Semantics - A Larger Role in the new Google Algorithm


or is the latest update with Google proof of genetic relationships in full force?

I've been quite open in my thoughts about knoll peaks and how use of deoxyribonucleic acid is being used to identify web sites and relationships between them. I've also shown how that same thoughts process when used with social demographics and neck colour can be used to help tailor results for a search engine user. Follow the title link for more...

But whilst lots of people initially thought that the latest update was to do with words, phrases and the relationships between them. So called Semantics, and LSI/A, on further analysis I believe it is not.

Yup. Words and phrases matter. The over use to equivalent words links in this thread (I hope) shows there is a million and one ways to say the same thing, but I believe that 4 questions can categorically be answered by this update.

  • Was this update a huge algo change?
  • Is the Sandbox standalone or a side effect of another algo change?
  • Is LSI/A the cause for the update?
  • Is Hilltop (in altered format) being used?

I'll do my best to answer each in turn.

Was this update a huge algo change?

In my (not so) humble opinion no. Previous algorithmic changes have meant that the recent update by Google was but one of the new style of dances. She may have been a disco fan previously but now she prefers the waltz

It's a slow dance, graceful to look at but takes a long time to undertake. You may look at a Waltz as boring but others, will look at it and see that it is a long thought out well prepared and elegant experience to behold.

Google now undertake infrequent updates. She may start doing the foxtrot in the future as her computing power increases to undertake dancing at will.

Is the Sandbox standalone or a side effect of another algo change?

So many sites have been released from the Sandbox on this update and after analysing them it becomes apparent what the differences are. Some people have said WHOIS information is to blame, others say content, others again say IP addresses of hosting.

I say all of them!

I believe (and have for some time) that the Sandbox is a side effect of Hilltopesque style algorithms and that removal from the sandbox happens when the Hilltop algo is happy. It may take a week or may take many many months, but it seems to me that once you comply with the requirements, and then the algorithm has computed that you have complied then you'll get out of your sticky swamp and jump into the SERPs.

Is LSI/A the cause for the update?

Nope, but LSI/A will appear to have an effect in the same way this lil post could have lots of varying words that when put together have a similar meaning.

Is Hilltop (in altered format) being used?

In my (not so) humble opinion, Yup!


Good show on LSI/A on right now - they're about halfway through and my brain is bleeding - time for bed...


I'd definately say the hilltop has been tweaked, and with good effect. Smaller sites linked from authorities, which have deserved a spot in the serps/algo are starting to appear. I like it, and it's a definate improvement for some of the cleaner serps I track.