Google Analytics dominates the search analytics market with over 50 million websites under its belt, which is a statistic that gets outdated every year due to the massive adoption of companies and individual websites. And we’re not talking about small blogs, we’re talking about big brands here because 45% of Fortune 500 companies use GA to control their businesses.
Yes, the market share is quite impressive.
But there’s one thing, though: There are some “weak spots” you need to understand before crowning Google Analytics the Holy Grail of your Marketing arsenal. There are a few crucial things to know before you can affirm how your website/e-commerce is actually performing because every business owner knows how misleading data can be harmful to one’s business success. If you assume that data is 100% veridical, then you got yourself a problem that may be very hard to track back to if you don’t know what you’re doing.
Therefore, I’ll make you understand how reliable the metrics of Google Analytics really are.
So let’s cut to the chase and reveal some important stuff every marketer and business owner should know by now.
Here’s what I’ll cover:
- Search Traffic
- Referral Traffic
- Direct Traffic
- The Unique Visitors
- The New Visitors
- The Normal Visitor
Let’s not waste our times anymore and begin this.
MISLEADING DATA: SEARCH TRAFFIC
I’ll start with a major topic in the online marketing industry: Search Traffic. In your Google Analytics dashboard you’ll see something similar to this image below:
Here’s a very easy-to-understand definition to it:
SEARCH TRAFFIC EASY DEFINITION:
Search traffic is traffic that came from a search engine website like Google, Bing, Yahoo! or others that Google categorizes as search engines.
So technically, search traffic should point out a number of visits that found you through search engine queries, just by typing keywords related to your niche. The desire in most marketers’ hearts is that hordes of users find his brand or his client’s through search queries, by finding interesting branded content created with the purpose to drive traffic. Well, unfortunately, that’s not what always happens. Because once you understand how users’ online behavior are and their patterns, this kind of data changes dramatically.
The problem begins with these 3 reasons below:
The user typed a branded term OR
The user is actually searching for something inside your Website OR
The user is investigating your rankings (competitors or newbies).
Let’s start with the first and take my own brand as an example for you to get this.
EXAMPLE: If someone typed “Peterson Teixeira” on Google then it would be a branded term in my case. But, here’s the thing: It could be someone who already knows my brand like a previous customer or a fan, and is just too lazy to type my brand’s website, creating the ILLUSION that I was found by a niche keyword like “Marketing” or “Copywriting” or some other keyword related to my brand. That’s one case. But here’s another VERY common as well: A user with a Smartphone may prefer to type PART of my URL on Google instead of typing it into the address bar because he doesn’t want to type the whole URL (lazy!). Or maybe the user prefers to type on the center of his smartphone screen on Google because he has fat fingers that make him click away or close the smartphone’s browser window accidentally.
So what you need to start realizing is that human behavior changes and reveals TRUE DATA if you actually pay attention. Therefore, in the examples above, such search traffic is not a product of my content marketing strategy or some other marketing campaign I made because THE REAL INTENT of the user was actually this:
BRAND BECAUSE I’M TOO LAZY TO TYPE EVERYTHING MYSELF
And it doesn’t stop there.
There are other possible reasons too you might consider remembering:
- THE USER FORGOT A WEBSITE URL HE WANTS — The user forgot which URL of your website has something he wants and asks Google to find it for him, like a resource page. This happens a lot because internal search engines of websites are TOO LOUSY to find the content desired by the user in comparison to Google’s search. So Google being an external search engine, is in fact far more accurate (obviously) in finding what the users wants inside that website than the actual search engine of that website.
- THE USER IS STUDYING YOUR WEBSITE’S RANKINGS — The user is a technical guy or a competitor who is doing his research on your brand. In this case, people are coming through search engines with a query, they click on the page Google gave them and they investigate WHY Google gave such page as a result. I personally do this a lot for research purposes and when I’m writing about complex topics like Neuromarketing.
- THE USER IS LAZY TO TYPE OR DON’T WANT TO CLICK/TAP ON THE WRONG PLACES — Just to finish it strong, remember that the majority of Google users come from mobile and every mobile user has some preferences on how to access their favorite websites. Some prefer to type in the center of the screen to prevent closing the window accidently or clicking on a notification for instance. That’s just ONE scenario. One case. There are lazy users as well who knows that with a few keywords they find you faster, like an abbreviation of your brand’s name or some specific keyword only your brand uses, for example. So keep that in mind. Users are lazy. They want practicality.
