5 ms·

> Google hasn't made much progress getting away from "page rank". Is that so? I thought conventional wisdom is that they now have much stronger signals about h

by baddox 6y ago

> Google hasn't made much progress getting away from "page rank". Is that so? I thought conventional wisdom is that they now have much stronger signals about how well results satisfy queries, from Google Analytics and Chrome.

threeseed 6y ago
For niche searches they might not simply have enough statistically significant data for "time on page" to be a useful metric.
tyingq 6y ago
It is so, but it is largely now either "brand authority" or "backlinks" now. They aren't as sophisticated as they want you to believe.
- solarkraft 6y ago
  Your example seems a lot less simple than you seem to believe. You created lots of high value interest instead of just getting some page to post links for you. In a way Google is correctly accessing real value here - you didn't cheat.
  - bigiain 6y ago
    Which is exactly what Matt Cutts and Samantha Fox were saying 20 odd years ago. Provide genuine value to people making web searches. Google will keep updating their algorithms to make searches result in better results for people. “Tricks” can sometimes make you rank quickly, but building genuinely valuable content is the best long term strategy. (And I acknowledge that not everybody is looking for long term value from sites they build, which means they shouldn’t follow that advice. If you are trying to rank a site for a one off event, for example, where you really don’t care about search traffic in 12/18/60 months time, your site will end up looking a lot closer to SEO-spam de jour.)
    - tyingq 6y ago
      To be clear, I researched popular folks in the "Arduino, control a physical thing that is visually interesting in this space". I then sent several folks a free sample of a thing, and also rough code, that costs me $100 or so. And the gamble paid off. One of them ran off with it. It benefitted them pretty much zero as far as I know. I totally gamed the system. It was not natural in any sense. I got ~3000 sales, and maybe $10k of upside, which was disproportionately in my favor. Most of the magic was this person's fruitful efforts to take my rough idea into something popular. But I like to selfishly think nobody really "lost" in this little microcosm.
      - HeyLaughingBoy 6y ago
        But how is that gaming the system? Sending free samples for review has been a thing since there were advertisements.
        SyneRyder 6y ago
        Reviewers are meant to send back their review units. A lot of professional reviewers refuse free review units entirely and purchase the items themselves. Google has posted a warning on their Webmaster Blog about providing free review samples, they will take action against companies that do it: "Bloggers should use the nofollow tag on all such links because these links didn’t come about organically (i.e., the links wouldn’t exist if the company hadn’t offered to provide a free good or service in exchange for a link)." https://webmasters.googleblog.com/2016/03/best-practices-for-bloggers-reviewing.html https://webmasters.googleblog.com/2016/03/best-practices-for...
        bigiain 6y ago
        Sounds to me like what was mentioned by tyingq was described accurately: "Maybe I even skirted the rules a bit by sending him a free sample of my cool toy and some rough code samples. He took it well beyond what I could. I didn't ask for anything." Certainly seems easy enough to argue complies with: "send bloggers free products to review or give away in return for a mention in a blogpost" If you don't ask for mentions in blogs, that Webmaster Blog post doesn't technically apply, right? There's gray area in choosing to send free samples to people you are pretty sure will mention/link to you in a blog post, but where you don't ask them to (or place any expectation on them to). So "skirting the rules a bit" sounds about right to me. There weren't "review units", they were "free gifts". No request or expectation of a "review".
luckylion 6y ago
Yeah, some variation of page rank is still the number one signal. That's why subdomain/subdirectory leasing at media companies is such a hot deal right now: you take completely unrelated content (e.g. coupons with affiliate links), throw them on a subdomain of a news site (e.g. CNN) and then just watch your content rank that would have never ranked on a "normal" domain. That "buying links" is the only deadly sin next to cloaking is more strong evidence for Google primarily relying on it. You don't get penalized for abusing Microdata, stuffing Emjois into your Title + Description or writing barely readable keyword-stuffed SEO texts, but get links from a well-known link-selling-site and watch your -100 penalty wreck your traffic.
- tyingq 6y ago
  Good insight. My intuition tells me that "buying links" is still the fastest path to ranking well, provided you are stealthy enough.
XCSme 6y ago
> I thought conventional wisdom is that they now have much stronger signals about how well results satisfy queries, from Google Analytics and Chrome. Do you have any official source for this? I remember they clearly saying that Google Analytics data is not used and will not be used for search in any way.

