The head of Google’s webspam prevention team, Matt Cutts was kind enough to conduct an impromptu 45 minute live Google questions and answers video chat on YouTube in September which I decided was so chalk full of content it would make a great article; it just took me a lot longer than I had expected to get done! Anyway, I tie it up with a fun ode to a future Movember Matt.
Also you can listen to some discussion on this Q&A on the Oct 3 2011 episode of SEO 101 on WebmasterRadio.FM which I co-host with fellow SEO veteran John Carcutt; you can find the show on iTunes here.
The Prelude to the Google Q&A
Before he did the Q&A Matt wanted to cover a few points which I outline below along with direct links to the most pertinent section of the video where he explains each point; so you can hear it in his own words (and I don’t have to write them all down). After these points you will find his Q&A where my short-hand should give you the answers you need; if not you always have the video!
Pagination: Matt mentions the use of rel=next and rel=previous to aid in the improved indexation of paginated pages. This is not a light topic so he understandably does not get into great detail but watch Matt’s short explanation on pagination here.
Reconsideration Requests: Matt discusses an article by Tiffany Oberoi and Michael Wyszomierski from the Google Search Quality Team called “Reconsideration requests get more transparent“. As the title aptly describes the Google will now be much more open about whether your site has been penalized or not should you submit a reconsideration request. Here is Matt’s description of the changes to reconsideration in his own words.
Requests for Crazy Ideas: if you have “crazy ideas” for how to search which Google has not done they invite you to add your ideas on Matt’s blog where he posted a request for those ideas recently: What cool new websearch ideas should Google launch in 2012?
Matt Cutts Answers Google Questions
NOTE: each of the linked titles below will take you directly to the relevant segment of the video – in case you want to hear it from Matt directly. Otherwise I have done my best to paraphrase his answers; a shorter version of the actual transcription.
“Right now we are still studying it” Matt says. He then notes they are using the “block” data in search results (they rolled this into the Panda update) but only if they are coming from clearly trustworthy individuals; those with a strong history in Google’s records. In terms of the +1’s they are still looking into the affect they can have and “its definitely promising, and we’re excited about it and we hope to find a good way to use that” but he stresses it is not something they want to implement it just because they can.
Google doesn’t really have a way to ensure your site does not show up for a particular word. Why? Well, there is no easy way to do this, he explains, because “in theory that could be a way you could silence critics or not show up for negative stuff.” He goes on to say they don’t want to deny the benefit of opposing viewpoints in their results which this could negatively affect.
Matt recommends visiting the Google Webmaster Forum and asking that question so someone can look into it. He also hints at there being a form of filter which will not allow just anyone to show up because it is suppose to be a positive indicator in results vs a negative; in other words spammers could (and in my opinion would) abuse this if it were wide open.
Should I create a native application or a mobile web site to work with all of the mobile platforms out there? (the original question was just “Native app vs Mobile web”)
Matt found this to be a tough question because native apps can offer a more fluid experience but it can also be overwhelming with the myriad of platforms to create apps for. He says, however, that HTML 5 shows a lot of promise. In fact, he says a lot of people are creating their product in HTML 5 and when they run into special functions which are better in an app format they will then go that route.
Good suggestion, he answers, but he has no idea but he would suggest that since the Google Plus team is implementing changes quite quickly.
It is too soon to tell since the feature has only just been implemented and the Google Flights team is still working on the implementation. For example, they are still lacking specific functionality that is needed to perfect the US implementation such as being able to search for first class flights or other such custom search options. When these features are worked out he expects they will roll it out quite quickly.
They won’t have any affect on SEO. The +1 on ads will show others who have +1’d it but that is all.
“I think Panda would be independent of whether any of those sorts of technical queries have gotten worse.” That said, he explains if you have some way to show a past and present version of search to show how it has degraded they are very interested.
Matt chuckles and states he shut down his Facebook account a while ago so he no longer uses it. That said, he has enjoyed what he has used of Google Plus in particular the Hangouts and the benefits of the API.
He does have his name on a patent which refers to using the age of anchors and domains to assist with finding higher quality content. That said, he wouldn’t advise going overboard on that. You still need fresh information because old information can get really stale, so they do qualify whether the search query deserves a higher level of freshness. He also states that older, more established domains may have a lot more links pointing to them which can increase the illusion that domain age is more important than it actually is.
When are the next round of Panda Updates expected? (original q: when is Panda 2.5 / 3.0 going to run?)
Matt first states he does not want to get into what number the update is since it is now a rolling update; it is being updated on a regular basis like all Google algorithm tweaks. He then mentions an article by Google’s Amit Singhal which lists ways to increase the quality of your website to avoid Panda Penalties (“More guidance on building high quality sites“). Further to that he discusses one element which is the number of advertisements you have on a page and above the fold; too many and your site/page is likely to be negatively affected (even though he confusingly says it is not a signal Panda is watching). After all, Google is all about search experience and they do not want to send traffic to sites that oversaturate their content with advertisements. At any rate, Matt states the volume and placement of advertisements is a signal they are watching.
