Google Analytics web stats : Fraudulent or just seriously wrong

Part 2 of Web page ranking and search engine optimization : SEO

Google Analytics web stats may be fraudulent, or they could be just seriously wrong by accident.  I will explain in this what I mean by that, based on some impressions I gained during close to two years of search engine optimization that is coming to the close.

No outsider can figure out why or how Google does things to come up with web- or site-stats that are as terribly wrong as those illustrated in the graphs shown next, a little farther down.  Keep in mind that, as I explained in the first part of this write-up, in July 2017 I began a thorough, systematic effort to do search engine optimization (SEO) at Dads&Things.  I used Yoast SEO Premium for that.  That is, as far as I can tell, the best and most popular software tool on the market for that (the best I ever had a chance to use).  It is tailored to fit Google’s requirements for SEO.

The data reflected in the subsequent graphs were collected by Google Analytics.  One application of that is installed at WordPress, the software that runs Dads&Things, the other at the host-service provider who hosts Dads&Things.  There is a third application of  Google Analytics that tracks the traffic statistics for Dads&Things.  That is the one I recently installed to monitor and observe traffic statistics for Dads&Things from my end.  I activated that on February 28, 2019.  Therefore, I don’t have enough data from that worth comparing to the data produced by the other two applications of Google Analytics that both produce such data, from all the way back in June 2017 until now.  The data from that third application of Google Analytics, for the same traffic to the same blog but from March 1st, 2019 and later, will be illustrated in a subsequent part of this write-up, when I get to that.

The following graphs pertain to the history of traffic statistics relating to one time interval, one blog, and one single stream of traffic that is look at from the perspectives of the host-service provider and the perspective of WordPress, the software application that runs the blog.  It stands to reason that two sets of data representing measurements of a single set of circumstances should be precisely identical.  They are not!  They are not even similar.  Here are the data:

How can things go so wrong?
Visitors to Dads&Things, June 2017 – February 2019,
as measured by Google Analytics,
used by WordPress and by the host-service provider

How can things go so wrong?
Page Views at Dads&Things, June 2017 – February 2019,
as measured by Google Analytics,
used by WordPress and by the host-service provider

How can things go so wrong?
Page Views per Visit at Dads&Things, June 2017 – February 2019,
as measured by Google Analytics,
used by WordPress and by the host-service provider

When I set out to work on my SEO effort, I knew that I would do a few or a lot of things wrong.  I had a lot to learn.  Yoast SEO was new for me.  The potential for making errors was great.  I hoped that I would make all of my errors right at the start.  That way I would not make them over and over, with every additional blog posting I tried to get up to the recommended standards demanded but not so well specified by Google.

SEO produced unexpected, grossly counterproductive outcomes

That was not what I had hoped for.  I had done a lot of work.  It was good work.  Some iterations happened, and some of those involved more than just a few pages.  I knew that I would learn by trial and error and that I would get better at it.  For that reason, I did not expect that all of the changes I made would immediately result in better page ranks, at least not at first.  That is what happened, but it was much worse than I had expected.  That was not because of the work I caused myself on account of the errors I made.  There were not that many of those.

The better I got at using Yoast SEO, the worse the new traffic statistics for the blog turned out to be.  Some people, I am sure, would have given up at that point, but I knew that there were many instances of problems on many of the 980 blog postings that needed fixing.  I had become comfortable with Yoast SEO and carried on, revised blog posting after blog posting.

I thought that at least I had a way to find and eradicate those errors, and that I had to get all of the postings to conform to a common standard.  Once done with that, I could then far more easily try something else to make the blog rank a bit better.  Chances were that, by first using Yoast SEO for standardizing all blog postings, I would reach the point of being able to make bulk changes requiring far less time.  I liked how nice everything worked and looked, and I most definitely liked working with Yoast SEO.  Doing it without Yoast SEO, I could not have done even close to as nice a job of SEO as I did.  I carried on, but I became more and more puzzled.

Alexa.com does web ranking.  I could see no rhyme or reason why the ranking trend for Dads&Things had absolutely no correlation to what I was doing.  It had no correlation to what Google Analytics for WordPress told me the traffic to the blog was doing.  Neither did it have any correlation to what Google Analytics, as used by the host-service provider, told me the traffic to the blog was doing. 

At first I thought that my SEO helped the web rank for the blog.  When I put new life into it, in May 2017, it had a web rank of 1.7 millionth, relative to other websites in the World.  As soon as I began posting to it again, the rank rose.  It kept rising, as I began to do SEO.  Then it fell again.  It did that a few times, eventually rose to about 450,000th place.  That impressed me.  The SEO I was doing had an enormous impact, but that was apparently just wishful thinking.  The rank fell again and continued its roller coaster ride.

The most curious thing was that the two different applications of Google Analytics produced so widely differing results, while they both measured  the same set of traffic data for one blog.  The two sets of data they produced differed by a factor of around a hundred.  Not only do the two sets of data differ greatly but they differ ever greatly more as time goes on. 

The gap between the two data sets kept widening even when I was not doing SEO for extended intervals.  While the traffic to Dads&Things is steadily increasing over time, Google Analytics tells me, through the version of the data destined for me to see, that the traffic to Dads&Things is steadily declining at a very substantial rate.  What is causing that?  My concern about what caused the rank fluctuations grew.  Still, the traffic volume kept increasing, at times at a substantial rate of growth.  That made up for it.  Obviously, a lot of people find Dads&Things without the help of Google.  Then something really weird and inexplicable happened.

