matt Cutts (@mattcutts) October 5, 2012
Panda update 19 hit on September 18, 2012 affecting 0.7% of English search queries, followed just over a week later by Panda update 20 which hit on September 27, 2012 affecting 2.4% of English search queries.
Panda update 20 was an actual algorithm refresh, accounting for the higher percentage of affected queries.
At the end of September Matt Cutts announced an upcoming change: low quality exact match domains were going to be taking a hit in the search results.
Up until this point, exact match domains had been weighted heavily enough in the algorithms to counterbalance low quality site content.
Additional Reading:
Panda version 3.9.1 rolled out on Monday, August 19th, 2012, affecting less than 1% of English search queries in the US.
This update was a data refresh.
Panda data refresh this past Monday. ~1% of queries noticeably affected. More context: http://t.co/nSjVRWGb
— Google (@Google) August 22, 2012
In August Google began displaying 7 results for about 18% of the queries, rather than the standard 10.
Upon further inspection it appeared that google had reduced the number of organic results so they’d have more space to test a suite of potential top-of search features including: expanded site links, images, and local results.
This change, in conjunction with the knowledge graph, paved the way for the top-of-search rich snippet results we see in search today.
Additional Reading:
Google announces they’ll be devaluing sites that repeatedly get accused of copyright infringement in the SERPs. As of this date the number of valid copyright removal notices is a ranking signal in Google’s search algorithm.
Additional Reading:
On July 24, 2012 Google Announces Panda 3.9.0 – a refresh for the algorithm affecting less than 1% search
Not technically an algorithm update, but it definitely affected the SEO landscape.
On July 27, 2012 Google posted an update clarifying topics surrounding a slew of unnatural link warnings that had recently been sent out to webmasters:
Additional Reading:
In June Google made two updates to its Panda algorithm fighting low quality content in the SERPs:
Both updates were data refreshes.
On June 7, 2012 Google posted an update providing insight into search changes made over the course of May. Highlights included:
Additional Reading:
A data refresh for the Penguin algorithm was released on May 25, 2012 affecting less than 0.1% of search queries.
Additional Reading:
On May 16, 2012 Google introduced the knowledge graph, a huge step forward in helping users complete their goals faster.
First, the knowledge graph improved Google’s understanding of entities in Search (what words represented — people, places, or things).
Second, it surfaced relevant information about these entities directly on the search results page as summaries and answers. This meant that users in many instances, no longer needed to click into a search result to find the information they were seeking.
Additional Resources:
On May 4, 2012 Google posted an update providing insight into search changes made over the course of April. Highlights included:
Additional Reading:
In April Google made two updates to its Panda algorithm fighting low quality content in the SERPs:
Panda 3.5 seemed to target press portals and aggregators, as well as heavily-templated websites. This makes sense as these types of sites are likely to have a high number of pages with thin or duplicative content.
Additional Reading:
The Penguin Algorithm was announced on April 24, 2012 and focused specifically on devaluing sites that engage in spammy SEO practices.
The two primary targets of Penguin 1.0? Keyword stuffing and link schemes.
Additional Reading:
The Penguin Algorithm was announced on April 24, 2012 and focused specifically on devaluing sites that engage in spammy SEO practices.
The two primary targets of Penguin 1.0? Keyword stuffing and link schemes.
Additional Reading:
After a number of webmasters reported ranking shuffles, Google confirmed that a data error had caused some domains to be mistakenly treated as parked domains (and thereby devalued). This was not an intentional algorithm change.
Additional Reading:
On April 3, 2012 Google posted an update providing insight into search changes made over the course of March. Highlights included:
Additional Reading:
On March 23, 2012 we saw the Penguin 3.4 update, a data refresh affecting 1.6% of queries.
Panda refresh rolling out now. Only ~1.6% of queries noticeably affected. Background on Panda: http://t.co/Z7dDS6qc
— Google (@Google) March 23, 2012
Panda Update 3.3 was a data refresh that was announced on February 27, 2012.
On February 27, 2012 Google posted an update providing insight into search changes made over the course of February. Highlights included:
Additional Reading:
The Venice update changed the face of local search forever, as local sites now up even without a geo modifier being used in the keyword itself.