Just so you can memorize this faster, I’ll use a real case. Sometimes I promoted my own brand by provoking users to do as the image below says:
A RESULT IN YOUR COUNTRY?
The result according to Analytics would be a search traffic, but it’s not the case as I proved above once more. Users have different behavioral patters online with search engines. This is why Marketing is not an exact science, my friend. Because t all relies on CONTEXT. Context. It changes everything. And there may be other reasons too as you may imagine because behavioral patterns change very often due to technology upgrades and innovations. Once some new technology appears and alters how society works, it ends up reflecting on how humans do things online.
Anyway, let’s continue.
MISLEADING DATA: REFERRAL TRAFFIC
Referral traffic is another point worth giving attention. Technically, referral traffic should show visitors who came from other websites but again, that’s not the case. Here’s the image for you to situate before we tackle this:
Here’s the easy-to-understand definition of it:
REFERRAL TRAFFIC EASY DEFINITION:
Referral traffic is a visit that came from another website that is not a search engine.
A referral must come from the browser and it must be a website (you’ll understand why I said this last sentence later). But with only this referral info, can you think of other famous marketing channels which marketers treat separately from websites and wouldn’t like it to be treated as referrals? Let’s see if you have guessed:
- WEB EMAILS — If someone opens a newsletter on Gmail, Yahoo!, Hotmail or other web email client then in essence, it is a website. Right? Therefore, it will count as a referral in your traffic analytics section and not an email or a newsletter as it should be.
- SOCIAL MEDIA — Just like email, social networks are also websites that inflates your referrals metrics because they’re, well, websites. And as such, they are treated as referrals as well if you don’t tag them properly with link tracking.
- SPAMMY WEBSITES — There are needy websites that send visits through their spammy websites just so their useless domains can appear as a referral link in your stats. The goal is to obviously make you curious about their link showing up to you as a referral and drive YOU to their website. Get it? So spammers create content with a link to yours, make people or bots click on it, and as a result your referrals get inflated.
- BROKEN ANALYTICS CODE OR CAMPAIGN CODE — If you make a mess with your analytics code (the analytics.js file) that is what Google uses to connect and interpret your website’s data, then you’ll get “direct traffic” instead. The same goes if your campaign codes has an error like a variable typo or something.
- HTTPS WEBSITES DO NOT SEND REFERRAL DATA TO HTTP WEBSITES — Websites that use the HTTPS Protocol don’t pass referrer data to HTTP-only websites by default, for security reasons. And since Google Analytics populates the referral report based on what is in the HTTP Header, no data gets captured and therefore, no referer shows up on referral stats. So if a HTTPS website linked to your HTTP website, you wouldn’t be able to tell. Therefore, people with HTTPS websites might be linking to you without you even knowing it. But the “good news” is that these visits are counted as part of the direct traffic because once Google has no information to work with, it will label it as “direct traffic”. That’s a rule. Google says: “No referral information? Then it’s direct traffic”.
Next misleading data then.
MISLEADING DATA: DIRECT TRAFFIC
Understand this: NEVER assume your direct traffic stats always means that someone typed your URL in the browser directly to go to your website because it is simply not true. As every analytics software, Google Analytics fails to provide perfect metrics on this as well. So first, let me address what we’re talking about with a screenshot:
Now here’s the easy-to-understand definition:
DIRECT TRAFFIC EASY DEFINITION
Direct traffic is traffic for which a referrer wasn’t specified due to lack of data.
OK. Now to make this even easier, I’ll copy and paste a short explanation about direct traffic that worth highlighting as well coming from Brian Clifton himself, an internationally recognized Google Analytics expert:
The use of “direct” can be ambiguous when analyzing your referral data. Out of the box, Google Analytics will not track email campaigns, links within files (such as PDF, DOC, XLS, PPT etc.), RSS referrers, or even your email signature for you—you need to make changes to those in order to track them. If you don’t do this, such visitors will be classified as “direct.”
Are you starting to understand why you need to track your links?
So here’s how misleading direct traffic can be in practice.
Google Analytics considers direct traffic the following:
- RSS FEEDS — Since RSS Feeds are not websites but instead a format for delivering regularly changing web content, a click on a RSS feed link will be categorized as Direct Traffic, inflating those particular stats instead of being categorized as referral traffic.
- FILES LIKE PDF, DOC, PPT, XLS etc — Google Analytics also can’t tell whether a person typed a URL into the address bar or if she came through a link embedded in a file like a PDF Ebook because for GA, both are the same since they don’t send any referral data. And you remember what happens when GA doesn’t have any data to work with it, right? It is also counted as direct traffic.
- DESKTOP EMAIL CLIENTS LIKE THUNDERBIRD AND MICROSOFT OUTLOOK — The very same thing applies to email clients like Thunderbird and Outlook (and other programs). For example, if people clicked on your e-mail signature it will be counted as another direct traffic source, instead of being considered an email campaign instead because these applications are not in the browser where all the GA magic happens.
- HTTPS WEBSITES DO NOT SEND HEADER DATA TO HTTP WEBSITES — This was explained in the referral section, but is good to remind you that HTTPS websites when communicating with HTTP websites (HTTPS -> HTTP not the other way around), won’t pass any referral data at all, making it look like a direct visit.
- BROWSERS SOMETIMES FAIL TO SEND HTTP DATA — Yes, browsers fail to send referral data too. A very nice and BOLD experiment made by Groupon (they removed themselves from Google’s index for a while) has proved that 60% of their direct traffic was actually made of organic traffic because browsers failed to send HTTP data. About 10-20% of Firefox, Chrome, and Safari desktop traffic reported as Direct is actually Organic, which reinforces one more misleading data point.
Now you know how much confused you can get while reading your analytics data if you don’t understand all this context from the start. It’s confusing for those who are unaware of all this. So keep in mind that there are a few discrepancies to consider for direct traffic sources as well.
That settles this section.
Now it’s time to change things a little bit from TRAFFIC to USERS.
MISLEADING DATA: THE UNIQUE VISITORS
Yes, more misleading data: Users. Google Analytics has some unreliable data on this which demands a deeper understanding of how users interact with websites and what logical and technological limitations Google has to face in order to properly track them. Therefore, we still need to pay closer attention to users as well.
So here’s the definition of UNIQUE VISITORS:
PS: I must use the official version from now on so you can understand things better. No easy stuff for you anymore!
The official Google Analytics definition of this Web metric is: “Unique Visitors is the number of unduplicated (counted only once) visitors to your website over the course of a specified time period.”
Now here’s what is misleading:
The number of unique visitors you get per day for a week DOES NOT equal the TOTAL number of unique visitors for that week.
Why? Because if Google Analytics counts unique visitors once a day, then you may have been interpreting your unique visitors data for the whole week as another inflated number.
For instance, you may have got 3,500 unique visitors to your website’s blog on a Monday. Ok. Good. But then these very same visitors came to visit your website over the rest of the week again, creating the ILLUSION of 3,500 x 7 = 24,500 unique visitors when in fact, they were all the same guys who came on Monday in the first place.
See how misleading that data can be if you don’t know what it means?
Because if the TIME PERIOD is set for A DAY, then a week (7 days) will multiply this number by 7. Therefore, you cannot make bold assumptions like “I have 500,000 unique visits per month” using raw Google Analytics data without realizing this particular point in your metrics.
So chop that data off as soon as you look at it because it’s not real.
Next misleading data.
MISLEADING DATA: THE NEW VISITORS
Another metric you need to be aware of is the “new visitors” metric. This one has flaws as well which raises some issues business owners ignore or fail to see (which is even more painful). But for now, let’s start with the definition. According to Google Analytics, a new visitor is the following:
The new user definition is: “Users that have had at least one session within the selected date range. Includes both new and returning users.”
Alright, what do you need to know here?
A visit is a session. So if a visitor is IDLE on your website for about 30 minutes, his visit will expire and the next click he makes will count as the beginning of a NEW visit. This will happen if you’re using Google Analytics’ default cookie value which is exactly 1800000 milliseconds.
For high-authoritative websites, this number may inflate a lot because those who also like to write or read high-authoritative content are used to open several links to do research or to interact with the content, leaving several browser tabs opened while they do something else. As a result, data gets hurt. Because some other link may steal the user’s attention until he comes back to your content, creating the illusion that several new users are accessing that website when in fact is just that one user who doesn’t move a finger in your website.