by baddox 6y ago

threeseed 6y ago

For niche searches they might not simply have enough statistically significant data for "time on page" to be a useful metric.

tyingq 6y ago

It is so, but it is largely now either "brand authority" or "backlinks" now. They aren't as sophisticated as they want you to believe.

solarkraft 6y ago
Your example seems a lot less simple than you seem to believe. You created lots of high value interest instead of just getting some page to post links for you. In a way Google is correctly accessing real value here - you didn't cheat.
- bigiain 6y ago
  Which is exactly what Matt Cutts and Samantha Fox were saying 20 odd years ago. Provide genuine value to people making web searches. Google will keep updating their algorithms to make searches result in better results for people. “Tricks” can sometimes make you rank quickly, but building genuinely valuable content is the best long term strategy. (And I acknowledge that not everybody is looking for long term value from sites they build, which means they shouldn’t follow that advice. If you are trying to rank a site for a one off event, for example, where you really don’t care about search traffic in 12/18/60 months time, your site will end up looking a lot closer to SEO-spam de jour.)
  - tyingq 6y ago
    To be clear, I researched popular folks in the "Arduino, control a physical thing that is visually interesting in this space". I then sent several folks a free sample of a thing, and also rough code, that costs me $100 or so. And the gamble paid off. One of them ran off with it. It benefitted them pretty much zero as far as I know. I totally gamed the system. It was not natural in any sense. I got ~3000 sales, and maybe $10k of upside, which was disproportionately in my favor. Most of the magic was this person's fruitful efforts to take my rough idea into something popular. But I like to selfishly think nobody really "lost" in this little microcosm.
    - HeyLaughingBoy 6y ago
      But how is that gaming the system? Sending free samples for review has been a thing since there were advertisements.
      - SyneRyder 6y ago
        Reviewers are meant to send back their review units. A lot of professional reviewers refuse free review units entirely and purchase the items themselves. Google has posted a warning on their Webmaster Blog about providing free review samples, they will take action against companies that do it: "Bloggers should use the nofollow tag on all such links because these links didn’t come about organically (i.e., the links wouldn’t exist if the company hadn’t offered to provide a free good or service in exchange for a link)." https://webmasters.googleblog.com/2016/03/best-practices-for-bloggers-reviewing.html https://webmasters.googleblog.com/2016/03/best-practices-for...
        bigiain 6y ago
        Sounds to me like what was mentioned by tyingq was described accurately: "Maybe I even skirted the rules a bit by sending him a free sample of my cool toy and some rough code samples. He took it well beyond what I could. I didn't ask for anything." Certainly seems easy enough to argue complies with: "send bloggers free products to review or give away in return for a mention in a blogpost" If you don't ask for mentions in blogs, that Webmaster Blog post doesn't technically apply, right? There's gray area in choosing to send free samples to people you are pretty sure will mention/link to you in a blog post, but where you don't ask them to (or place any expectation on them to). So "skirting the rules a bit" sounds about right to me. There weren't "review units", they were "free gifts". No request or expectation of a "review".

luckylion 6y ago

Yeah, some variation of page rank is still the number one signal. That's why subdomain/subdirectory leasing at media companies is such a hot deal right now: you take completely unrelated content (e.g. coupons with affiliate links), throw them on a subdomain of a news site (e.g. CNN) and then just watch your content rank that would have never ranked on a "normal" domain. That "buying links" is the only deadly sin next to cloaking is more strong evidence for Google primarily relying on it. You don't get penalized for abusing Microdata, stuffing Emjois into your Title + Description or writing barely readable keyword-stuffed SEO texts, but get links from a well-known link-selling-site and watch your -100 penalty wreck your traffic.

tyingq 6y ago
Good insight. My intuition tells me that "buying links" is still the fastest path to ranking well, provided you are stealthy enough.

XCSme 6y ago

> I thought conventional wisdom is that they now have much stronger signals about how well results satisfy queries, from Google Analytics and Chrome. Do you have any official source for this? I remember they clearly saying that Google Analytics data is not used and will not be used for search in any way.