Additional Reading:
This update devalued pages in search that had too many ads “above-the-fold.” Google said that ads that prevented users from accessing content quickly provided a poor user experience.
Additional Reading:
On January 10, 2012 Google announced Search, plus Your World. Google had already expanded search to include content personally relevant to individuals with Social Search, Your World was the next step.
This update pulled in information from Google+ such as photos, profiles, and more.
Additional Reading:
On January 5, 2012 Google posted an update providing insight into search changes made over the course of December of 2011. Highlights included:
Additional Reading:
On December 1, 2011 Google posted an update providing insight into search changes made the two weeks prior. Highlights included:
Additional Reading:
The Panda 3.1 update rolled out on November 18th, 2011 and affected less than 1% of searches.
On November 18th, 2011 Panda Update 3.1 goes live, impacting <1% of searches.
On November 14, 2011 Google posted an update providing insight into search changes made over the couple preceding weeks. Highlights included:
Additional Reading:
Google puts an emphasis on more recent results, especially on time-sensitive queries.
Additional Reading:
On October 18, 2011 Google announced that they were going to be encrypting search data for users who are signed in.
The result? Webmasters could tell that users were coming from google search, but could no longer see the queries being used. Instead, webmasters began to see “(not provided)” showing up in their search results.
This change followed a January roll out of SSL encryption protocol to gmail users.
Additional Reading:
In October Matt Cutts announced there would be upcoming flux from the Panda 3.0 update affecting about 2% of search queries. Flux occurred throughout October as new signals were incorporated into the Panda algorithms and data is refreshed.
Additional Reading:
On September 20, 2011 Google released their 7th update to the Panda algorithm – Panda 2.5.
Google added pagination elements – link attributes to help with pagination crawl/indexing issues.
Note: this is no longer an indexing signal anymore
Additional Reading:
On August 16, 2011 Google announced expanded display of sitelinks from a max of 8 links to a max of 12 links.
Google rolled out Panda 2.4 expanding Panda to more languages August 12, 2011, impacting 6-9% of queries worldwide.
Additional Reading:
Google rolled out Panda 2.3 in July of 2011, adding new signals to help differentiate between higher and lower quality sites.
On June 28, 2011 Google launched their own social network, Google+. The network was sort of a middle ground between Linkedin and Facebook.
Over time, Google + shares and +1s (likes) will eventually become a temporary personalized search ranking factor.
Ultimately though, Google+ ended up being decommissioned in 2019
Additional Reading:
According to Matt Cutts Panda 2.2 improved scraper-site detection.
What’s a scraper? In this context, a scraper is software used to copy content from a website, often to be posted to another website for ranking purposes. This is considered a type of webspam (not to mention plagiarism).
This update rolled out around June 16, 2011.
On June 2, 2011 Google, Yahoo, and Microsoft announced a collaboration to create “a common vocabulary for structured data,” known as Schema.org.
Additional Reading:
Panda 2.1 rolled out in early May, and was relatively minor compared to previous Panda updates.
On April 11, 2011 Panda 2.0 rolled out globally to English users, impacting about 2% of search queries.
What was different in Panda 2.0?
Additional Reading:
Google introduces the +1 Button, similar to facebook “like” button or the reddit upvote. The goal? Bring trusted content to the top of the search results.
Later in June Google posted a brief update that they made the button faster, and in August of 2011 it also became a share icon.
Panda was released to fight thin content and low-quality content in the SERPs. Panda was also designed to reward unique content that provides value to users.
As a result sites with less intrusive ads started to do better in the search results, sites with”thin” user-generated content went down, as did harder to read pages.
Per Google:
As “pure webspam” has decreased over time, attention has shifted instead to “content farms,” which are sites with shallow or low-quality content.”
Additional Reading
This update focused on stopping scraper sites from receiving benefit from stolen content. The algorithm worked to establish which site initially created and posted content, and boost that site in the SERPs over other sites which had stolen the content.
Additional Reading:
Overstock and J.C. Penney receive manual actions due to deceptive link building practices.
