Why Shortner Links Show Inflated Statistics
Let’s start with an understanding that no two analytics programs will ever give you the exact same result.
We’ll discuss the details of this in a minute, but for now it’s enough to understand that each analytics platform has its own idiosyncrasies, so much so that if two platforms DID agree on the number of visits, I would be worried that something has gone terribly wrong.
But first, what exactly am I talking about here?
Shortner Vs. Reality
Here is what goo.gl shortner is reporting for a recent post I published.
According to google’s own shortner, I’ve had 639 visits to that page in the last 5 days.
But here is what my Google Analytics is reporting.
324 visits to that same page.
In theory, more people could be coming to that page via other channels (RSS, direct links, email, facebook, etc), and yet, it looks like more people came via the shortner…or do they?
The Alexa Case Study
In order to illustrate the challenge, and the idiosyncratic nature of measuring traffic, I like to use Alexa.
In case you’re not familiar with it, Alexa is in fact a standard of measuring online traffic. When Triberr neared 1000 most visited sites in US, Dan and I were ecstatic.
And while Alexa is the standard, it serves as a great example of how fundamentally flawed these platforms can be.
In order to calculate your site’s “popularity”, Alexa looks at the % of users who have installed the Alexa toolbar AND are visiting your site.
So a site like google, that everyone goes to, will probably be ranked quite high.
However, if you have a very popular non-techy site, your overall ranking will most likely be underrepresented.
In fact, I often recommend to those concerned with accuracy, to uninstall Alexa toolbar, because simply having Alexa toolbar installed on your computer, and going to your own site on regular basis as we’re prone to do, will skew the ranking.
So you see, each tool has it’s own shortcomings and limitations. But that doesn’t mean they don’t have value.
How Shortners Work
Shortners (bit.ly, goo.gl, t.co, etc) wait for someone -or something- to “click” on them and they record that click as a visit.
It’s very much a “front loaded” function. And it’s very simplistic in nature.
In fact, these shortners are so front loaded that one of the challenges is eliminating search bots that are “clicking” and following the shortner to its destination. This is why I refer to them as “front loaded”.
Shortners are not smart enough to distinguish a person from a bot. However, thats not why shortner stats are inflated.
Google and bit.ly (to name just two) maintain up-to-date bot lists which eliminate the inflation due to bot clickthroughs. It’s especially easy for google to keep that list up to date since they are the ones unleashing those bots to begin with.
So, if we compare that to the way Google Analytics works, you will notice quite a difference.
How Analytics Work
Let’s start with an illustration.
We all have this code stored somewhere on our blogs. I store mine inside an area specifically designated for GA tracking code.
This piece of code is the key in understanding the discrepancy amongst so many traffic-measuring tools.
This one piece of code is executed with every single page on your blog, every time a human visits your site.
Once the code is executed, the Javacode snipped performs an E.T. Phone Home function and calls a Secure Google Server to transfer data over. This data is then consolidated and displayed as a report inside your GA in the form of valuable visitor insight.
- Operating System
- Bounce Rate
- and much, much more.
Well, my friends, if the script doesn’t get executed, it doesn’t get counted as a visit to your blog.
But why wouldn’t the script get executed, you might ask?
When your server passes the information to Google Secure Server for tracking purposes, it does so in the form of a cookie.
However, if I’m sufficiently paranoid, and decide to block all my cookies, then my data can never be passed to google, and therefore, you will never see me as a visitor.
How many people actually do it?
A quick google search turned up over 10% figure.
So who is right? Shortner or analytics? And the answer is neither. Or both. Whichever you prefer.
Here is my real verdict.
You can NEVER take statistical information as 100% accurate, and the information that you do collect, you should look at it in terms of trends, rather than accurate representation of reality.
I can’t stress that enough.
In-depth analytics are used to track trends. Nothing more, nothing less. And if more, then it’s only gravy.
But have you considered the fact that traffic doesn’t really matter?
Traffic Doesn’t Matter
Hear me out…
What can traffic do for you? What if I could send 100s of thousands of visitors to your blog right now? Is that something you would want?
If you’re smart, the answer is no, simply because that many visitors would crash your servers. And if your servers didn’t crash, how many of those visitors would actually “buy you”?
What we really want is for the reader to know us, trust us, and buy from us… whatever it is that we’re selling. Could be a service, product, or a simple notion that you’re an expert of some sort.
And you could accomplish that much faster and more effectively by having your post syndicated to many dozens of blogs, rather than trying to bring more eyeballs to your one blog.
Not that the two are mutually exclusive. Once people see you, know you, and like you elsewhere, they will in fact seek you out and come to your blog.
What am I suggesting here? Guest posting? Hell no…guest posting is slave labor which yields minimal results.
I’m talking about Triberr’s new ReBlog feature. More on that here.