Google has a new way of getting feedback from users about the quality of its search results. Some users are reportedly seeing an experiment the search engine is running, which lets them compare two results on a page, and choose which deserves to rank higher.
Alex Chitu at Google Operating System shares a screenshot of what this looks like:
For this particular search, I’m actually seeing result 2 rank higher than result 1. It’s hard to say if this is directly related to the experiment.
For a long time, Google has included a link at the bottom of its search results pages to send feedback to the company, but lately they’re trying some different approaches.
Last month, Google announced a new form for gaining feedback on its Search policy.
Shortly after that, Google’s Matt Cutts tweeted out an other form to let people tell Google if there’s a small website they think should be doing better in Google.
If there's a small website that you think should be doing better in Google, tell us more here: https://t.co/s80BibIBhN
— Matt Cutts (@mattcutts) August 28, 2013
Of course Google also has its army of raters.
Image: Google Operating System
Google says on its new “How Search Works” site that it launched 665 “improvements to search” in 2012.
In a graphic looking at data from 2012, Google explains that it had 118,812 “precision evaluations,” which are described as “The first phase is to get feedback from evaluators, people who evaluate search quality based on our guidelines. We show evaluators search results and ask them to rate the usefulness of the results for a given search.”
From there, it had 10,391 side-by-side experiments.
“In a side-by-side experiment, we show evaluators two different sets of search results: one from the old algorithm and one from the new, and we ask them for details about which results they prefer,” Google explains.
You can see Google’s Search Quality Raters Guidelines (at least a “cliffs notes version“) here (pdf).
Google ran 7,018 live traffic experiments in 2012. “If the evaluators’ feedback looks good, we move forward with a ‘live traffic experiment,’” Google explains. “In these experiments, we change search for a small percentage of real Google users and see how it changes the way they interact with the results. We carefully analyze the results to understand whether the change is an improvement to the search results. For example, do searchers click the new first result more often? If so, that’s generally a good sign.”
From there, came the 665 actual launches.
“Finally, our most experienced search engineers carefully review the data from all the different experiments and decide if the change is approved to launch,” says Google. “It sounds like a lot, but the process is well refined, so an engineer can go from idea to live on Google for a percentage of users in 24 hours. Based on all of this experimentation, evaluation and analysis, we launched 665 improvements to search in 2012.”
Typically, Google has been providing monthly lists of “search quality highlights” showing some of the tweaks they’ve made, but they haven’t done it in months. Despite today’s effort in transparency, it remains to be seen whether we’ll see exactly what Google has been up to since October.
Google publishes a list every month to keep us aware of the changes coming to their search quality. April is no different and it’s a massive update containing 53 changes from domain diversity to better HTML5 resource caching.
This month’s update contains one particularly interesting change though. Google will be increasing the base index size by 15 percent. Under the codename “indexing,” Google describes the change as thus:
The base search index is our main index for serving search results and every query that comes into Google is matched against this index. This change increases the number of documents served by that index by 15%. *Note: We’re constantly tuning the size of our different indexes and changes may not always appear in these blog posts.
Related to the increased base index size, Google will also be introducing a new index tier. As you might be aware, Google keeps their index in these “tiers” where documents are indexed at different rates depending on their relevance. The new tier should increase the comprehensiveness of search for all users.
There’s another 51 changes to go and they range from fascinating to interesting. One of particular note is Google’s introduction of instant answers for the MLB season. Just trying typing sf giants score into Google and get the latest updates.
Google tweeted out a link to a video of some footage from one of its search quality meetings. The company says this is part of its efforts to be more transparent how how its search engine works.
@google
A GooglerWatch uncut video from one of our search quality meetings (for the first time ever!) http://t.co/y53oAAx0
30 minutes ago via web · Reply · Retweet · Favorite · powered by @socialditto
This particular video deals with spelling for long queries. The meeting, held in Mountain View, is where the search team decides to launch specific algorithm changes.
Google says in the YouTube description:
“As part of our continued effort to be more transparent about how search works, we’re publishing video footage from our internal weekly search meeting: ‘Quality Launch Review.’ We hold the meeting almost every Thursday to discuss possible algorithmic improvements and make decisions about what to launch. This video is from the meeting that happened on December 1st, 2011, and includes the entire uncut discussion of a real algorithmic improvement to our spell correction system.”
Google has been talking about these transparency efforts for months, and recently started a monthly series of blog posts outlining various algorithmic changes they’ve made. From the sound of it, we can expect more of these videos as well, which should give us an even deeper look at Google’s strategies and mindset.
If your’e a webmaster looking to get the most out of your SEO efforts, I’d suggest keeping an eye out for those monthly posts, these videos and the Webmaster Help videos Google’s Matt Cutts puts out. In fact, add Duane Forrester’s videos and Bing’s search quality series to the list as well.
Google recently started a series of blog posts highlighting various points of progress it has made in its efforts to increase search quality. The results have essentially been lists of monthly changes Google has made to its algorithm and interface. This has been extremely interesting to watchers of the search industry and webmasters who are concerned with search engine optimization.
Bing is now starting a similar series, though if the first entry is any indication, we’re going to see a much more in-depth explanation about things in general, as opposed to lists of specific tweaks.
“Quality improvements in Bing are often subtle but often those little changes are the result of years of research,” says Dr. Harry Shum, Corporate Vice President of Bing R&D. “In the coming weeks and months, you will hear from members of my team on a range of topics, from the complexities of social search and disambiguating spelling errors to whole page relevance and making search more personal. We will also highlight the ideas and projects we have collaborated with colleagues from Microsoft Research and academia to advance the state of the art for our industry. We hope this will not only be useful information for our blog readers, but that they will spark conversations that help us all move the search industry forward.”