Overstock offered a 10% discount to universities, students, and parents — as long as they posted anchor-text rich content to their university website. A competitor noticed the trend and reported them to Google.
JC Penney had thousands of backlinks built to its site targeting exact match anchor text. After receiving a manual action they disavowed the spammy links and largely recovered.
Additional Reading:
Google confirms that they use social signals including accounting for shares when looking at news stories, and author quality.
<h3style=”font-size: 18pt;”>2010 December – Negative ReviewsIn late November a story broke about how businesses were soaring in the search results, and seeing their businesses grow exponentially – by being as terrible to customers as possible.
Enraged customers were leaving negative reviews on every major site they could linking back to these bad-actor businesses, trying to warn others. But what was happening in search, is all those backlinks were giving the bad actors more and more search equity — enabling them to show up as the first result for a wider and wider range of searches.
Google responded to the issue within weeks, making changes to ensure businesses could not abuse their users in that manner moving forward.
Per Google:
“Being bad is […] bad for business in Google’s search results.”
Additional Reading:
NYT – Bullies Rewarded in Search
This temporary feature allowed users to see a visual preview of a website in the search results. It was quickly rolled back.
Google suggest starts displaying results before a user actually completes their query.
This feature lived for a long time (in tech-years anyways) but was sunset in 2017 as mobile search became dominant, and Google realized it might not be the optimal experience for on-the-go mobile users.
Google made a change to allow some brands/domains to appear multiple times on page one depending on the search
This feature ends up undergoing a number of updates over time as Google works to get the right balance of site diversity when encountering host-clusters (multiple results from the same domain in search).
On June 10, 2010 Google announced Caffeine.
Caffeine was an entirely new indexing system with a new search index. Where before there had been multiple indexes, each being updated and refreshed at their own rates, caffeine enabled continuous updating of small portions of the search index. Under caffeine, newly indexed content was available within seconds of being crawled
Per Google:
“Caffeine provides 50 percent fresher results for web searches than our last index, and it’s the largest collection of web content we’ve offered. Whether it’s a news story, a blog or a forum post, you can now find links to relevant content much sooner after it is published than was possible ever before.”
Additional Reading:
The May Day update occurred between April 28th and May 3rd 2010. This update was a precursor to Panda and took a shot at combating content farms.
Google’s comment on the update? “If you’re impacted, assess your site for quality.”
Additional Resources:
In April of 2010 Local Business Center became Google Places. Along with this change came the introduction of service areas (as opposed to just a single address as a location).
Other highlights:
By April of 2010, 20% of searches were already location-based.
Additional Reading:
Google announces search features related to newly indexed content: Twitter Feeds, News Results, etc. This real time feed was nested under a “latest results” section of the first page of search results.
On August 10 Google begins to preview Caffeine, requesting feedback from users.
Additional Reading:
Essentially the Vince update boosted brands.
Vince focused on trust, authority and reputation as signals to provide higher quality results which could push big brands further to the top of the SERPs.
Additional Resources:
Watch – Is Google putting more weight on brands in rankings?
Read – SEO Book – Google Branding
Google introduces “suggest” which displays suggested search terms as the user is typing their query.
Additional Reading:
The Dewey update rolled out in late March/early April. The update was called Dewey because Matt Cutts chose the (slightly unique) term as one that would allow comparison between results from different data centers.
The Buffy update caused fluctuations for single-word search results.
Why Buffy?Google Webmaster Central product manager and long-time head of operations, Vanessa Fox, notoriously an avid Buffy fan, announced she was leaving Google.
Vanessa garnered an intense respect from webmasters over her tenure both for her product leadership and for her responsiveness to the community – the people using google’s products daily. The webmaster community named this update after her interest as a sign of respect.
Additional Reading:
Old school organic search results are integrated with video, local, image, news, blog, and book searches.
Additional Reading:
An update to how the filtering of pages stored in the supplemental index is handled. Google went on to scrap the supplemental index label in July 2007.
Additional Reading:
This was an update to the Google search infrastructure and took 3 months to roll out: January, February, and March. This update also changed how google handled canonicalization and redirects.
