Marketers have long looked to referral traffic to understand campaign performance and guide where to invest further resources. Conducting a referral traffic analysis may feel less urgent now, when new COVID-19 induced customer behaviors bring marketing campaigns to a standstill. But a pause is exactly the right time to reassess your analysis to ensure the data represents real traffic.
It goes without saying that the pandemic has upended many digital campaign strategies, especially digital ads. eMarketer noted the trend when Google reported its first loss from digital ads — a blow to a medium considered a reliable aid in building referral traffic.
The question you should ask is which traffic sources can potentially be supported with a limited marketing budget. Using the following fundamental approaches to referral traffic analysis may offer you some promising ideas.
The History on Referral Traffic
Since the early days of analytics, analysts examined metrics to note at what point website traffic arrived and interpret the traffic’s value. Referral traffic expanded over time to include devices and media, which resulted in more unstructured data. Bots have also emerged. All of this raises a quality question: does the traffic to a website represent real people?
The answer is especially significant for regressions and machine learning models, which thrives on clean data for model accuracy.
So how should you start a review? By taking a few steps to organize your data, you can clean up referral traffic and get some answers.
Related Article: Facebook Leads in Social Referrals — Maybe
Rank Referral Traffic Sources
Ranking referral traffic sources is the first step to organize results — this shows which share of traffic holds the potential for improvement. For example, if a campaign is sharing content aimed at B2B, I would want to see if Medium or LinkedIn are among the largest traffic contributors.
Referral traffic can be ranked by traffic volume over a time period of interest, or by conversions to note which source brought more effective impact to user experience, such as registrations. Each choice frames the order of referral source to address.
For large sites consider the flow chart, a Sankey diagram to discover what pages interested visitors after they arrived from a given channel.
Use Secondary Metrics to Vet Traffic Quality
Sometimes referral traffic visits may not represent actual people. Using a second dimension or metric can help determine if the traffic is real. With Google Analytics, you can go to Acquisition, then sort referrals by average session to see which sources best contribute to engagement. You can also arrange by pages per session.
Audit this list to note sources from augmenting sites — these are sites partnered with your campaign for sharing content. Doing so gives a sense of the value of traffic being received from the relationship — review the posting schedule for consistency — decreases in traffic can reflect a paused campaign.
Related Articles: Use Lookalike Audiences to Reinvigorate Your LinkedIn Ad Campaigns
Identify Spam Among the Referrals
Spam often gets mixed into referral traffic from different sources. Spam can be sites that became associated with your site over time. Then there are different classes of programming bots which can claim a percentage of traffic online. When crawler bots visit your site it can trigger the analytics tag. Ghost metrics — fake data sent directly to an analytics solution — appear in the reports as a result. Non harmful bots can at times serve an important purpose, such as language bots. But too many bots indicate false leads.
There are two ways to address spam. Google Analytics provides a bot filter — make sure it’s switched on.
The second method is to include spam sites in a non-referral list on an htaccess file. The htaccess file is a text file in the hosting directory of a website or web app. Its purpose is to tell the search engine what sections of a site to crawl and more importantly for the current context, which sites not to crawl.
Related Article: How to Effectively Use Google Analytics TreeMap Reports
Filter Out the Traffic You Don’t Want
Referral exclusion filters will exclude traffic from spam sites and other unwanted sources. To use it in Google Analytics, navigate to the admin page of the account, then click on a given property that contains the analytics script. The filter appears in the view column. You can indicate the site by clicking the add button.
A referral exclusion list will only block the site from being included in its calculations. To block sites arriving via search, add the site URLs to the exclusion list in the htaccess file.
But Also Filter for the Traffic You Do Want
Given the multiple ways consumers access a site, consolidating referral sources in a filter can save time in assessing traffic behavior.
A report column can gather multiple dimensions for URL shorteners. For example, in Twitter, people can use a variety of URL shorteners in their tweets, creating multiple streams of traffic in the process. You then face the extra step of gathering metrics for all Twitter related traffic in a report as a result.
Applying a filter can gather related URLs into one dimension entry, organizing reporting into a more convenient view that saves the work of making repeated calculations or adjustments.
This is done through setting a custom channel grouping. Custom channel grouping is a text box to enter the URLs. Sometimes you’ll require a regular expression, a programming syntax that identifies a sequence of characters from a URL. You can label the grouping, as well as indicate the degree of match, select a display color for how the grouping appears in a report, and preview the set up.
One point about filters: keep a main profile view without filters as a back up in case a filter or grouping doesn’t work at first installation. Google Analytics data cannot be retrofitted once you’ve changed the settings.
Referral traffic metrics will never be completely perfect. Sorting genuine human activity associated with a website or app will always be a dull, but necessary chore. But applying a few tactics can ease the way for establishing the right starting points to protect your digital marketing campaign budget.
Pierre DeBois is the founder of Zimana, a small business digital analytics consultancy. He reviews data from web analytics and social media dashboard solutions, then provides recommendations and web development action that improves marketing strategy and business profitability.