The first entry comes from Jan Pedersen, Chief Scientist for Core Search at Bing, who talks about how Bing determines “whole page relevance,” which it uses to determine not just where to rank a result on the search results page, but whether to just return a link or an “answer”.
“As with any relevance problem we start with the question of how to measure if Bing has done a good job,” explains Pedersen. “We could do this by simply asking human judges to compare the output of competing blending algorithms and assess which is better. This turns out to be a difficult judgment task that produces quite noisy and unreliable results. Instead we look at how people behave on Bing in the real world. Based on how they respond to changes we make an assumption that a better blending algorithm will move people’s clicks towards the top of the page. This turns out to be the same as saying that a block of content, or answer, is well placed if it receives at least as many clicks as the equivalently sized block of content below it — or, as we say internally, if its win rate is greater than 0.5. So a good blending algorithm will promote an answer on the page upward as long as its win rate is greater than 0.5. Armed with this metric, we can run online experiments and compare the results of competing blending algorithms giving us a realistic data set.”
Shum did note in a blog post announcing the new blog series that Bing does measure search quality with a mix of offline human judges (presumably similar to Google’s raters) and online user engagement.
“Next we investigate the available inputs into an online blending function that improves this metric,” continues Pedersen. “We can, and do, use historical anonymous click data, but this is not sufficient because it does not generalize to rare queries, or to new content with no history. So, we add in three kinds of additional inputs: confidence scores from the answer provider, query characterizations, and features extracted from other answers and web pages that will be shown on the page. For example, to learn where to place an image answer in the search results for a given query, we consider the confidence score returned from the image search provider, the ranking scores of nearby Web pages, and whether the query is marked as referring to the sort of entities that are well described by images (people, places, etc.).”
Bing actually uses over a thousand signals for blending search functions, according to Pedersen. This is in line with what Bing’s Duane Forrester has said in the past – roughly a thousand signals.
“Finally, we consider the offline and online infrastructure that will be used to create and run a blending function,” adds Pedersen. “We use a very robust, but high-performance learning method, called boosted regression trees, to automatically produce a ranking function given training data. This allows us to use many signals with the confidence that each additional signal will incrementally improve our blending function. Our training sets are fairly large, since they are mined from our billions of anonymous query session logs, so we use our large-scale data mining infrastructure, called Cosmos, to prepare the data and run offline experiments. Once a new blending function has been generated by our offline learning method, it is deployed to a serving component internally called APlus that puts all that data into action and runs after all candidate content blocks that have been generated, where it can be tested via online experimentation and finally placed into production.”
Pedersen says Bing has been focusing on applying all of this to new inputs for “temporarily relevant” answers. Think news stories that die down after a while.
If you’re really into learning about the inner-workings of search engines, it’s a pretty interesting read, and is frankly not he kind of thing we see from Bing very often. It looks like that’s changing now.
There’s not much here in terms of SEO guidance, at least in the first post, but SEO enthusiasts will no doubt want to keep an eye on the series and stay on the lookout for info and tips that could be applied to SEO strategies. Google’s change lists are a bit more useful in this regard.
Here’s a pair of interesting Google videos.
The first one is a presentation from Google’s “Uber Tech Lead for Search Quality and User Happiness” Dan Russell. Here’s the abstract for it:
How well DO people search on Google? Although popular opinion is that “everyone is above average,” that obviously can’t be true. The truth is that self-perceptions of search expertise are often wildly over-estimated and that people, on average, actually use only a fraction of the potential of Google. They both don’t know much of what’s possible, and don’t understand where internet search capabilities are headed. Since the rate of change and improvements isn’t slowing down, in this talk I’ll examine where we are, and where we’re headed, and conclude with some heuristics for teaching research skills in the years ahead.
The video lasts about 36 minutes.
The second video comes from the Khan Academy, which Google’s first employee (after co-founders Larry Page and Sergey Brin) just announced he was leaving Google for. Russell also appears in this video, which is much shorter, about searching for similar images:
As you may know, Google has been putting out a monthly list of algorithm changes it has been making, as part of the company’s initiative to be “more transparent”. Google will never put out the entire secret sauce of its algorithm (without a court order, at least), so webmasters can at least be thankful that they’re being thrown a handful of bones in the form of a monthly list.
Have you seen effects from Google’s most recent algorithm changes? Let us know in the comments.
Some have apparently already been feeling the effects of Google’s algorithmic early this year. Webmasters were quick to point the finger at the old panda, but Google assures us that this is not the case. A spokesperson for the company told WebProNews there have been no Panda updates in 2012 so far (though I’d expect one to launch before too long).
By the way, now that 2011 is over, do you think Panda has done a good job at cleaning up search results?
In the latest edition of the series, on the company’s Inside Search Blog, they highlight 21 changes made in the month of December. The list goes as follows:
The image search landing page quality signal change is quite interesting. We ran a great article on optimizing for image search by Michael Gray last year, and that’s full of tips to consider for this less talked about element of SEO, but the adjustments, as unspecific as they may be, reflect Google’s Panda-style focus on quality in search results. This, to me, is saying they’re applying same kind of thinking they do with regular web search to other parts of Google, more than ever before.
Here’s the list of questions Google has presented in the past to consider asking yourself, when evaluating quality.
Note that “better spam detection for image search” is also on the list.
Also note the codenames used throughout the list. Most you probably won’t have to remember like Panda and Caffeine, but it’s still nice to have something to reference for the future.
With regards to the “more rich snippets” item on the list, you may want to check out the series of videos Google recently put out on how to do rich snippets.
Which changes do you think are the most significant? Is your site being helped or hurt by changes? Let us know in the comments.
Following the previously reported upon controversy surrounding a marketing campaign for Google’s Chrome browser, Google has apparently decided to devalue its Chrome landing page (in terms of PageRank). At least temporarily.