Additional Reading:
The Jagger Update rolled out as a series of October updates.
The update targeted low quality links, reciprocal links, paid links, and link farms. The update helped prepare the way for the Big Daddy infrastructure update in November.
Additional Reading:
In October of 2015, Google merged Local Business Center data merges with Maps data.
A number of SEOs noted fluctuations in September which they originally named “Gilligan.” It turns out there were no algorithm updates, just a data refresh (index update).
Given the news, many SEOs renamed their posts “False Alarm.” However, moving forward many data refreshes are considered updates by the community. So we’ll let the “Gilligan” update stand.
Additional Reading:
Google relaunches personal search. This time it helps shape future results based on your past selections.
Additional Reading:
Google launches the ability to submit XML sitemaps via Google Webmaster tools. This update bypassed old HTML sitemaps. It gave Webmasters some influence over indexation and crawling, allowing them to feed pages to the index with this feature.
Additional Reading:
The May 2005 update, nicknamed Bourbon seemed to devalue sites/pages with duplicate content, and affected 3.5% of search queries.
The Allegra update rolled out between February 2, 2005 and February 8, 2005. It caused major fluctuations in the SERPs. While nothing has ever been confirmed, these are the most popular theories amongst SEOs for what changed:
Additional Reading:
In early January, 2005 Google introduced the “Nofollow” link attribute to combat spam, and control the outbound link quality. This change helped clean up spammy blog comments: comments mass posted to blogs across the internet with links meant to boost the rankings of the target site.
Future Changes:
Additional Reading:
The Brandy update rolled out the first half of February and included five significant changes to Google’s algorithmic formulas (confirmed by Sergey Brin).
Over this same time period Google’s index was significantly expanded, by over 20%, and dynamic web pages were included in the index.
What else changed?
Austin followed up on Florida continuing to clean up spammy SEO practices, and push unworthy sites out of the first pages of search results.
What changed?
Many SEOs also speculated that this had been a change to Hilltop, a page rank algorithm that had been around since 1998.
Additional Reading:
Google’s Florida update rolled out on November 16, 2003 and targeted spammy seo practices such as keyword stuffing. Many sites that were trying to game the search engine algorithms instead of serve users also fell in the rankings.
Google split their index into main and supplemental. The goal was to increase the number of pages/content that Google could crawl and index. The supplemental index had less restrictions on indexing pages. Pages from the supplemental index would only be shown if there were very few good results from the main index to display for a search.
When the supplemental index was introduced some people viewed being relegated to the supplemental index as a penalty or search results “purgatory”.
Google retired the supplemental index tag in 2007, but has never said that they retired the supplemental index itself. That being said it’s open knowledge that Google maintains multiple indices, so it is within the realm of reason that the supplemental index may still be one of them. While the label dissapeared, many wonder if the supplemental index has continued to exist and morphed into what we see today as “omitted results”Sites found they were able to move from the supplemental index to the main index by acquiring more backlinks.
Additional Reading:
In July, 2003 Google moved away from monthly index updates (often referred to as the google dance) to daily updates in which a portion of the index was updated daily. These regular updates came to be referred to as “everflux.”
Esmerelda was the last giant monthly index update before Google switched over to daily index updates.
Google’s Dominic update focused on battling spammy link practices.
Google’s Cassandra update launched in April of 2003 and targeted spammy SEO practices including hidden text, heavily co-linked domains, and other low-link-quality practices.
Google began allowing banned sites to submit a reconsideration request after manual penalties in April of 2003.
Additional Reading:
Google’s first named update was Boston which rolled out in February of 2003. The Google Boston Update improved algorithms related to analyzing a site’s backlink data.
Google’s first documented search algorithm update happened on September 1, 2002. It was also the kickoff of “Google Dance” – large-scale monthly refreshes of Google’s search index.
SEOs were shocked by the update claiming “PageRank [is] DEAD”, this update was a little imperfect and included issues such as 404 pages showing up on the first page of search.
Additional Reading:
Google launches their search toolbar for browsers. The toolbar highlights search terms within webpage copy, and allowed users to search within websites that didn’t have their own site search.
Additional Reading:
