When we begin to analyze the results of a site and performance for each channel, as indicated in a Web Analytics process , some things are easier to interpret than others.
For example, if the traffic of the Social Media channel fell or rose, we come into a greater degree of detail, see exactly what the network had variations and investigate the exact causes (were more posts or less posts, scope and interactions fell or rose , and so on). In (not provided) mask some things, you can get an idea of keywords ranging ranking or landing pages that received less traffic.
Already direct traffic is something that leaves many lost professionals. Usually we understand as direct traffic one where someone entered the site on their own to your browser, not coming for clicks on any other source. When this traffic source has variations, marketers generally find it difficult to make an interpretation of what actually caused people to enter more or less on the site and which explains this behavior.
Our idea with this article is to show a little better various items impacting direct traffic and how you can understand it a little better.
To credit the traffic to one source or another, the Google Analytics looks for information in two different ways, one following the other.
1 – The first is the URL itself. Usually he does not find anything here, but to find it is to this source that it gives priority and will credit the traffic. This is done through a system where companies can place a espécio to “label” the link that promote to identify the source of the way they want.
Basically you can place parameters in the URL link that force Google Analytics to mark a specific source if there is a click there.
For example, the link https://resultadosdigitais.com.br has no marking and Google Analytics to follow step 2 to identify the source of traffic.
In the case of link https://resultadosdigitais.com.br/?utm_source=newsletter&utm_medium=email&utm_campaign=disparo-janeiro , regardless of the click happens on Facebook, Google or any site, there indicated parameters will force Google to understand the click came from an email newsletter.
The pros do it to better identify the results of specific campaigns and is always valid label banners, ads, email campaigns, etc. To create this label just use the URL Builder , Google.
2 – In most cases the links have not labeled and the GA part to the second way to identify: a page reference.
When you load a page in the browser, it makes a request to a server to deliver this page and this request includes a field in the header that identifies the address you choose to link. It’s basically this heading and what was stated in it that the referrer GA read to infer where the visitor came.
In practice, whenever the GA does not have this information and can not interpret, he credits the visit as direct traffic.
Understanding how direct traffic is counted in Google Analytics you can see that there is room for mistakes and not always what it appears as direct traffic is the setting that uses the market: people who typed the address in the browser.
It is very common that direct traffic becomes “dirty” with some flaws:
On the other hand, the original direct traffic is also not accounted often. This is because, from the moment the user comes to the site and Google Analytics determines the source of traffic, it gets a cookie, a kind of stamp that identifies the source, among other things.
The problem is because direct traffic, unlike everyone else, do not overwrite the cookie and just crediting the visit to the same source as before.
If someone came to your site by clicking on a campaign Adwords , for example, and one week later entered the site by typing in the browser, the visit will be counted again as coming from Adwords. That’s why a lot of people who have already made announcements and stopped still seeing the GA few visits paid even having stopped investing.
Often the failure to account happens because the browser does not make the request through all the information correctly.
Groupon in the US experimented blocking the site from appearing on Google and the correlation of the fall in organic visits with the fall in direct traffic was high. The estimate presented by them was that for internal pages up to 60% of which is credited as direct traffic is actually organic . Overall, they reached something close to that 25% of searches made on the Internet Explore are being wrongly credited to direct traffic as 10-20% for Chrome and Firefox and over 50% in mobile browsers.
The real direct traffic should be a good indicator of the value of your brand.
It consists of people who have already bought your company and turn to look more, by mouth recommendations mouth, by people who are so fond of their content marked as favorite or by offline relevant appearances (a matter of magazine or newspaper, something they heard on TV or radio).
A good job of branding and media relations should contribute here in these numbers and can be evaluated.
The number of visits via direct traffic should also be high in cases of products online, where people come to sign in and access their accounts.
With all that said, you should still get the question, and how to evaluate and work properly with all this?
The first recommendation is to try our best to get all the dirt from your direct traffic. This includes two main actions:
The second thing is to try to better understand which pages people who arrived via direct traffic are entering. In general, people enter the Home, blog or some other main pages that you may have. Separate this segment and interpret it as the true direct traffic.
The rest should probably be seen as another source that was incorrectly assigned. It is very unlikely that users actually enter a long URL as it is usually a specific blog post, for example. In this case you need to assess more specifically a case to try to understand what is the correct source.
In the case of social media and organic search, the trend is that the assignment errors are only a percentage and comparative curve of the channels, as we have indicated in the Marketing BI just above, indicate a correlation.
Already in other cases we need to think if there are campaigns that lead to the page and that eventually are not with the parameters set correctly.