Danny Sullivan at Search Engine Land acquired a statement from Google, which says:
We’ve investigated and are taking manual action to demote www.google.com/chrome and lower the site’s PageRank for a period of at least 60 days. We strive to enforce Google’s webmaster guidelines consistently in order to provide better search results for users. While Google did not authorize this campaign, and we can find no remaining violations of our webmaster guidelines, we believe Google should be held to a higher standard, so we have taken stricter action than we would against a typical site.
Later, Google’s Matt Cutts jumped on Google+, while trying to take a “digital break” for a holiday in Central America. I his post, he said:
I’ll give the short summary, then I’ll describe the webspam team’s response. Google was trying to buy video ads about Chrome, and these sponsored posts were an inadvertent result of that. If you investigated the two dozen or so sponsored posts (as the webspam team immediately did), the posts typically showed a Google Chrome video but didn’t actually link to Google Chrome. We double-checked, and the video players weren’t flowing PageRank to Google either.
However, we did find one sponsored post that linked to www.google.com/chrome in a way that flowed PageRank. Even though the intent of the campaign was to get people to watch videos–not link to Google–and even though we only found a single sponsored post that actually linked to Google’s Chrome page and passed PageRank, that’s still a violation of our quality guidelines, which you can find at http://support.google.com/webmasters/bin/answer.py?hl=en&answer=35769#3 .
In response, the webspam team has taken manual action to demote www.google.com/chrome for at least 60 days. After that, someone on the Chrome side can submit a reconsideration request documenting their clean-up just like any other company would. During the 60 days, the PageRank of www.google.com/chrome will also be lowered to reflect the fact that we also won’t trust outgoing links from that page.
Other Googlers have responded to Matt’s explanation on Google+. Here are a few of the responses:
Great explanation from +Matt Cutts. This is embarrassing, but a good illustration of two things:
1. Why I like working at Google. The Search Quality Team tries to apply the Webmaster Guidelines fairly – even on other Google products.
2. Why you should pay attention to what any marketing, advertising, or SEO companies might be doing on your behalf.
Was mortified yesterday to read Danny Sullivan’s story, and sad that he thought we were paying bloggers to promote our Chrome films in such a crude and horrible way. Having been in Google marketing for over 5 years I can testify that we don’t do advertorial or pay bloggers! It’s one of the things that makes me love my job here, we let the products speak for themselves and don’t overpromise. So this was pretty shocking. It was a media buy that went very wrong, we never agreed to do anything else than paid video ads – but clearly we need to be more on top of our media buys. As Matt explains in his post we’ve demoted the Chrome page in the search rankings in response.
This is the kind of integrity and transparency that makes me proud to work at Google. Show me another company that would have acted in this way.
This was the right thing to do. Glad to see we did it.
Of course, the whole search quality aspect wasn’t discussed in Matt’s post. One of the posts from the Chrome campaign, which I referenced in another piece, which has pretty “thin” content, is still ranking for “chrome small business benefits”.
I don’t think Chrome losing its PageRank will help quality there, though it is still at the top of the page, via AdWords.
Google tweaks its search algorithms over 500 times a year. You may have already known that, but Google is sharing a new video today designed to give people a “deeper look” into how Google makes “improvements” to its algorithms.
I don’t think everyone would agree that they’ve all been improvements, as we see complaints about this every day, but Google makes a lot of changes aimed at improving search results, nevertheless.
“There are almost always a set of motivating searches, and these searches are not performing as well as we’d like,” says Engineering Director Scott Huffman. “Ranking engineers then come up with a hypothesis about what signal, what data could we integrate into our algorithm.”
The video does provide some unique behind the scenes footage of Google engineers plugging away on their computers, presumably working on the algorithms.
Google briefly talks about the process of raters. “These are external people that have been trained to judge whether one ranking is more relevant and higher quality than another,” says software engineer Mark Paskin.
“We show these raters a side-by-side for queries that the engineer’s experiment might be affecting,” explains Google Search Scientist Rajan Patel. “We also confirm these changes with live experiments on real users.”
“We do this in something called a sandbox. We send a very small fraction of actual Google traffic to the sandbox. We compute lots of different metrics,” says Paskin.
“In 2010, we ran over 20,000 different experiments. All the data from the human evaluation and the live experiment are then rolled out by a search analyst,” says Huffman.
Sangeeta Das, a quantitative analyst says, “For each project, it’s usually one analyst assigned from the moment that we’re talking to the engineers, trying to learn about their change.”
“We then have a launch decision meeting where the leadership of the search team then looks at that data and makes a decision,” says Huffman.
“Ultimately, the goal of the search eval analyst team is to provide informed, data-driven decision, and present an unbiased view” says Das.
“If our scientific testing says this is a good idea for Google users, we will launch it on Google,” says Google Fellow Amit Signhal.
The video then looks at the “did you mean” and “showing results for” features as an example.
Google hires employees all the time, but interestingly the company has basically put out a help wanted ad on its Webmaster Central blog, looking for academics to help it improve search quality and fight spam.
“Yes, we’re looking for help improving Google search—but this time we’re not asking you to submit more spam reports,” writes Kaspar Szymanski of Google’s Search Quality team. “Although we still appreciate receiving quality spam reports, today we’ve got a different opportunity for you to improve Google Search: how about YOU join our team and do the webspam fighting yourself?”
Earlier this month, Google launched what it called “the biggest refresh in 10 years” to its spam report form. We looked at this more closely here.
It’s worth noting that the job is in Dublin. Szymanski posted the above image, saying that it’s an actual work environment photo taken at the Dublin office.