Such an eye-opener, isn’t it?
Now, let’s check the main metric most people like to pay attention to: The Normal Visitor.
MISLEADING DATA: THE NORMAL VISITOR
The visitor metric is one of the most “wide” ones you have to understand because there are a lot of different scenarios on which you can associate a visit to the same person. And for businesses, it’s a huge problem to have different scenarios that messes up with part of your core data, hence the need to understand the whole context behind visits.
Well, by easy-to-understand definition:
A Visit is a Session. It’s when someone enters your website to browse. Every time a person enters your URL is a visit.
Google decided to trade names from Visits to Session a while ago just in case you have found some information elsewhere that doesn’t match your GA dashboard.
But here’s a good screenshot to remind you of what we are talking about, just in case:
Ok, now let’s open your mind once again.
Look, we all have different devices from which we access THE SAME WEBSITE. Right? So you may access Youtube from both your smartphone and your desktop computer. Cool. But for Google Analytics though, if you access those at the same time, you’ll be considered as 2 visits when in fact you’re the same person. Get it? Two devices, two visits. One person.
Shocking? Not so much. But you probably forgot about the “real world” behind technology, right? So here’s some interesting data. There was a study that Sun Microsystems made which estimated the percentage of users who were using more than one computer in a month to visit the same website, as high as 30% of all visits (!!!). With that kind of metric, maybe your traffic numbers will probably go down a bit. That’s ok.
So basically there are a few problems with this metric:
- There are users who own multiple computers/devices with internet access.
- The user behavior has an enormous impact on the accuracy of your GA data due to the user access to his browser cookies, which is what Google Analytics uses to gather data from a user to count visits.
With such things in mind, let’s expand once again your understanding of all this:
- THE SINGLE USER WHO HAS MULTIPLE COMPUTERS/DEVICES —
- SEVERAL DIFFERENT USERS WHO ARE USING THE SAME COMPUTER/DEVICE —
- INTERNET SAVVY VISITORS WHO REJECTS OR DELETES COOKIES —
You know people access the Internet from several different devices like from a mobile phone, a desktop computer, a friend’s desktop computer, a tablet etc. Therefore, a single user can generate several visits because he is using different places to provide information to Google Analytics, counting each visit as a different one when in fact, he’s the same person behind different keyboards.
Haven’t you used a computer’s friend to access your email account or a certain website before? I bet you did. Well, there is that too. People may end up sharing computers or devices just to access a website which obviously puts your GA data in checkmate again. For some cases though, the data is more trustworthy, like in Internet cafes were the machines/computers are set to erase cookies as soon as a user ends his session.
There are 2 main scenarios here: (1) Machines that are behind a firewall which may block/reject cookies from Google Analytics making it impossible to track the user who is using that machine. This may happen if someone is accessing a website from within a company’s headquarters (his work/job environment) that has very strict security measures. Thanks to Web Attacks through Malicious Cookies happening all the time, certain companies need to protect themselves since crackers like to steal people’s cookies to do some evil stuff with it. That’s reason nº1. Reason nº2 is quite simple: (2) The user hates having cookies on his computer and he deletes it from the browser himself by his own free will. By the way, you really cannot believe how UNEXPECTED the stats are for cookie deletion.
Now, after I’ve shown you all this let me ask you: Are you shocked?
Probably, right? Well, these are the main points you should consider memorizing whenever you yell “AHA! My visits just raised by X% this month!” because as you can clearly see, Google Analytics cannot be that reliable yet. Technology needs to improve. There are still some issues and limitations. So upgrades need to be made before we can trust more on its data like a blind man would trust his guide dog.
You can monitor your web traffic with Google Analytics with a good considerable amount of precision, but it’s still far from being perfect due to technology limitations. Google can’t tell who you are yet, but I believe that maybe in a very near future, browsers may use the device’s webcam to read your eye pupils to fill cookies with some personal data, aiding tracking software like Google Analytics to reach to a more mature state. This will give businesses more accurate data to work with than what we’re seeing today. But anyway, that’s not the case yet. For now, just stay with everything you read to prevent yourself from making stupid marketing decisions.
Now you have no excuse anymore.