Google’s “Random Algo” Explained

  •   10th April 2018
  •   : Blog
what is the Google random algo

So what is Google’s “random algo” or “rank modifier” that it also goes by? well in the video at the end of this page (HIGHLY recommend viewing on this subject, if you prefer to watch than read, scroll down now) we go into depth on what it is and I highly recommend watching the video, but in essence, its designed to try and fool / worry the unknowing SEO provider and / or SEO client. Basically, if Google spots anything that could be SEO, they roll a dice on three metrics…

  1. positive or negative
  2. time frame
  3. position adjustment

Here’s some examples based on different scenarios…

An SEO gains a brilliant backlink to the site, the anchor text is natural, the content of the site and theme related to the target site, high quality original content on the page that links to the target site, etc. Google spots this, and this new link gives the target page a boost in trust, authority and power to take the target keyword from position 8 on page 1, to position 5, but the random algo kicks in, and rolls the dice on the three metrics and picks…

  1. positive or negative = negative
  2. time frame = 3 weeks
  3. position adjustment = 10

So instead of jumping to position 5, which this backlink has really given, the random algo pushes their 8 position down to position 18 for 3 weeks. If the link is still there in 3 weeks, this random algo is removed and they gain what they should of, and jump to position 5. If you didn’t know about the random algo, you may panic that this link has sunk the ranking from position 8 to 18 and remove it and never wait out the random algo and now that position 5 is lost, AND, you may of just tipped off Google you’re doing SEO because if this link wasn’t for SEO gain, why was it removed soon after a dip in rankings.

what is the Google random algo
Here’s another example….

An unknowing SEO sees a great gig on fiverr where they can get 100,000 backlinks to their site for just $5, they’ve read that backlinks are powerful for SEO so they order 10 of these for $50, so 500,000 backlinks. These are made and point links to the target site all using the same anchor text, and all the links are from spammy websites. At the moment they’re ranking position 27 for their target keyword and they fire the 500,000 links to the target page, Google spots this and these link reduce trust and should take their position 27 to position 90 BUT, the random algo fires up yet again…

  1. positive or negative = positive
  2. time frame = 2 months
  3. position adjustment = 7

So their position 27 moves up to 20, the SEO thinks “wow this is brilliant” and buys even more so now there’s another 500,000 spammy links (upon seeing this Google would fire off another random algo layered on top of the first). After 2 months the random algo falls off, and the page drops to position 90. Google fooled the SEO into thinking they were good links so they would do more.

Now this is where it gets “fun”. Backlinks are not the only thing to fire off a random algo, other metrics such as: social signals, page / html / site changes, different user metrics, etc and to make it even more complex, every one of these gets its own random algo. For example, lets take just one metric, backlinks, and take the situation where a backlink is added daily to link to a target site, so every day a random algo fires, then daily the random algos could be something like…

day 1 = positive / 8 days / 3 positions

day 2 = positive / 32 days / 1 position

day 3 = negative / 2 days / 76 positions

…. etc – so when this is happening on every backlink found and processed, every social signal, every site change, the somewhat “simple” idea Google has come up with here is just beautiful (from their side at least lol) in making SEO testing take a lot longer (as you need to stop SEO tests and then wait it out for the random algos to “fall off”) and makes it easier for SEO to panic and shoot themselves in the foot, and finally it makes it harder for clients buying SEO services to see what is working / isn’t. For example, a client may rank at position 20 for a keyword, start with a great SEO provider who does amazing work and out of “bad luck” with the random algos, actually see things get worse for a time. If the SEO and client aren’t a) aware of the random algo and b) able to test and verify rank drops aren’t due to other reasons (as you can’t and shouldnt just write off a rank drop as “probably the random algo” as there could be other reasons such as site issues, page issues, content issue, negative seo etc) then the client may think the SEO service isn’t working and stop (which is what Google wants).

If you’d like to get even more information about it, why not check out Googles patent on this at https://patents.google.com/patent/US20090171943

Any questions on the above please feel free to shoot me an email at scott@scott.services, I’ve love to hear from you, and all the best with your SEO, Scott 🙂

This post is also available on our Google+ page at https://plus.google.com/u/0/+ScottDSmith

Tags: , , , , , , , , , , , , , , , , , , , , , , , , , , , ,