“Here’s what we’re looking for: open-minded academic graduates willing to work in a multinational environment in our Dublin office,” says Szymanski. “Looking at a site’s source code should not scare you. You should be excited about search engines and the Internet. It’s also essential that you share our aversion to webspam and the drive to make high-quality content accessible. PlayStation or foosball skills are a plus.
So what does Google look for in a spam fighter (or “Search Quality Associate”) exactly?
The requirements are listed as:
“As a Search Quality Associate, you have the opportunity to make a significant and direct impact on the quality of Google’s search results through search quality evaluation,” Google says. “You will be working on the cutting edge of search and the forefront of the web ensuring quality information is provided to millions of internet users, and you will be expected to keep pace with constant change in a fast-paced work environment, bringing innovative ideas to improve access to relevant information on the web. You are a web-savvy individual who is a take-charge team player, as well as a quick learner and strongly interested in providing a better search experience for Google users.”
The position requires he candidate to review assigned sites for quality and content and improve the quality of search results by evaluating sites “to identify areas of concern and interest.”
Since the Google Panda Update first launched back in February (and really for some time before that), there has been a lot of discussion about search quality throughout the industry – the quality of the content that search engines are returning in their results.
This is the whole reason the Panda update exists. It’s all about improving the quality of results. Some will dispute the success of that, but it is the reason for better or for worse.
But what about Bing? It doesn’t command nearly the search market share that Google does, but as it powers Yahoo search, it’s really the only major competitor in town.
Bing talked a bit about its own views on content quality this week, and content producers might do well to check take notice of that as well – especially those who may have been hit by the Panda update, but are still doing ok in Bing.
Whereas Google had a list of questions one could ask themselves to asses the quality of their site, Bing has published a list of things to avoid, which reads as follows:
The rest of Bing’s advice basically boils down to focusing on creating a good user experience and letting Bing know about your content. “Whether you call them rich snippets or by their proper names, the act of marking up your content to tell the engines more details about the content is a wise investment,” says Bing’s Duane Forrester. “By following the plan outlined at Schema.org, you can embed meta tags around your content. Visitors won’t see them, but the search engines will, enabling us to understand your content and use it in unique ways to create more engaging search experiences. Take some time and review this idea to see if you can leverage the great content you’re creating in new ways.”
If you’re living up to Google’s definition of quality, you probably won’t be doing too bad in Bing either, and if you’re doing well in Google, you’re probably getting a lot more search referrals from Google than you could ever get from Bing anyway, but it’s still helpful to get a look into Bing’s own thinking on this issue.
Blekko, the alternative search engine that aims to challenge Google and Bing by reducing spam and low quality content in search results via human curation, has refreshed its index and results pages in an update it refers to as “Zorro”. We picked CEO Rich Skrenta’s brain about the update, search quality, and blekko’s goals in general.
“Zorro is a major upgrade in our relevance,” Skrenta tells WebProNews. “blekko users have spent the past six months curating the web on category by category basis, telling us the best sites for broad categories like health and personal finance as well as narrower categories like gluten free.”
“Zorro incorporates those human-curation efforts into our result set for non-slashtag queries by boosting pages from the curated sites – even for non-slashtag queries,” he adds. “With Zorro, we can boost results from multiple slashtags to make results better. ex.: https://blekko.com/ws/pregnancy+tips”
“The net result is further reduction of spam,” he says.
On how the Zorro update improves the search experience compared to competitors, Skrenta tells us, “Other search sites rely wholly on algorithmic intelligence for results. We are incorporating human-curation efforts directly into our results. Given the amount of SEO gaming being done, we believe that only humans can accurately differentiate a clever spam site from a quality site.”
“We initially integrated slashtags for 10 sites into our standard results,” he says. “Now we are incorporating hundreds. Our users have created over 100k slashtags since we launched. That data wasn’t available on day 1.”
Blekko’s mission is to provide search results without spam. That takes a lot of filtering. There is a discussion going on around the web right now about how what we see on the web is becoming more and more filtered. Eli Pariser calls it the “Filter Bubble,” and DuckDuckGo, another alternative search engine and peer of blekko’s launched a site discussing this very topic. This is more about search engines and other sites (including social networks like Facebook) filtering what we see by tailoring content delivered to us on a personalized level. A lot of people don’t like the idea of having this content filtered. With blekko doing its own kind of filtering, we wondered what Skrenta might have to say about this.
“We’ve reached a tipping point on the web where it is easier to white list the set of good sites than black-list the set of bad sites,” he tells us. “ex. the top 100 health sites will answer all your health questions. You don’t want to search outside that set of sites. Our efforts with Zorro combine the best of curation and algorithmic intelligence to deliver spam free results.”
Blekko somewhat famously (at least within the search industry) blocked a number of sites deemed “content farms” from its results. Even today, blekko’s home page carries the message: “Slashing out…spam…content farms…malware.” Among the sites blocked were a few from Demand Media, including eHow, which has consistently carried the “content farm” label, despite the company’s best efforts to position it in a different light.
Demand Media has made it a point to clean up eHow’s quality (more on this initiative here), so we wondered if blekko’s banning of eHow, or any site, can be reversed. “We constantly review sites for quality and our users continually identify quality sites for us,” Skrenta says.
Along with the Zorro update, blekko launched a little game called “3 Engine Monte“. It’s available via a link on blekko’s home page, and invites users to enter a query and then choose from a set of three results sets, which one they like the best. One is from Google, one is from Bing, and the other is of course from blekko.
We asked if the majority are picking blekko most often. “We are currently collecting data and will have results soon,” Skrenta tells us.
When asked about long-term goals for blekko, Skrenta simply says, “Clean up the web from spam, category by category.”
Alternative search engine Blekko has launched a major update, called “Zorro,” which includes an expanded search index with integration of the slashtags, which have been the staple of the site.
