It’s the perennial struggle – the nagging question that keeps higher education marketers awake at night: Am I really making the best use of my marketing dollars?
An accurate attribution model is more critical than ever as marketing channels grow in number and complexity and make it difficult to discern where your marketing spend ends up. Yet the complexity of modern marketing isn’t the only challenge to proper attribution.
For higher ed marketers especially, traditional channels are heavily influential in the conversion process. But these channels are not usually trackable in attribution models, which begs the question – exactly how influential are traditional channels?
Without having an exact picture of how your traditional channels contribute to conversion, you’re operating largely on trial and error when allocating your marketing spend. In fact, you may be wasting incalculable amounts by pouring money into a channel that’s not performing as you believe it to be – or you may be overlooking your most influential channels in favor of nonperformers.
However, these risks can all be mitigated through proper channel attribution. Let’s dig deeper into the question of attribution models and how they can help you form a clearer picture of your marketing performance.
A Little Background
You’re likely already familiar with the most common types of attribution models: first touch, last-touch, and linear (or equal). These can give you a general idea of which channels are involved in conversion, but unfortunately, they can’t really describe accurately the extent to which each channel contributes to conversion.
The problem with these models is that they assume each interaction is a conversionbased engagement, but this is not always the case. Each touchpoint is an opportunity for influencing a decision – but where attribution models often break down is that they cannot assign a dollar value to each point of interaction. They simply aren’t able to examine the influencers on such a granular level.
And, more broadly speaking, these standard models are effective for tracking digital channels, but less so for traditional channels. For schools that use a large mix of direct mail, TV ads, and other traditional outlets along with digital, this creates a serious predicament. As the famous John Wanamaker quote goes, “Half the money I spend on advertising is wasted; the trouble is, I don't know which half.” This is the situation for many institutions that aren’t able to track the ROI of their traditional channels.
Boiling the issue down to this level, you discover that a completely new attribution model is needed – one that both takes traditional marketing into account, and gives a more accurate picture of which channels are contributing to conversions (and how much they contribute). There’s only one solution that fills all these needs: a custom model that is tailored to your institution.
FIRST TOUCH:
A model in which the first touchpoint in the conversion process is given all the credit for the conversion
LAST TOUCH:
A model in which the last touchpoint in the conversion process is given all the credit for the conversion
LINEAR:
A model in which each touchpoint of a conversion is credited equally
Building the Perfect Model
Now, lest this seem like an easier-said-than-done solution, we’ll demystify the process a little bit by describing the ideal custom model. To be effective, a custom model must be data-driven, cross-channel, variable, and constantly optimizing. Let’s break down each of these tenets:
1. Data-driven. A custom model must begin and end with data. In-depth behavioral data and other research will give you a “baseline” for how students interact with your institution, and this will be the springboard from which you develop your model. Gathering and analyzing this data gives you the background you need first to create the model (point 3), and then to optimize it (point 4).
2. Cross-channel. The pain of trying to track and then attribute success to traditional channels is one that marketers feel daily. However, by linking traditional channels to digital, you create touchpoints that are actually trackable and can be incorporated into your attribution model. Then you can finally and decisively determine how your traditional channels fit into the mix.
3. Variable. The downside to many multi-touch models (i.e., the linear model we discussed earlier) is that the weight assigned to touchpoints is fixed. Different models may assign value differently, but that weight is generally a fixed percentage. For your custom model, you need a variable weight that is extrapolated from your data, and which can be updated and refined according to new data that you receive.
4. Optimizing. Finally, a custom model is only strong if it is able to adapt. The more you track how your students interact, which channels they touch most, and which channels hold the most influence, the more accurate your model will become. In short, testing, tracking, and analyzing data lead to a smart attribution model – one that changes as your prospects do.
Assessing the Outcome
The constant data-gathering and tracking that go into developing a custom model may seem highmaintenance. And it’s true that an effective attribution model cannot remain static, but needs to be consistently tweaked for maximum efficiency. However, the benefits of custom models are so great that any ongoing effort pales in comparison.
With a custom attribution model in place, one that is able to track traditional and digital channels accurately, you curb the need for constant trial and error. You can know with certainty where your money is going, which channels are performing, and how to shift your focus to better meet your target audience. Gone is the guesswork – instead, you gain a tool that has the ability to help you achieve a more effective marketing spend, a better return on investment, and greater peace of mind.
913.648.8333