Google Analytics - Understanding Filters, Traffic and Results - Review

Today I am going to talk about 3 really cool things, which I would say, form the backbone of the way we analyze data in GA. CXL Institute has done a brilliant job at explaining these 3 topics that go a long way in having those intelligent conversations in office on how to make a website better. The 3 topics are: Filters, understanding and customizing traffic and understanding results.

Understanding Filters

Let us start with Filters. What is the purpose of using Filters? There are multiple – to make data easier to read, to present cleaner data and to fix fractured reporting.

In case I have many domains, I want to analyze traffic by domain and the first filter I am going to show you does exactly that – improves the readability of a view. In the last blog, I talked a bit about real-time reports on GA. These reports become very useful now as they can be used to test if these filters are working.

In the screenshot above, “analytics.google.com” is the domain and whatever follows is the request URI. Now “Filter”, under “Views” has a very cool feature, where we can merge the two. Under “custom”,  (.*) grabs whatever is there in the domain name,  we use it for “Hostname”. We do a similar exercise for the “Request URI” option to grab the remaining part and then we merge the two using a simple formula $A1$B1 under “Output to -> Constructor”. When we analyze the content now in Real-time reports, we can see the full-page address. There is another big advantage to this exercise. If someone is trying to send spam traffic (different hostname but same request URI) to the website, this method easily filters out that traffic. We can go one step further and exclude this spam traffic completely by using another filter type “Include”, where we put in the hostname we want to see in our reports. The “|” symbol, which means “or”, allows us to include multiple hostnames.

Now let us move on to cleaner data. Certain pages show up under “Behavior -> Site Content -> All pages” with slash and then without a slash. How do we fix that? We go back to Filters, find a Request URI that does not have a slash and add a slash at the end. That helps in cleaner reading. Another very important point to note is that sometimes sources/mediums show up once with the first letter in uppercase format and then in lowercase format. That is not adding any value to my reports. We make everything lowercase to remove any discrepancy and thus remove fractured reporting.

Understanding Traffic

The next topic on today’s agenda is: understanding traffic. Sources and mediums are the two major points here. Sources are “google, LinkedIn, baidu, Twitter, Facebook etc. Mediums show us the type of traffic – referral/paid/organic and so on. Campaign URL builder is a great way to track source, medium, content, term, campaign. The source is always the brand name, medium talks about type of traffic, the campaign name is generally the purpose/products, term is the subject line (under GA reports, shows up as “keyword”) and content is used as differentiators. If we are running paid campaigns, the medium is generally labelled as “cpc” as paid campaigns on “Google Ads” come in like that.

Important point to note here: “Share” traffic for any medium is basically the link that we share on that medium which leads back to the webpage we are looking for. It is not paid traffic and we are not labelling it as “organic” even though it is. This is to differentiate between the organic traffic that comes via search engines versus social media platforms. Make sure that everything is lowercase so that GA does not do fractured reporting.

I would like to mention one more important point here to fix fractures traffic. Sometimes sources show up in different formats like “m.facebook.com”, “l.facebook.com” etc. The first one is mobile traffic, while the next one might be via desktop malware scanner. That way, the story becomes very complex to understand and to simplify it, what we do is combine filters. Under custom, we do a “search and replace”, find “^.*facebook.com$”, which basically means that we should find anything that begins with some expression “facebook.com” and ends with “facebook.com”. We replace this expression simply with “facebook”. That fixes fractured traffic and makes reporting much simpler.

Understanding Goals

There are 4 kinds of goals – Destination, duration, pages/session, and event goals. There is another category that itself has a vast scope: Ecommerce results. Destination and event goals show up in real-time reports.

Goals refer to some sort of completion. For destination goals, it is the destination URL that matters. The goal template depends on “Industry” chosen. While selecting domain, there are three options: Exact Match, which means “Equals To”, Begins With which is used paste in the URL. If we want to check leads now, funnel reports show how we reached those leads. Here, we choose the lead page URL, switch on the “Funnel” option and in the steps add in the pages. (the URLs that lead to the destination URL). I f I have a homepage in the first step, that URL shows up in the subsequent page URLs as well. So, to differentiate the homepage from others in this case, I am going to put in ($|\?). This is like an exclusion parameter.

Duration Goals – we cannot see these goals in real-time. One misleading metric for this goal is the average time spent on a page. If I visit a homepage, spend 3 minutes there and leave that page, duration goal will record the time spent as 0. This happens because the moment we land on that page, GA records a timestamp but after that there is no way to record the duration for which we were there on that page unless we visit another page. But even then, the duration of stay on that next page will not be recorded if we bounce off it. GTM works better with duration and event goals but we shall talk about it in another blog.