On the new results pages, users will see sites that others have hand-picked to be included in one or more slashtags. So far, users have created over 100,000 of them. The company considers this a human site review element to the “war on spam.” The slashtags appear at the top of the search results, and for each site boosted by the slashtag, next to its URL.
“Millions of users and hundreds of millions of searches have given us insight into what is truly great quality content on the Web and what is poor quality spam,” said Blekko CEO Rich Skrenta. “This new version of blekko bakes in that intelligence in every search so the spam gets weeded out and the best content comes to the top in every search.”
In addition to that, search results will display sites’ logos. Blekko says this is so users can “quickly choose results from trusted brands that offer the best content.”
Though stumbling a bit in some months, Blekko has continued to grow since its launch. Compete has Blekko.com at 211k unique visitors in the U.S. last month, after launching last fall.
Finally, the Zorro update comes with a “new relevance model for ranking.” This includes “a ramped up adspam algorithm, identifying millions more pages on the Web that contain multiple ads and little content.” Such pages, Blekko says, have been permanently elminated from the index so they’ll never appear in search results.
Blekko is so confident in its search quality, it has also introduced a little game called 3 Engine Monte, which lets you enter a query and see three different sets of results. One comes from Blekko, and the others from Google and Bing. The user’s job is to pick which one they like best, and they think users are apt to pick Blekko fairly frequently, I’m assuming.
3 Engine Monte doesn’t appear to take into account various kinds of search results offered in the other engines through Universal search. For example, if I search for “mexican food” it doesn’t show me all of the local stuff and the images that Google actually shows me if I go to Google and perform the search. As far as the purse “ten blue links” type results, I’ll give Blekko credit for offering better ones (and Bing as well for that matter) in some instances that I tried.
Interestingly, it does show you which results in the competitors’ results that it has banned. The game can be accessed from the home page of Blekko.
Google has pulled its search engine at google.kz out of Kazakhstan, where the country’s government is requiring all .kz domain names to be operated from servers located in the country. Now when you go to google.kz, you’re redirected to google.com/webhp?hl=kk.
Google says its users in kazakhstan may see a decrease in search quality, but that the company does not want to contribute to a fractured Internet. Here is the entire explanation from Google SVP, Research & Systems Infrastructure, Bill Coughan, as posted on the official Google Blog:
The genius of the Internet has always been its open infrastructure, which allows anyone with a connection to communicate with anyone else on the network. It’s not limited by national boundaries, and it facilitates free expression, commerce and innovation in ways that we could never have imagined even 20 or 30 years ago.
Some governments, however, are attempting to create borders on the web without full consideration of the consequences their actions may have on their own citizens and the economy. Last month, the Kazakhstan Network Information Centre notified us of an order issued by the Ministry of Communications and Information in Kazakhstan that requires all .kz domain names, such as google.kz, to operate on physical servers within the borders of that country. This requirement means that Google would have to route all searches on google.kzto servers located inside Kazakhstan. (Currently, when users search on any of our domains, our systems automatically handle those requests the fastest way possible, regardless of national boundaries.)
We find ourselves in a difficult situation: creating borders on the web raises important questions for us not only about network efficiency but also about user privacy and free expression. If we were to operate google.kz only via servers located inside Kazakhstan, we would be helping to create a fractured Internet. So we have decided to redirect users that visit google.kz to google.com in Kazakh. Unfortunately, this means that Kazakhstani users will experience a reduction in search quality as results will no longer be customized for Kazakhstan.
Measures that force Internet companies to choose between taking actions that harm the open web, or reducing the quality of their services, hurt users. We encourage governments and other stakeholders to work together to preserve an open Internet, which empowers local users, boosts local economies and encourages innovation around the globe.
It’s not that surprising that Google would make such a move, as the company has promoted an “open” web consistently for years, and after the ordeal with China, they made it clear that they’re not above pulling out of a country, and frankly, China is a much bigger economy than Kazakhstan.
Still, it is interesting that this comes at the sacrifice of search quality, and that Google is openly pointing this out, at a time when Google’s search quality has been heavily criticized and iterated upon relentlessly by the company, with recent algorithm updates.
We’ve been covering a Google’s Panda Algorithm update a lot since its initial launch (in some ways even before its launch). We thought it might be useful for some to provide something of a round-up of coverage as a one stop shop for those looking to learn more about the algorithm update, its impact on websites, and related stories.
So here is a list of our articles related to the Panda update, content farms, and search quality from the past several months (updated continuously).
Despite New Panda Guidelines, Google Still Burying Authoritative Results
Panda Update Already Happened. Not Every Google Tweak Is Still Panda.
What Would Google Search Quality Be Like Without AdSense?
Helium Raises $10 Million After Being Victimized by Google Panda Update
Don’t Expect a Lot of eHow Content to Be Removed from YouTube
Google Panda Update: New Advice Directly From Google
Google Competitive Practices Brought Into the Panda Conversation
Demand Media Announces 20% Decline in eHow Search Referrals Due to Panda
Demand Media Earnings Report – Revenue Up, Questions Remain
Demand Media Deletes eHow Articles, Edits Others in Quality Clean-up Initiative
Google Algorithm Update – Is Bounce Rate a Ranking Signal?
Demand Media Search Data Released Ahead of Earnings Call
Google Panda Update – Webmasters Still Trying to Crack the Code
eHow Getting Serious About Quality
Mashable Affected By Google’s Panda Update? Not Likely.
