Attribution reporting – it’s all about who and what gets credit. Who gets the credit for those 5 new leads that came in yesterday? Who gets credit for the 20% increase in applicants for the cybersecurity program? It all depends which attribution model your enrollment team uses and how reporting is handled throughout your department. At the end of the day, you need insight into which marketing and recruitment tactics are producing the greatest results so that you can adjust your precious budget (and time!) accordingly.
But exactly which model(s) should schools be using? Let’s first walk through an example to identify some of the major bright spots and pitfalls of traditional marketing attribution models.
To paint the picture, let’s say you just launched a marketing campaign for your brand new cybersecurity program. Your campaign consists of several marketing channels, including 3 email announcements, a program guide, LinkedIn ads, 2 blog posts, and a couple of landing pages. The goal of your campaign is to spread awareness about the program and ultimately generate qualified applicants. Soon into the campaign, new contacts come pouring in to your CRM, including one new applicant named Jessica. Here’s how Jessica engaged with your campaign prior to applying:
Now you’re faced with the question: “Which marketing channels should we credit to Jessica’s application to the cybersecurity program?”
Let’s take a look at a variety of attribution scenarios to help answer that question.
In this model, the last page or source of engagement before applying is given 100% of the credit for the conversion. In our example, Jessica’s last interaction before applying was with a LinkedIn Ad. In this instance, the LinkedIn Ad would receive ALL of the credit for Jessica’s application conversion.
Where this model falls short is that it fails to give any credit to engagements that occurred with marketing channels at the top or even middle of the marketing funnel. In reality, Jessica could’ve made up her mind that she was going to apply to the cybersecurity program on Day 4 when she was reading your program guide!
This first source model gives 100% of the conversion credit to, you guessed it, the first source of engagement. Sure, Jessica opened the first email you sent on Day 1, but are you really sure the first email fully convinced her to apply? In using this model, the answer would be yes. This model is flawed because all of Jessica’s subsequent engagements with your marketing channels after her first engagement with your email are ignored.
There are, however, a couple of use cases for First Interaction. Consider an approaching application deadline and your deployment of a push-to-apply email. Your goal is to analyze the direct effectiveness of that email as it relates to converting prospective students into applicants. You analyze the email in an isolated manner where all other marketing channels are ignored. In this case, all you care about is if the first interaction with your push-to-apply email resulted in an applicant.
In this model, the pages or sources that were visited most recently are given more credit, and pages or sources visited less recently are given less credit. This is a more complicated model, and it assumes that the information a prospective student engages with days prior to applying is more important than the information they engaged with weeks or even months ago.
We assume that Jessica’s interaction with your LinkedIn Ad and her visit to your program’s website page are more important than the first email you sent her and the cybersecurity program guide she downloaded.
This model isn’t useful for getting a holistic look at Jessica’s prospective student journey. Where it could be useful is looking at the effectiveness of an inquiry communications email series whose goal is to push prospective students like Jessica to apply to your program, and if your ultimate goal is to measure the effectiveness of those emails.
The position based model gives 40% of the credit for the application conversion to the first and last engagements, and 20% of the credit to all other engagements. Again, the assumption here is that Jessica’s decision to apply to the cybersecurity program was in large part a result of her first and last engagements. While this model is a step in the right direction and a good attempt at spreading out the attribution credit to more than just two tactics, it’s still too simplistic in the way it weighs certain engagements.
This is where we start to have some real fun with attribution models. The linear model is a type of multi-attribution model where equal credit is given to each and every engagement before Jessica submits her application. Over the course of Jessica’s prospective student journey, chances are she’ll be exposed to a digital advertisement, a web page, a guide, an email, a phone call, a direct mailer, and even a billboard. Each of these tactics are given the same amount of credit.
This model is better (and more complex) than anything previously mentioned because it considers all touch points across Jessica’s prospective student journey. But where it falls short is in allotting equal credit to all touch points. Usually, there are some key steps in the prospective journey that are unequivocally more important than others.
This multi-attribution model gives 22.5% of credit to the first engagement, 22.5% to new contact creation (new contact added to your CRM), 22.5% to application started, 22.5% to last engagement, and the remaining 10% is split equally among other engagements over the course of the prospective student’s journey.
This model succeeds because it helps effectively answer questions like, what is driving prospective students to my site? What is getting visitors to formally inquire and download educational guides? This model gives some credit to each engagement in the prospective student’s journey, and it gives more credit to certain conversion points that result in a greater likelihood that the prospective student formally applies to your program.
The bad news is that this model is hard to achieve! It requires deep analytical insights into your prospective student’s journey and data at each engagement point along the way. A typical CRM won't provide those insights, nor will a traditional email automation tool.
Some companies try to get attribution reporting down to an exact science, using algorithms, AI, and complex data crunching to figure out which marketing channels provide the best ROI. At the end of the day as marketers, we’re all trying to make grand assumptions about the points in which our audiences consciously make decisions to purchase a product (or start an application).
In the education industry, there are certain engagement and conversion points that significantly increase the likelihood that a prospective student will formally apply. For example, inquiring about a program, attending a campus event, or speaking to an admissions representative are high-intent actions, and should be given significant credit because they, more often than not, result in applications. However, not every new prospective student is ready to take one of these high-intent actions, so attracting and nurturing them is just as important.
Truth is, there’s no one-size-fits-all attribution model. You should, however, use a multi-attribution model, where more credit is given to more important steps in the prospective student’s journey. Depending on your school’s marketing approach, the best attribution model falls somewhere in between the linear and full-path attribution model. Using one or a combination of these multi-attribution models is key to making clear, evidence-based evaluations of the effectiveness of your marketing tactics.