“This article was first published on our private email list on the 16th of September 2013, 2 weeks before Hummingbird was officially launched, I am posting it now on our blog for posterity and so I can link out to it in future articles”.
There is a new algorithm that has just been named and it is claimed to
be the biggest redesign of the algorithm since it was first launched
in 2001. Its called Humminbird:
“Google told us the name come from being “precise and fast.”
Now, if you look back to the email I put out on the 16th of September
with the subject “another update and more confusion” you will see that
I predicted a NEW algo had been launched and talked about what I
noticed and many of the details I suggested we’re eerily 100% right.
The following is printed again from that email I sent on the 16th
Well It’s clear now G are playing with the dials and on the 12th,
there was another mild update, what they are doing is somewhat of an
enigma from the chatter out there but this week we’ve seen some new
light on the situation.
On the 4th September whatever they did seemed odd and possibly wrong,
I.e. not one of their best or clearest updates, from then on it looks
like G rolled that update back on the 12th, we’ve certainly seen
results that show that clearly (one site we monitor was at the top
along with 4 others – all exactly the same in link profile), then this
one site vanished for no real
reason and has now been brought back to its former positions.
It seems if anything that one site was effected by some sensitivity to
dupe content issues, but that is a bit vague, though that site did
suffer a technical problems brought on by a silly web developer
uploading dupe content briefly a few months ago until I noticed it.
The other issue is two fold now there are all these “in depth articles
taking up more real estate” and clearly there needs to be a big push
to understanding what formula is being used to choose those selected,
secondly there has certainly been another push in the middle between
the EMD and BRANDS, with lower quality domains ‘brands’ seemingly
being negatively effected, thogh its all link based, so companies that
have worked hard on
their brand citation in their link profile are doing better.
Ive long been saying the break down should be like this:
33% links BRAND,
33% links UNIQUE ANCHORS
33% links white noise, ‘click here’ etc.
The other thing is the speed of changes from new links being found to
actual changes in results.
Maybe they have just done some deep crawls but we are noticing 2
quicker events, one on the recoveries from penalties, previously in
years gone by they were slower, now we sometimes see positive results
in weeks from taking action (I’m not talking about doing
reconsideration requests but that is quicker too) but I’m talking
about removing toxic links, it seems if you can find anything toxic
immediately you will not be affected by any negative positions. The
other thing is speed at which they calculate Penguin (over
optimization of anchor texts), it seems like this is becoming constant
flux rather than a specific update rolled out every so often.
They are being too quiet during this period also. This is unlike them
in general, normally updates are called a bit later and confirmed but
so far there is absolute silence from the beast. This indicates
uncertainty in what they are playing with and not the typical updates
that we are used too. Something new and exotic maybe that is being
aimed at spammers but that they
have not quite perfected yet so are testing live and will roll out
officially in a few more weeks. I suspect something like that is
happening and we haven’t seen the last of this ‘thing’ yet.
I suspect this will be official in a few weeks and they will call
something like 5% of the SERPs will be affected but when in reality
its like 40% of the commercial terms being thought over which are