Google Panda Update – A Broader View of U.S. Traffic Patterns
Panda Reference in Google Earth Day Doodle
Google Panda Update Victim HubPages Tweaks Approach to Boost Search Visibility
Panda Update: HubPages Asks Google Why YouTube Gets Away With Softer Content Policy
Google Panda Update Helps Local Search Results
Google Algorithm Update Fallout, eHow Response
Google Panda Update Winners, Losers, and Future Considerations
EzineArticles Hit By Google Panda Update Again
Google Panda Update Hits Demand Media’s eHow This Time
Google Panda Update Benefits Google Properties
Ranking in Google Now That Panda Has Gone Global
Google Panda Update Winners: Video, News, Blogs, and Porn
Google Panda Update Victim Xomba Loses AdSense Ads Too
Google Panda Update Officially Goes Global (In English)
Google Panda Update Launched in More Countries
Google Panda Update – Made for Big Brands?
Google Does Guest Post for Panda Victim HubPages
Examiner’s Approach to Content Quality Post Panda Update
MerchantCircle Goes From Panda Victim to Blekko Curator
EzineArticles Tells Authors How to Get Accepted Post Panda
Google Panda Update Hit E-Commerce Sites Too
Google’s Algorithm Impact Over the Years in Graphic Detail
Why Panda is the New Coke
EzineArticles Traffic Update Post Panda
Google Panda Update: Lack of Consistency on Quality?
Google Panda Update Still Encouraging Higher Quality
Suite101 CEO Writes Open Letter to Google’s Matt Cutts
Google “Panda” Algorithm Update – What’s Known & What’s Possible
Calacanis on Google Algorithm Aftermath and Impact on Mahalo
Google Algorithm Update to Get “New Layer” to Help “Falsely Caught” Sites
EzineArticles Aims to Get Rankings Back, Following Google Algorithm Update
Did Google’s Algorithm Update Go Far Enough on Content Farms?
Is This Google Algorithm Change About Content Farms or Not?
Quora vs. eHow: Where’s the Better Quality?
Decreasing Google Dependence: A Growing Trend
Demand Media Redesigns eHow with Quality Control Feature
Will Google Fill In Its Own Search Gaps, Demand Media-Style?
Retrevo Says Google Update Pushed Bigger Brands Up
AdSense and Its Relationship to Search Rankings
Google Algorithm Update Casualties Speak
Google Update Costs Mahalo Employees Their Jobs
Google Algorithm Update Helps (Not Hurts) eHow
Google Finally Cracks Down on Content Farms
Demand Media Responds to Google Content Farm Update
Demand Media Goes on the Defensive About Content Quality
Content Marketer or Content Farm?
If Google Tweaked Its Search Results …
An Inside Look at wikiHow Content Quality Control
User Feedback: The Next Google Ranking Signal?
Why It’s Easier for a Startup (Than For Google) to Take Action on Content Farms
wikiHow On Why Wikis Deliver Higher Quality Than Content Farms
DuckDuckGo Follows Content Farm Banning With Promoting wikiHow Content
Blekko Queries on the Rise, More So Since Content Farm Blocking
What If Content From One Company Dominated Google’s Search Results?
Confirmation: The Google Algorithm Change Was Not for Content Farms
Blekko Bans eHow and Other Content Farms
Demand Media CEO: Google Not Talking About Us
Google, Bing, and Blekko Talk Content Farms and Search Quality
The Real Problem With Content Farms is Google
Nobody said we’re done yet either. We’ll update the article as new developments, insights, and research occur.
Meanwhile, we might as well use the comments section to expand the article as a useful resource. If you have any of your own Panda-related stories or insights, please don’t hesitate to share them with the rest of us.
I don’t think many people will argue that Google’s AdSense program has been a major catalyst in increasing the amount of content/search spam on the web. This may not have been Google’s intention for the service, but it has clearly contributed. I’d love to see the ratio of sites that were hit by the Panda update that displayed AdSense ads to sites that were hit and didn’t display these ads.
That’s not to say that simply using AdSense will get you penalized. Of course Google doesn’t want that. It makes money from these ads, but it is interesting to see how AdSense publishers of all kinds have been impacted by the update.
One can’t help but wonder what Google’s search results would look like if sites using AdSense ads were removed. Would the quality be better? Maybe. Maybe not. It would be interesting to see either way. Obviously that will never happen, unless Google one day pulls the plug on AdSense, which is also highly unlikely.
Google recently released a list of questions “that one could use to assess the ‘quality’ of a page or an article,” in light of the Panda update. How many sites do you come across regularly that meet all of these criteria and run Google AdSense ads? To recap, here’s the full list:
It is certainly possible to have a “quality” site and use AdSense ads. There are plenty of examples out there, but is that the norm?
Interestingly enough, Google is reportedly turning away some advertisers that were hit by the Panda update from advertising with AdWords. Aaron Wall of SEOBook tells an interesting story about a guy this has happened to. Here’s the situation as Wall presents it (pulling no punches):
That’s not everybody’s experience, but it’s also not the only such complaint we’ve seen. It’s not hard to find a similar analysis in any webmaster forum or comment section on a related article.
@aaronwall
aaron wallhttp://www.youtube.com/watch?v=yBrbgh3CO-I Don’t Be Evil, just be corporate 3 hours ago via web · powered by @socialditto
We reported before, that another Panda victim, Xomba, had its AdSense ads completely removed following a bogus takedown notice, though Google did restore them shortly thereafter.
For another Panda victim – HubPages, a Googler went so far as to write a guest post on the company blog telling writers how to produce better content for AdSense. Granted, that was before the global roll-out of the update.
One thing regarding Panda and AdSense seems pretty clear. Don’t overdo it on the ads. Don’t “have an excessive amount of ads that distract from or interfere with the main content.”
[Image credit: kawanet]
We’ve covered HubPages a number of times since Google’s Panda update originally launched, as they were one of the victims that lost significant search visibility as a result. Since the update, the site has been making various adjustments to its editorial policy and certain features.
Interestingly enough, a Googler actually posted a guest post to the HubPages blog before the Panda update rolled out globally, providing tips for HubPages writers to put out better content for AdSense. Since the global roll out, HubPages announced changes like the removal of a “news capsule” feature, and tightening up its policy on affiliate links.
