TRAFFIC
Figure 6: Negative clickthrough-rate experiment
Each year, they do deep analysis of their
data set to come up with the top ranking
factors for the year.
In 2014, they suggested click-through
rate had the highest correlation of any
signal with rankings.
It’s important to remember that they are
not saying correlation equals causation, i.e.
just because it correlates that means it’s the
reason for something happening. However
if you add up the other evidence, it’s clear
that click-through rate and subsequent
engagement are huge signals.
Conclusions on engagement rank
To me, it’s obvious this is a massively
important signal that no-one seems really
to be talking about. I find that weird!
Anyhow, I think Google are very clever
at understanding what a real or a fake
click looks like. From these experiments,
it seems that they understand where each
click has come from, because obviously
they’re looking at a user’s IP address and
geo-locating it.
Google must also have other systems and
processes in place for catching fake click
through manipulation. Why? Because of
pay-per-click fraud. If Google allowed PPC
fraud to get out of hand, advertisers would
lose trust in them and therefore Google
would lose massive amounts of ad revenue.
If you take this extremely well-developed
and understood technology and apply it
to organic search results, and you have
something that is pretty rock solid as a
protection against manipulation.
This may explain why Terry Kyle’s
experiments went from reasonably
successful to producing negative results.
Perhaps Google spotted his fake crowd
sourced clicks?
Figure 7: Searchmetrics ranking correlations 2014
20
iGB Affiliate Issue 50 APR/MAY 2015
There is another factor to consider.
Websites have lots and lots of keywords in
Google’s index. If Google is tracking all
of these keywords, it must be very hard to
fake a complete keyword set for a particular
website. That means Google can aggregate
the click-through rate and engagement data
for a whole site, and therefore give it more
trust, leading to higher rankings for the site
as a whole.
Now what?
Depending on your approach to SEO, there
is the high road or the low road. The high
road means sorting out engagement on
your website; the low road is coming up
with ways of manipulating the system to
your advantage. Let’s look at each one of
these in more detail.
● The high road: making your site
more engaging.
Start with the big picture: your mission as a
website and how you deliver your promise
to the user. A website doesn’t have to be
pretty in order to deliver on its mission.
There are affiliate sites which look horrible.
However, they deliver on their mission.
Perhaps they list more bonus codes than
anybody else, or maybe offer more in-depth
information on a particular bookmaker
than anybody else.
The point is that it’s not how pretty a
site is, it’s how well it delivers on what it’s
supposed to. Google Quality Rater guidelines
talk about this quite a lot, where they talk
about a site having satisfying content.
The important question is, if your site
makes a promise to a user, will that user be
satisfied with what they see?
Overall, there’s a simple rule, which is
that a website (generally) makes a promise
to be good at something. If the promise you
make is kept, then great. If not, Google
will probably find out.
Let’s say that you’ve got a site which
does keep its promises, the next thing is to