Web rank history for fathersforlife.org
(Dads&Things is a subdomain of fathersforlife.org)

Monthly Page Views at Dad&Things
Per Google Analytics Dashboard WordPress
Note: Page views range from 613 to 4,824 per month (value of for Feb 2019 is an artifact of troubleshooting, during which I had to disable Google Analytics tracking)
Note also that actual monthly page views as per the host server range from 21,595 to 114.170

Although the web rank for fathersforlife.org had remained relatively steady in the range from 500,000th to 550,000th place, from about September 2018 to February 2019, on February 11, 2019 it began a precipitous decline for absolutely no apparent reason.  I had done nothing that could have caused that to happen.  That does not worry me too much.  The blog has more than a million page views a year, in spite of Google down-ranking the domain.  Search engines direct an abysmally low volume of search traffic to the blog (presently in the range from 1.7% to 2% of all traffic to the site).

Overall traffic volumes are high and growing, but I now I had become intensely curious about who was moving the goal posts, what they where, to what they were being changed, and to where they were being moved.  I posted  a request for advice on what was going on.  Tommy Wennerstierna, in Sweden, responded and pointed out something of great importance.  I have to look at that in depth, before I can report on it, which I will do in another part of this write-up (to be posted a few days from now.)

Are absurd consequences of SEO an artifact of revenue-splitting procedures?

A re-cap of the behavior of the traffic trends of concern.

Dads&Things, net-changes after two years of doing SEO

Impressions

1. Anyone who relies on ad-revenue derived from his website or blog and thinks that he can use Google Analytics as an audit tool should be concerned that there is more than one set of books that relate to the aspects of the Internet traffic to his domain.
2. One instance of discovering two sets of books is not acceptable evidence of general practice, but if there is one instance, there are likely many more.
3. In my case, the two sets of books, relating to aspect of Internet traffic to a single, common blog each, contain data that vastly differ from that in the other book
4. One set of books indicates a struggling blog, with a low and declining web rank and minimal, declining and vanishing traffic volumes.  That set of books is intended for public consumption, for establishing and tracking of web rank.  It prompts some web rating services to declare that the website and blog are not doing well, that they are in fact being penalized by Google.  Yet, page rank for the blog declined from 5 (on a scale from 0 to 10) to a page rank of 0, within just a few days.  It wasn’t my fault.  Nothing I lately did could have caused that.  The information provided by Tommy Wennerstierna explains it.  The release of another collection of Draconian algorithms by Google is the cause of it.
5. The other set of books, maintained for the host-service provider, indicates a blog of considerable vitality, with a substantial volume of traffic that is growing at a substantial rate.  That set of books is not used for web ranking.  It is inaccessible to the public and accessible only to the host-service provider and his client.
6. The first set of books indicates that the Internet traffic to the blog is declining at the rate of about 45% per year for the monthly page views and for the number of visits.
7. The second set of book indicates that the Internet traffic to the blog is growing at the rate of 94% per year for the monthly page views and at the rate of 53% per year for the number of visits.  The difference between the two growth rates shows that the visitors have an interest in the number of pages they view per visit that grows at the rate of 25% per year.
8. It is obvious that, for any party who derives revenues from the Internet traffic to a blog, the second set of books is unquestionably the one they will choose to go by.
9. Search engine optimization (SEO) had no perceptible effect on boosting the web rank for Dads&Things, but it appears that, with the help of Yoast SEO, I greatly improved the quality and structure of Dads&Things, even though my SEO project was just the first go-around, and many more refinements shall be made.  During  a 21-month interval the monthly numbers of page views rose from 32,643 to 92,493, visits rose from 9,436 to 19,202, and page views per visit increased from 3.5 to 5.1.  Those results far exceed the expectations I had when I set out to do SEO for Dads&Things.  More will be done, and I am looking forward to seeing the results of that.
10. In the first part of this write-up I declared that SEO is a misnomer, that it is intended not to optimize search engines but to tailor websites and blogs to cater to what Google’s requirements are.  There is no doubt in my mind that SEO is of far greater importance for websites and blogs, collectively, than it is for Google.

Contrary to those who make a living from advertising on web pages, by attaching advertising to the fruits of the labour by many others, the website of FathersforLife, and especially its blog, Dads&Things, are very much alive, receive a lot and ever more visits, and increasing numbers of people read the articles they present (close to 2 million page views a year),

That is primarily due to everyone, not so much and increasingly less so because of Google.  It is due to word-of-mouth advertising (call it keyboard-to-keyboard advertising, if you wish, it’s part of what samizdat evolved into).

So, bookmark Dads&Things and help with getting the word around.

Continued in the next part:
Requesting help with an inexplicable problem

_______
P.S. I don’t make any money on this, never asked for any, receive no funding from anyone and am not affiliated with any organization, government or otherwise.  I am  a pensioner, 82-years old, going on 83.

#Censorship #MediaBias #SEO #WebStatistics

See also:

• Google search bias quantified, constitutes censorship
• Internet censorship affects all, censors and censored