I have been thinking about why technology in higher education can become so complicated even when the individual processes do not initially appear complicated at all.
A university admits students. It advises them. It manages financial aid, academics, communications, advancement and eventually alumni relationships. None of these things are particularly difficult to understand independently.
The difficulty appears when we consider that the same person moves through all of them.
A student does not experience herself as an admissions record and then become an advising record. She does not become a different human being when she graduates and enters an advancement system. Her relationship with the institution is continuous even when the institution’s information about her is not.
I think this distinction is important to understanding what Salesforce can actually do well in higher education.
There is a tendency to describe institutional technology according to what the technology contains. A CRM contains records. An integration moves data. A dashboard displays information. An automation performs an action when particular conditions are met.
All of that is true, but it does not tell us whether the institution understands what the information means.
This is where I think Salesforce implementations can go wrong.
An institution can have an extraordinary amount of information about a student and still have a remarkably incomplete understanding of her experience.
Admissions may know one thing. Advising knows another. Financial aid knows something else. A faculty member may possess information that exists nowhere in the formal system at all. Each department can therefore possess an accurate fragment while the institution as a whole possesses an inaccurate picture.
More information has not necessarily produced more understanding.
It has produced more information.
The distinction seems obvious when written this way, but I do not think we take it seriously enough when designing systems.
Every Salesforce implementation eventually requires somebody to decide what distinctions matter. We decide what deserves to become an object, what becomes a field, what constitutes a relationship, which events should trigger an automation and which people should have access to which information.
These appear to be technical decisions because they are implemented technically.
They are also interpretive decisions.
We are constructing a model of how an institution works.
This is why I do not think good Salesforce architecture begins with Salesforce.
It begins with the institution.
How does this university actually understand a student? Where does information originate? When does its meaning change? Do two departments use different words for the same thing? More dangerously, do they use the same word for different things?
What information is technically available but practically invisible?
What information is repeatedly recreated because nobody knows it already exists?
Which processes are necessary and which processes exist because somebody designed a workaround ten years ago and everybody eventually forgot that it was a workaround?
These questions become even more important with AI.
There is enormous interest in bringing generative AI into higher education, and much of that interest is justified. A person should not necessarily have to search through several systems, records and documents to reconstruct information that the institution already possesses.
But an AI model does not magically make institutional information coherent.
If the underlying architecture contains conflicting definitions, missing context or poorly governed information, AI can simply become extraordinarily good at presenting that confusion back to us in natural language.
That may actually make the problem harder to detect.
A badly designed report looks like a badly designed report.
A beautifully written AI response can make uncertain information sound understood.
This is why I think the interesting future of Salesforce in higher education is not simply automation.
It is institutional understanding.
The question is not how much information an institution can collect about a student. Universities already possess enormous amounts of information.
The better question is whether the architecture allows that information to become meaningful at the moment another human being needs to understand it.
That requires technical architecture, certainly.
But it also requires language, institutional knowledge, judgment and an unusual amount of listening.
At North Shore, this is the part of Salesforce work that interests us most.
Technology does not become intelligent merely because it contains more information.
Sometimes intelligence begins with understanding what the information actually means.