Home – New Forums Marketing mastery The Best Google SEO Video in the Last 2 Years?

  • This topic is empty.
Viewing 12 posts - 1 through 12 (of 12 total)
  • Author
    Posts
  • #994282
    JohnW
    Member
    • Total posts: 2,642
    Up
    0
    ::

    Hi All,
    This is a continuation and expansion of a discussion that started but went way off topic on “The #1 thing people do wrong with Social Media Marketing…

    We all want more, relevant generic search engine referrals to our websites. A big barrier is understanding how Google works. Culling fact from fiction, myths and rumours is tough when most SEO content is published by people trying to sell their services/products rather than sorting out the facts first.

    A few particularly contentious SEO issues at the moment are…

    Does Google use these signals in its ranking algorithm:

    a. Are “clicks” on links in its search results part of the algo?
    b. “Engagement” figures like session time, bounce rates, pageviews, page depth visits
    c. If its in a patent is it in the algo?
    d. “Engagement” factors in social media (Eg: Facebook, Twitter, G+)?

    From my research, the short answer is “NO” to each of them, according to various Google’s spokespersons.

    The reasons for each “NO”, involve a negative answer to one or more of the following five questions that G posses for every algo change proposed by its engineers:

    • Do they improve search results?
    • Are they reliable signals?
    • Can they be “gamed”?
    • Are they widely adopted on websites?
    • Can Google measure them?

    IMHO, the best SEO video in the last 2 years answers most of the above in detail and with logic.

    A. G Answers to a-c

    They are all in this video.

    23 Mar 16: The Video: Google Q&A+ #March
    It’s over an hour long and tragically I’ve already watched it 4 times in 3 days.

    There are some remarkable aspects to this video. Eg:

    a. The Q&A panel includes some of the best known SEO consultants in the US and UK

    b. A couple of the panalists are/were originators of certain of the algo ranking signal rumours under discussion here

    c. We are told what are G’s top three ranking signals, (but they forgot to start with “it depends”)

    d. G’s spokesperson gives a lot of info about G’s basic principles – when and why they may or may not use algo signals

    e. Right at the end of the video we are given the most important SEO tactic clue for small businesses in Australia (IMHO).

    The G spokesperson facing questions from the SEO panel of experts is Andrey Lipattsev.

    The panel includes Ammon Johns UK, Eric Enge USA, Rand Fishkin USA.

    To save you time, I’ve culled the video time points for a few of the “most important” comments (IMHO)…

    1. Q: Are click-through rates part of the ranking signal? (10 min 24 secs)

    Ans: No, it is not part of the signal.

    Reasons:

    • It is an unreliable signal.
    • It is too easy to “game”.
    • Result quality would be compromised.

    Discussed again in the video at 22 min 02 secs.

    If CTR used, you would end up with a small set of results (I.e. Poor quality search answers).

    2. Statement: (16 min 50 sec) The answer to every SEO question should start with “it depends”.

    3. Q: Are “engagement” parameters used? (Like Analytics CTR, bounce rate, visit page depth, ave. pages viewed) (19 mins 20 secs)

    Ans: No,

    Reasons:

    • “it’s impossible to measure”.
    • We can’t and don’t use these.
    • “We don’t have the technology to record the data”.
    • Our system not widely enough used therefore data would be unreliable.

    4. Q: If its in a patent, is it used in the algo? (25 min 50 secs)

    Ans: No.

    Reason:

    It does not matter how many patents have been filed, if a proposed algo change can’t pass the questions about improving results (etc.), it will not be implemented.

    5. Q: RankBrain and the 3 most important ranking signals. (28 mins 10 secs)

    They forgot to preface this queston with “it all depends”…

    Top three stated as: Content and links to your site – in no order and RankBrain

    Discussing the algo: If you look at a slew of search results, certain elements (of the algo) come into play differently.

    We SEOs don’t know how RankBrain works, so no one can claim to offer SEO services that include it.

    Other issues discussed like voice vs typed search queries. As these query terms change for typed vs voice actuation AND with how RankBrain may come into play, so the search results may vary.

    6. Discussion of limited content for search results (43 min 20 secs)

    These experts don’t seem to know much about how G algos work outside USA – various questions about this issue discussed.

    In small markets, not enough people publishing good content (47 min 52 secs)

    Example of Arabic countries equal 5-6% of world population. Content in G many orders less than this.

    Note that Aust’s Internet access is about the same as the Arabic figure. Small business owners should ask prospective SEO’s how this may impact on strategies and tactics for their sites.

    B. “Engagement” in social media (Eg: Facebook, Twitter, G+)

    I’m not aware of any G statements that contradict this article:

    May 14: Social Signals Are Not An SEO Ranking Factor

    The bottom line is NO. G does not treat these social media pages any differently to any other web pages. Subsequent G discussions indicate that they may be important to temporarily trigger the “query deserves freshness” algo component. As to their value as external links, they can’t help because they employ “no follow”, links with a minor G+ exception.

    The article runs us through the history of G’s statements about social media links as algo signals with links to various videos of relevance.

    It refers to the articles published by Moz and Search Metrics that possibly started the rumour. (I’m intrigued at how often the SEO Tool sellers seem to “find/start” these correlation but not causation statements.)

    C. How G “thinks”.

    I’ve always found that if you understand the problems that G faces in identifying and delivering relevant content and the basic principles it uses to address the issues, you are likely to arrive at pretty accurate conclusions. For that reason, I highly recommend this video…

    20 Jan 16: Google Q&A+ (January)
    Panelists: Andrey Lipattsev +Bill Slawski +Ammon Johns +Eric Enge

    It includes:

    • How much do people misinterpret what a SE is?
    • Mobile search: Good and bad sites
    • Mobile sites: The problems developers and marketers face

    Sorry to hit folk with so much heavy duty SE issues. Hope some of you find the refs. useful.

    I’d love your comments and any items you have to update this post.
    Regs,
    JohnW

    #1195870
    Aidan
    Member
    • Total posts: 1,125
    Up
    0
    ::

    Great video John, thanks for bringing it to my attention, it reinforces my points in some recent discussions about the engagement factors influence (or actually lack of influence) on rankings.

    Sadly there is too much nonsense around, I know of one ‘real smart’ guy who built his pages so the user has to click to an unnecessary second page to get all the info he might need on he services offered and it was designed that way so Google would see less bounce rate – holy schmoley – when did Google ever suggest the user experience should be made cumbersome?!

    #1195871
    JohnW
    Member
    • Total posts: 2,642
    Up
    0
    ::

    Hi Aidan,
    Then there are the huge volume of mobile phone searchers who simply want a phone number or address…

    A phone number should be on every page. Five seconds from landing, the searcher should be gone and phoning you. One page, 100% bounce and low session duration are good site quality signals for these search phrases.

    It is totally illogical to use session duration, bounce rates and pageviews to indicate any sort of search result quality in this smartphone search era. If a SE did use them, mobile site accesses would cause web pages to drop in ranking results.

    If they are going to be consistent, the “engagement is a ranking factor” believers should be barring access to their sites by mobile phone users.

    Here is what Google has shared with us…

    Micro-Moments

    Mobile has forever changed what we expect of brands. It’s fractured the consumer journey into hundreds of real-time, intent-driven micro-moments.”

    Regs,
    JohnW

    #1195872
    MD Clean
    Participant
    • Total posts: 308
    Up
    0
    ::

    To sum up, Google wants relevant information so it can deliver relevant search results and customers want relevant information written in a way that convinces them you are the right fit to solve their problem.

    I secretly theorize that Google does not count social signals because social platforms are it’s direct and indirect competition and if social signals were counted, their competition would grow stronger while Google itself would become weaker.

    #1195873
    JohnW
    Member
    • Total posts: 2,642
    Up
    0
    ::
    MD Clean, post: 230648, member: 58759 wrote:
    To sum up, Google wants relevant information so it can deliver relevant search results and customers want relevant information written in a way that convinces them you are the right fit to solve their problem.

    I secretly theorize that Google does not count social signals because social platforms are it’s direct and indirect competition and if social signals were counted, their competition would grow stronger while Google itself would become weaker.
    Hi Paul,
    By social signals are you referring to “likes” and “followers”?

    If so, I think it is simpler than that…

    Google knows that social media “likes” and “followers” are meaningless and highly gamed parameters that would therefore detract from search results accuracy.

    Then there is the issue of how universally a social medium platform is used.

    I believe the regular usage of Twitter and G+ in Aust is only around 10%. Too low for G to use as universally representative signals. I doubt that even Facebook would pass the universally used test. (NB: One of the reasons stated in the first video for not using signals from Analytics was that it was not widely used enough.)

    If referring to social media pages being accessed from G search results, no problem they are there but you usually have to enter an unusual search term or use the social medium name in the search query to find them.

    That goes to the nature of social media pages, their content and accessibility. Typically, their content rambles over all manner of unrelated topics. Their security settings frequently hide their content from SE access.

    If people are searching for topic specific answers, who wants to waste time scrolling a million kms down a Fb or Twitter page on the off-chance it contains a buried answer?
    Regs,
    JohnW

    #1195874
    MD Clean
    Participant
    • Total posts: 308
    Up
    0
    ::

    Hi John,

    No, I am really looking at the bigger picture – there are some laser focused FB, Instagram, Pintrest etc pages out there that do not have the same weight as a blog post on a website from a Google POV.

    #1195875
    JohnW
    Member
    • Total posts: 2,642
    Up
    0
    ::
    MD Clean, post: 230658, member: 58759 wrote:
    Hi John,

    No, I am really looking at the bigger picture – there are some laser focused FB, Instagram, Pintrest etc pages out there that do not have the same weight as a blog post on a website from a Google POV.
    Hi Bert,
    Humans may see social media as “laser focused” but SEs can’t because many of them are inherently incapable of delivering various “importance” signals that SEs use to “weigh” web pages/sites.

    SEs are designed to assess “weight” based on a page’s relevance to a set of search words. Here are a few ranking signals that are problematic for social media:

    • Search terms in a domain name (You have to use the social medium’s domain name)
    • Search topic specific information pages. (You essentially only get one. That means only one page title, URL, H1 tag, etc. Therefore, your social media page cannot be relevant to multiple search terms.)
    • Lack of words. This limitation kills Instagram or Pintrest in SERPs
    • Content may be published in code formats that SEs may not read. (Eg JavaScript and Ajax)

    Any social medium that is image based can never achieve a SE assessed high “weight” for anything but most unusual search terms.

    Eg: Instagram Help Centre

    “Your profile will be indexed on the web, but search engines aren’t allowed to index your photos.”

    Call it a limitation of SEs if you want but it is not driven by subversive intent. Social media decided to publish their pages in the way they do long after SE limitations were known to them.
    Regs,
    JohnW

    #1195876
    Karen Francis
    Member
    • Total posts: 67
    Up
    0
    ::

    Hey all,

    I was offline when this went live and am only just catching up on the news, it’s interesting the takes from a variety of SEO specialists about how this answers questions.

    Here’s the full transcript if anyone would prefer to read rather than watch the video.

    Search Engine land has now listed the top 3 search ranking factors as:
    1 & 2: Links & Content
    3: RankBrain

    I like that Rand talks about the experiments he has done inviting users to click on items from a search and sees the impact it makes with a quick rise, then a fall in the following days. I’ve participated in some of his experiments and seen the results live. He has a whiteboard Friday about how sites can go to a higher “new normal” in trials he has done.

    In the discussion about CTR as a ranking signal, from the transcript:
    “the point of that is that it seems to me that if you’re using click-through rate as a main measurement of other ranking factors, to better measure search quality, it doesn’t really matter to me that much whether it’s a direct ranking factor or an indirect ranking factor. It still is used in evaluating search quality.

    Also from a link to Search Engine Watch embedded in the transcript:
    “User engagement metrics are not a solid ranking signal, but you shouldn’t rule them out. They signal to Google that your website is serving the answers or solution to what your users are searching for, making them an invaluable metric, moreso than a link.”

    Personally, I wouldn’t regard CTR as more important than a link for ranking success, but I would regard it as a metric that requires improvement for site visibility and success. Often it comes down to what are the competitors of the site doing, and which areas can be improved to have the site you are working on be positioned above them in the results.

    #1195877
    Aidan
    Member
    • Total posts: 1,125
    Up
    0
    ::

    Hey Karen,

    Grab coffee and read or listen again to Andrey (Google)! Really listen to what he is saying.

    For one instance:

    “The disadvantages that I’ve most often seen described for this approach on a clear, pure ranking factor basis is that we’d need to have broad enough and reliable enough data about bounce rates, click-through rates, depth of view for the vast majority of pages and the vast majority of websites everywhere, in order to be able to make meaningful comparisons all the time. That is impossible…”

    The other guys are NOT Google and they freely admit the CTR experiments don’t work consistently and when they do they are very temporary in their affect.

    Andrey even says he’d like to watch those tests happen to find out why Google is being temporarily thrown by it!

    It was also confirmed very clearly by Gary Iles (Google) at SMX stating:

    “Google uses clicks made in the search results in two different ways – for evaluation and for experimentation – but not for ranking.”

    (Which echoes what Paul Haar (Google) says about their testing environment.)

    Iles also commented clicks do influence personalized search results, but CTR is too easily manipulated to be used for ranking purposes.

    There is precisely ZERO point in us arguing otherwise.

    Cheers

    #1195878
    Karen Francis
    Member
    • Total posts: 67
    Up
    0
    ::

    Thanks Aidan, I’ll make some time in the next few weeks to go over it again. Trust me to take a digital break when Google releases something useful :)

    #1195879
    JohnW
    Member
    • Total posts: 2,642
    Up
    0
    ::

    Hi Karen,
    The other video is also a very valuable source of Google info.

    20 Jan 16: Google Q&A+ (January)

    Don’t ignore it…
    Regs,
    JohnW

    #1195880
    Snakeman
    Participant
    • Total posts: 329
    Up
    0
    ::

    “A phone number should be on every page. Five seconds from landing, the searcher should be gone and phoning you. One page, 100% bounce and low session duration are good site quality signals for these search phrases.”
    I will plus one that comment and this is independent of the immediate SEO parameters as well.
    All the best

Viewing 12 posts - 1 through 12 (of 12 total)
  • You must be logged in to reply to this topic.