Personalization: a definition and how it is solved

As the topic of personalization has developed alongside AI, it’s become clear to me that higher ed institutions—when it comes to their martech stack—need a better definition of what personalization really means.

I prefer to think of personalization as existing across four layers:

1. Unknown and Unauthenticated
2. Known and Unauthenticated
3. Authenticated
4. Authenticated and Known Deeply

Unknown and Unauthenticated

In this stage, you don’t know the identity of your user, and they don’t have any type of authenticated account with a login or verification process. This is where your identity resolution and tracking process begins. Many vendors operate in this space using technology that places a cookie in a user’s browser and later attempts to identify users by monitoring form submissions. Often pre-packaged, these solutions can require an institution to rely on a third-party form submission tool—which, in my view, is more hassle than it's worth. There are also other paths to identity resolution.

Known and Unauthenticated

There are multiple ways to use web and martech tools to discover the identity of website users over time, especially in higher ed. This requires the right tech stack (inquire at Fathom This to learn more). Once you’ve identified a user, they move into this category: known but unauthenticated. While mostly reliable, this step still faces challenges like cross-device tracking, privacy tools (think Apple Mail), and more. But a link can be made, deepening the user relationship.

Authenticated

Your user has created some type of account. For example, a Slate account they use to apply to your institution. At this point, deeper tracking becomes possible. You know who the user is, can track activity across devices, view their interests, and identify when they’re active on your site. In my experience, this is where a lot of opportunity to learn about users is left on the table, mostly because of a lack in system interoperability and integration.

Authenticated and Known Deeply

By now, you've collected a wealth of data about your user. You should be able to personalize interactions more effectively—but this often falls short in higher ed. Admission communications teams tend to personalize through broad segmentation (e.g., emails to all applicants or all transfer students). While we have access to more user data, it doesn’t mean much if personalization efforts don’t reach users in a meaningful, 1:1 way.

How does AI fit in?

I often ask: how does AI fit into this personalization model, and how does it provide value? The truth is, AI is evolving rapidly, and we're still figuring it out. But as you narrow in on a user’s identity, web and martech managers can extend system capabilities to provide true 1:1 communication.

You can target the specific needs and problems of the people you know. AI can help web and martech managers:

• Decipher user needs
• Answer specific questions using contextual data (think: personalized chat)
• Refine web searches to return highly relevant results
• Filter through personalization data noise

AI-powered tools can also be deployed to optimize customer and student interactions.

Getting to this point requires a tech stack that brings it all together—including your CRM (Slate or otherwise). If you're interested in learning more about systems like this and personalization in higher ed, contact us.