Matt doesn’t know but he does know they are working on it and that it is “certainly a priority”.
Matt explains there is an engineer at Google named Panda who it was named after.
“If you say everything is a high priority then we might not take that as seriously.” But if the bulk of pages listed are of a normal priority and a few things are of a higher priority it will have a greater chance of having an affect. That said, he suggests not spending too much time on this if you have a blog or a content management system because you can use Ping systems to tell Google when pages are new.
Google will only use the Meta Description if it is appropriate for the term searched or perceived search intention. As a result, there are some occasions where Google will write its own description for a ranked page to provide insight into why it was ranked for the particular query.
This is not immediately anything to worry about, however, it could be your site has canonicalization issues whereby using the Canonical Tag on specific pages would consolidate the PR onto the best pages. It could also be your site has a spider trap where the search engine spider is finding infinite loops of the same data which results in a great deal of spider activity leading to little benefit for Google or your website.
Hangouts are a Google Plus feature which, up until yesterday, only allowed 10 people to connect on a single group discussion with video. Recently, however, Google changed Hangouts to allow an unlimited amount of people to connect in the chat area of this hangout and participate (without video). This has expanded the use of Hangouts greatly to allow one-on-one interaction with celebrities.
In Matt’s opinion rel=author is good on any website with unique content because it allows Google to begin to build trust with the author and identify the author to searchers. He ends by saying he sees absolutely no downside to using the rel=author for any website and “overtime it may even have an impact on our search results.”
He doesn’t know exactly how it will interact with rel=author and thinks they should “mostly be separate.”
He would love to do this a lot more frequently and feels if he can continue to do these and have them not too stressful then it would be more manageable. He notes he misses Danny Sullivan’s old daily searchcast and maybe he could have some fun and do a show or two with Danny or Bryan White… he will see how it goes.
Do a DMCA complaint. Do a spam report; especially if they are an actual spammer who is scraping content completely. Also, when you publish content be sure to be the website that PINGs Google first (see earlier note about PUBSUBHUBUB and other Google-sanctioned Ping services to let Google and other sites know when you have published the content. This will increase your odds of being considered the original publisher of a piece of content.
— from a later followup to the above scraper question where someone wanted more info on the Ping applications to use: Ping-o-Matic is built into WordPress and Feedburner users (a Google product) also pings Google and other sources upon posting new content. He also mentioned PUBSUBHUBBUB again and described its process.
Yes, no problem. Anything more than 4 and you will definitely not have a good result. Also, make sure that chained 301 redirects are only 301s and not inter-mixed with 302 redirects.
Matt laughed and admitted they have a history of naming things after engineers and said he actually knows there is a programmer at Google called Orkut and has seen him around. Then he states that Orkut.com (another social network that Google owns) is on the way out but continues to be quite popular in Brazil and to a lesser degree in India. He ended the answer vaguely stating he thinks it is fantastic it is still being enjoyed… but he states nothing about what will happen to it (anytime soon).
“In general it is every 3 or 4 months.” He says Google literally has a script that automatically runs every 3 or 4 months to update the PageRank that the public sees. That said, the calculations and the human aspect (where they ensure links flagged as purchased are not counted) that make up PageRank are constantly running and updated.
You can use rel-author in the news industry. In fact, the New York Times has already implemented rel=author and he expects other organizations to follow suit – if they have not already. He states it is natural for Google to want the news industry to do well because it helps people spend more time online, use search (i.e. Google) more and want to come back and use it again. He said the news organizations are talking about other options with Google as well but there is no certainty when they will be released.
Go to http://www.google.com/jobs/ and apply there around January when the Summer internship discovery process begins.
“I don’t know.” He is sure there is an option to remove a page cache after a certain period of time using a tag but he isn’t certain of the tag used to do that.
He says this is a complex question with a few layers to the answer and he will try to do a full video on this. In general, Google tries to find and deliver the best content regardless of where it is found. He followed this with stating his prior position was that Google didn’t behave different no matter the TLD used, however, that is not quite as true now. He states it is a more of a nuanced situation now, where if a particular TLD is often spammy it could affect how your search results rank if you are using that TLD. Ultimately, whether you buy your company name under a new TLD is up to you but Google will have to see how each of the new registrars respect copyright/trademarks before he can reliably state they are safe to buy for the long term. He finishes up by restating that Google’s mandate is to show good content no matter the domain.
From here on the show just gets fun and ties up with a few humorous stories and a Movember promise!
>> Matt takes a fun break to, on request, show the cool Android cookie jar shown in the background of his video. He then talks about his Porn Cookies and what they mean – this is some really funny Google history well worth listening to (this one is the best)
How will Matt Cutts look with a moustache? Well I just had to find out (here is a link if the photo doesn’t show up)
Ross Dunn, CEO, StepForth Web Marketing Inc.
Increasing the bottom line online for businesses since 1997
Connect with me on Google Plus!