There are those times in digital marketing when an analytical review, which shows traffic and conversion spikes, brings bliss that cannot be rivaled by many other means. Working in such a competitive space often makes the wins we find that much better. However, what hurts most of all is when we find that data reporting inefficiencies have painted a picture in our digital data that isn’t quite true. So, many times I have to review an analytics profile with a fine tooth comb because I understand that previous issues with how site traffic and conversions were reported inaccurately in the past can make a quick view show vast improvement that isn’t necessarily true.
I’ve found six areas that can wreak havoc on your understanding of digital success. Understanding how to ensure you are looking at the real picture will keep you from any embarrassing situations of premature celebrations of success.
This is one of the first stops I make when reviewing an analytics profile for the first time. It is important to understand how much of your traffic is coming from customers, new visitors, etc. and not from internal staff traffic, vendors, call centers, in-store kiosk users, etc. While it may be worthwhile to have a “Master” profile to gather all site traffic without filters, it is advised that you feature a profile which excludes traffic from these mentioned parties.
There are a lot of traffic types that can end up in Referring Site visits. While I typically am preoccupied with Organic Search and Paid Search visits, it helps to understand what other channels of traffic are bringing to the table. However, incorrect traffic numbers could be at play in this channel if you are not instructing search engines of your domain’s relationships with other owned properties, sub-domains, etc. For example, you may have users on your site who through the conversion or lead funnel move from the parent domain to, ex. checkout.mysite.com and then return to mysite.com/receipt to receive a receipt or thank you message. In my example above, if I had not placed “checkout.mysite.com” in the Referral Exclusions section of Google Analytics, then we would see that a user entered the site through a channel such as Organic Search and then we would also see another referral from checkout.mysite.com. It is important that you understand domain jumps that may be happening in a typical site visit on your site as you could be inflating your traffic count. Along with this, it is also important to ensure that if this exists that you have a good understanding of cross-site analytics tracking.
Direct traffic has really been increasing, you must have a really strong brand? That would be nice, but what might be happening is that you may have traffic arriving from PDF links, or you may have traffic arriving from Email template links that your analytics profile does not understand and thus is dumped into the Direct traffic channel. You might think, “Wow, our Social traffic is really converting well. Those crappy posts we place in Facebook are working well.” Actually, you don’t understand that another team has been advertising in Social Media outlets and without proper URL tagging, you don’t know if advertising traffic is working well or if it is from your Organic Social posts. Ideally, you should consider where all site traffic may be coming from. Utilize Google’s URL builder so that you can set up UTM structures to let Google Analytics know that traffic belongs to an advertising effort, a text messaging system, online PDF brochures, etc. Utilizing UTM structures, specifically the UTM Medium, will allow you to tell your analytics profiles where traffic should be posted.
Default Channel Settings
As we have reviewed above, not understanding your site traffic referring mediums can create a cloudy data picture. You should also consider the “Source” as well. When reviewing traffic by channel, you may see traffic being lumped into the channel titled “Other”. Again, this is another occurrence of your analytics profile not understanding which channel to place this traffic. It is worth mentioning that if you are receiving hundreds of thousands of visitors you may be capping out on your pageview limits in Google Analytics and visits are being sent to the Other channel. If that is the case, then it is time to upgrade to Google Premium. For the rest of us who wished that we were getting that much traffic, you need to look inside the Other channel’s traffic, but by Source. Again, there may confusion on where this traffic is supposed to be reported. The Source of these visits may show you what channel they should report under. Next, you will need to review your default channel grouping settings and revise these to move specific traffic by Source into the appropriate channel.
Test your Goal URL Locations
Are conversions really as great as they appear? Granted, if your goal URL locations are an exact page destination then they are likely reporting accurately. However, to ensure that these goals are true, ensure that the goal URL is not indexed in search engines. This can inflate goals. Additionally, if you can apply a funnel where the goal URL is only reached when specific URLs are traversed, this can help with goal accuracy as well.
Where goals can be reported with great inaccuracy is when regular expressions are used due to goal URL locations having varied structures. To help ensure that all goal URLs are appropriate, review the Pageview section of Google Analytics and provide a Regular Expression filter and review the URLs that are provided. If you see many URLs that are not what you see as goal URLs then your regular expression needs some work.
Duplicate Content Issues
Seeing a huge jump in Organic traffic for a search critical page on your site can be an exciting moment. Inversely, big drops can send you into a frantic, “what happened” moment. These occurrences typically happen when reviewing year-over-year data. Seeing these transitions will warrant that you make the following considerations.
Do you have duplicate content versions vs. your previous data set based on having URLs that have trailing slashes and non-trailing slashes?
Do you have duplicate content versions vs. your previous data set where you have uppercase/lowercase page versions?
Do you have duplicate content versions based on the same content on separate URLs where Google is now ranking a different page version vs. your previous data set?
Issues such as these can occur, which do not show surges/drops at the channel level, but by-page fluctuations.
While I have seen some real head-scratchers over the years when it comes to data anomalies, I have to admit these are the most common that I see on a regular basis. Keep these points in your “digital toolbag” and call upon them next time you see that big spike or drop.