Over the last five years, there has been a profound shift in higher education away from brick-and-mortar universities to the digital classroom. A few years ago, only about 20 percent of overall higher education enrollments were through self-paced online classrooms. According to research, in the last year, as the pandemic forced the adoption of remote or hybrid learning, the education sector has in a decline of enrollment.
Two key themes have emerged from this shift. While the pandemic is a major cause of the increase in online enrollment, students are choosing online studies because of the ease of access and reduced cost. And, because online courses have drastically reduced these cost barriers, new institutions can now compete with more established schools. In some instances, these new digital universities are more agile, which allows them to grab a lion’s share of prospects.
The other theme we’re seeing is a less complex enrollment process, which reduces decision time. Not only is a prospect visiting your site more informed than in the past, but the time they take to decide on a school has almost halved. This is a critical insight for the digital marketer. It’s essential for marketers to identify highly interested and convertible prospects during the same website visit. One cannot rely on offline engagement and nurturing processes to drive conversions.
To truly earn more conversions, marketers have turned to personalized content that uses a visitor’s real-time intent score.
What is Intent Score
Intent Score is a mechanism that gauges how willing a visitor will complete a conversion goal, such as a request for more information. The score is calculated based on their website visit behavior, channel of entry, and other contextual information. The score is indicated by a number between zero and one.
When you know the visitor’s conversion probability, they can be classified into three categories: Low Intent, Medium Intent, and High Intent. Obviously, visitors in a high intent bucket have a better chance of conversion than those in a low intent bucket. From there, these segments can be further broken down into lookalikes: Similar visitors that share multiple intent characteristics.
How does this benefit a higher education marketer?
As a visitor’s activities are analyzed, and their intent is determined, a personalization treatment can be deployed to accelerate the visitor’s journey towards the conversion goal.
But, beyond real-time personalization, there are additional benefits of the intent score for the higher education marketer:
What do I need to build a Scoring Model? Data Prerequisites
First-party, log level data using Google Analytics 360 or Adobe’s tool
Cloud-based Computing Resources
The core computing part can include Compute Engines, Kubernetes, or Container Orchestration services and databases. These can handle your data preprocessing, model training, and model hosting for intent prediction.
Website experience optimizer
This is a tool or collection of tools that dynamically customize a user’s experience for real-time personalization.
A team of UX experts can tweak and optimize your model based on user study. This team can also recommend personalization options.
Case Study: Online higher education institution identifies issues with internal search Deployment
Using Intent Score, web visitors were grouped into three buckets (Low Intent, Medium Intent and High Intent). Each group was separately targeted based on their behavior. For example, if a user in a medium intent bucket was a lookalike to the user in a high intent bucket, the missing behavior of the former was artificially induced to push them into the high intent bucket.
In real time, the model was deployed and integrated alongside the website experience layer. When a user arrived at the website, a custom code would capture their behavior, and send that as a request to the real-time API. This API processed the request, calculated the Intent score, and sent back the score.
The outcome of the Intent Score was to classify the visitor into an action or treatment group. This grouping feeds into CRO platform and CDP for website personalization.
Intent score and Lookalike Segment Data are also sent to a custom dimension in Google Analytics, which was used to create audiences for Google Display and Google Adwords.
The score can be visualized in a dashboard and compared with other leading indicators. You can also find abnormalities or quality issues within the up- and down-stream processes.
Result: UX Optimization using Intent Score
The higher education institution’s website employed specific pages with three drop-downs menus that filter the specific degrees based on Degree level, Areas of Study, and more. Lookalike analysis revealed that medium intent users spent more time on certain pages compared to high intent visitors.
The correlation analysis revealed that the more time spent on specific pages decreased the intent to convert. We can conclude that some of the users were struggling to find the information they were looking for using the internal search tool. Prior to this discovery it was believed that, because the search tool was used a lot, it was a great influencer and contributor to the conversions. This triggered the business decision to conduct UX research and revamp the entire search experience.
Today’s higher education marketer needs real-time content personalization to convert as many website visitors as possible. By using the website behavior and the other contextual information to develop an intent score, marketers can deploy a successful marketing strategy, and generate tangible business benefits.