Now, HubPages has released an overview of its recently added standards. These include:
HubPages says it will send emails to authors letting them know which “Hubs” require revision, and they’ll have two weeks to amend them. Otherwise, they’ll be unpublished. However, they can be resubmitted for publication later and will be reviewed again.
They’ve also added automatic alerts, letting users know when they’re at risk of violating one of their guidelines.
Earlier this week, we looked at how Demand Media’s eHows is taking content quality more seriously, based on a job posting paying $17-20 per article (up from $10-15) with strict qualifications requiring a degree in business, finance, or law, and “extensive experience in business writing.”
YouTube announced today that it is transcoding all new videos into the WebM format, as well as other supported formats, which include MPEG4, 3GPP, MOV, AVI, MPEGPS, WMV, and FLV.
In the announcement, they say they are working to transcode the rest of the YouTube catalog.
“Given the massive size of our catalog – nearly 6 years of video is uploaded to YouTube every day – this is quite the undertaking,” says software engineer James Zern. “So far we’ve already transcoded videos that make up 99% of views on the site or nearly 30% of all videos into WebM. We’re focusing first on the most viewed videos on the site, and we’ve made great progress here through our cloud-based video processing infrastructure that maximizes the efficiency of processing and transcoding without stopping.”
While clearly the transcoding videos is the real news here, I find that stat Zern dropped somewhat interesting. 30% of all videos account for 99% of views. Not all that surprising, but interesting still.
I find that stat perhaps even more interesting, conisdering some points HubPages just brought up about YouTube, with regards to Google’s Panda update.
“While we believe the democratization of publishing and earning potential is an important part of the progress of the Web, we want to avoid a situation where a portion of content negatively impacts the rankings of high quality content,” said HubPages CEO Paul Edmondson. “It appears HubPages has been impacted by this while YouTube has not, despite HubPages having a more strict content policy.”
For context, HubPages was one of the top sites negatively impacted by the Panda update. Clearly, some people have indeed had negative experiences with HubPages content, but I believe there is some higher quality content in the mix – as there usually is with sites labeled “content farm”.
He makes an interesting point about YouTube. The stat Zern drops, kind of backs up the point. Views aren’t necessarily equivalent to quality, but quality videos do tend to gain views. By nature, people want to share quality content with others. If only 30% of YouTube’s videos are making up almost all views, it would seem that there’s a lot of lesser quality content on there.
It’s hard to say where this content places in any given search results page, but it’s something to think about.
“Actually, HubPages is to articles what YouTube is to video,” Edmondson told WebProNews in a recent interview. “Like YouTube where enthusiasts post videos of their choice, our community write articles about whatever they wish and are passionate about. This covers a wide range of content from poetry to recipes, and pretty much everything in between. Writers choose what they write about, and they own their content. In return, they stand behind the content, build readership and interact within the HubPages community.”
With the recent international roll-out of Panda, YouTube was among the top gainers (as were other competing video sites).
As you know, Google has rolled out the Panda update internationally in the English language. SearchMetrics has put out a list of the top victims in the UK. We’ll be exploring more of this in time, but one thing that realy sticks out in this report is the presence of eHow on the list.
Demand Media’s eHow, a site often referred to as a content farm, was expected to be one of the main targets of the initial U.S. Panda update by many people, after Google announced it would be making algorithmic adjustments to go after content farms. eHow, while not considered a content farm by Demand Media, is usually one of the first sites mentioned when the topic comes up in discussion on the web.
eHow actually managed to gain traffic after the update, which left a lot of people very surprised. Now that Google has rolled out the algorithm update internationally, and made more adjustments to the U.S. version, it would appear eHow wasn’t so lucky this time.
According to the SearchMetrics data, eHow.co.uk took a 72,30% hit in visibility. eHow.com took a 53,46% hit. Keep in mind this is UK data, but it has also affected U.S. sites.
Also keep in mind that Google did make some tweaks in the U.S. “In some high-confidence situations, we are beginning to incorporate data about the sites that users block into our algorithms,” said Google’s Amit Singhal. “In addition, this change also goes deeper into the ‘long tail’ of low-quality websites to return higher-quality results where the algorithm might not have been able to make an assessment before. The impact of these new signals is smaller in scope than the original change: about 2% of U.S. queries are affected by a reasonable amount, compared with almost 12% of U.S. queries for the original change.”
They’re using the domain blocking data in search rankings. Now, if eHow is the most commonly referenced site when people talk about content farms, it stands to reason that it is probably one of the most blocked domains too. Google says it only uses this in “high-confidence situations” (whatever that means). I take this to mean situations where they’re confident it’s not just people abusing the feature for competitive purposes.
We don’t have the hard data about eHow’s performance in U.S. Google, but searching for some queries we’ve looked at in the past, it does look like eHow has taken some kind of hit – at least on some content. For example, the “level 4 brain cancer” result we’ve referenced in several articles has moved down. Granted, it’s still the second result, and with another ehow article ranking directly below it at number 3, but it’s not the top result anymore. Now the top result is from the MGH Brain Tumor Center. Clearly something’s happened.
I’m still waiting to receive comment from Demand Media about the impact of the Google’s update.
Demand Media has been putting a great deal of emphasis on an increase in quality at eHow. This has been going on since before the Panda update ever hit. Recently, they launched a new redesign and a user feedback feature aimed at sending questionable content back through the editorial process or to the cutting room floor.
Demand Media also been making partnerships aimed at boosting credibility and engagement (Rachael Ray, Tyra Banks, Getty Images, Buddy Media), and making significant acquisitions (CoverItLive).
The company has also acknowledged that there is some older content out there that isn’t necessarily up to their current standards, and they’ve been working to improve that as well.
Whether the content quality level at eHow is improved with the newer content is a matter of opinion, but there’s no question that they’re spending a lot of money on boosting perception.
Since Google launched the controversial Panda update in February, anxious webmasters and publishers have been waiting for the day when it would go from just a U.S. change to a global change. That day came this week, as Google announced it had expanded the “search quality” algorithm worldwide for English language users.
Have you been impacted by the global roll-out of the Panda update? Let us know.
Many sites were reported to have lost major traffic and search rankings as a result of the U.S. roll-out. We’ve expected these sites to be further devastated as Panda made its way to more countries.
Side note: just in case there is any confusion (there has been in the past) since Google didn’t use the word “panda” in the latest announcement, Matt Cutts did call it that:
@mattcutts
Matt CuttsToday Google rolled out “Panda” algorithmic improvement globally to all English-language Google users: http://goo.gl/RLhHW 1 day ago via web · powered by @socialditto
Believe it or not, not everyone’s quick to comment on how their traffic has been impacted. EzineArticles was one of the sites that was hit hard, and quickly began making changes to its content approach following the U.S. update. They talked about this on the company blog. No new blog post yet.
The last time I reached out to CEO Chris Knight for comment on the site’s traffic post-Panda, he didn’t want to offer any – although he did provide another update on the blog in March.
Mahalo CEO Jason Calacanis, who had to resort to layoffs because of the Panda update did comment on the situation. “We were impacted starting on February 24th and haven’t seen a significant change up or down since then,” he tells us. “We support Google’s effort to make better search results and continue to build only expert-driven content. This means any videos and text we make has a credentialed expert with seven years or 10,000 hours of experience (a la Malcolm Gladwell).”
He pointed to the following examples:
http://www.mahalo.com/learn-to-play-guitar-beginner-course/
http://www.mahalo.com/how-to-make-profiteroles/
“Everything we’re doing is now expert driven. Period,” Calacanis says. “Google’s Panda update has ended the age of casual, non-expert-driven content. We’re hopeful that as Google continues to tweak their algorithm our experts and their content will be rewarded.”
Morris Rosenthal of Foner Books (a Panda victim whose story we looked at here and here) tells us, “As far as the algo goes, I think my initial take was 100% on. It’s just brand names and social sites now. I’m even seeing Google Answers ranking for many queries, despite the fact they shut Answers down in 2002! But it was ‘social.’”
Last week, we looked at a report from marketing firm iCrossing, which seemed to indicate benefits for big brands as well.
We’ve reached out to Suite101, and Demand Media for comment on the update, and have nothing so far.
Google’s Amit Singhal told webmasters, “Based on our testing, we’ve found the algorithm is very accurate at detecting site quality. If you believe your site is high-quality and has been impacted by this change, we encourage you to evaluate the different aspects of your site extensively. Google’s quality guidelines provide helpful information about how to improve your site. As sites change, our algorithmic rankings will update to reflect that. In addition, you’re welcome to post in our Webmaster Help Forums. While we aren’t making any manual exceptions, we will consider this feedback as we continue to refine our algorithms.”
Following the initial launch of the Panda update, we looked at some things that are known and some things that are possible, with regards to what publishers and webmasters should take into account. Given Google’s specific reference to their quality guidelines in the latest announcement, this seems like the best place to start. Here’s Google’s list of “specific guidelines”:
They also get into “basic principles,” which deal with avoiding obvious black hat techniques. I’d suggest reading this whole page of webmaster guidelines, which deal not only with quality, but also design/content, and technical guidelines.
“Following these guidelines will help Google find, index, and rank your site,” Google says. “Even if you choose not to implement any of these suggestions, we strongly encourage you to pay very close attention to the ‘Quality Guidelines,’ which outline some of the illicit practices that may lead to a site being removed entirely from the Google index or otherwise penalized. If a site has been penalized, it may no longer show up in results on Google.com or on any of Google’s partner sites.”
Cutts also lent his endorsement of this advice from ex-Googler Vanessa Fox.
Google is also experimenting with better communication with webmasters who do reconsideration requests, according to Cutts:
@mattcutts
Matt CuttsWe’re running an experiment with better communication for site owners who do reconsideration requests: http://goo.gl/FrNzp 1 day ago via web · powered by @socialditto
The link goes to a Search Engine Roundtable article looking at an email a webmaster received from Google’s Search Quality team. Here’s an excerpt of the email:
If you’ve experienced a change in ranking which you suspect may be more than a simple algorithm change, there are other things you may want to investigate as possible causes, such as a major change to your site’s content, content management system, or server architecture. For example, a site may not rank well if your server stops serving pages to Googlebot, or if you’ve changed the URLs for a large portion of your site’s pages. This article has a list of other potential reasons your site may not be doing well in search.
If you’re still unable to resolve your issue, please see our Webmaster Help Forum for support.
While the Panda update has really only been around for a couple months, a lot has happened on the Google search quality front since its initial launch. For one, they announced the +1 button, and clearly said it would be used as a ranking signal.
They also started letting users block domains. This was also brought up in the latest announcement, as Google is now using it as a ranking signal. Signhal’s specific words on that, were ” In some high-confidence situations, we are beginning to incorporate data about the sites that users block into our algorithms. In addition, this change also goes deeper into the ‘long tail’ of low-quality websites to return higher-quality results where the algorithm might not have been able to make an assessment before. The impact of these new signals is smaller in scope than the original change: about 2% of U.S. queries are affected by a reasonable amount, compared with almost 12% of U.S. queries for the original change.” (emphasis added)
That is a big difference, but it will be interesting to see how some of the sites reported to have lost the most from Panda, fare over time.
Do you think Panda has improved Google’s search quality? Share your thoughts.