How We Got Here, And Why We Get It
The dean who built that rogue microsite? She needed to recruit for a new program and couldn’t wait eight months for the official process. The department that published contradictory requirements? They were responding to a faculty request with a 48-hour deadline and no clear guidance on where the information should live. The outdated PDF still ranking in Google? Someone created it to solve a real problem three years ago, and nobody was ever assigned to maintain it.
These aren’t failures of individual judgment. They’re the result of systems that were never designed to handle the volume and complexity of modern content needs.
Most higher-ed web and marketing teams didn’t have (and some still don’t have) a seat at the strategic table. They’re handed mandates without authority.
The path of least resistance becomes the only path. Over the years, those paths accumulate into a tangled mess that nobody wanted, and everybody inherits.
So when I talk about “starting with why,” I’m not pointing fingers at the people managing these sites. I’m trying to offer a framework that might make this all feel achievable.
The Real Breakdown Happens Before Anyone Touches the CMS
An institution can create the most beautiful sitemap in the world. It can have the most considered taxonomy, labels, menu hierarchy…. but then something happens.
Someone requests a page.
A new academic program needs a landing page. A dean wants to showcase an initiative. A faculty member believes their research deserves its own microsite. An administrative office has a policy change to communicate. The request arrives, and… then what?
Usually, the content just gets created. Nobody stops to ask: Why does this need to exist? What job is it supposed to do? Who’s it actually for? How will it stay accurate? Where does it fit with everything else we’ve already published?
Those questions don’t get asked because there’s no time to ask them. Because the requester has more institutional power than the web team. Because “just make the page” is faster than a philosophical discussion about content strategy when you’ve got twelve other requests in the queue.
We get it. We really do.
But here’s what’s changing: AI is about to make all of this visible in ways it never was before.
"AI search visitor (tracked to a non-Google search source like ChatGPT) is 4.4 times as valuable as the average visit from traditional organic search, based on conversion rate."
AI Doesn't Navigate Like Humans
For years, institutions could live with content contradictions because students navigated linearly. They clicked through menus, used site search, maybe stumbled on a PDF buried three levels deep. If two slightly different versions of program requirements existed somewhere on your site, most students would only ever encounter one.
AI search flips that completely.
When a prospective student asks ChatGPT or Perplexity about your application requirements, the AI doesn’t browse your site the way a human would. It uses what’s called a “fan-out” method—simultaneously scanning your admissions page, department sites, faculty pages, archived content, and those PDFs nobody remembers uploading. It gathers information from everywhere, then synthesizes it into a single answer.
The AI has no idea which version is authoritative. It doesn’t know that the official policy lives on the registrar’s page while the department’s version contradicts it. It just surfaces what it finds, and the student has no way to know which answer is correct.
This is actually a bigger deal than it might sound. US News and World Report’s 2025 research found that 47% of Gen Alpha prospective students are already using AI search to compare schools. We wrote more about how content sprawl affects your AI visibility and what you can do about it—but the short version is that every structural weakness, every “we’ll clean that up later,” every compromise you made because you didn’t have the authority to push back is now showing up in the answers students are getting about your institution.
This isn’t meant to make you feel worse about an already difficult situation. It’s meant to give you ammunition. Because now you can point to something concrete when you’re trying to explain why content governance actually matters.
Semrush research found that most of ChatGPT's citations came from URLs ranking beyond position 21 on Google—well past page two of traditional results.
A Simple Sequence That Actually Helps
When things feel chaotic, it helps to have a mental model. Here’s one that’s worked for a lot of teams I’ve talked to:
Start with why. Before content gets created, someone needs to be able to say what it’s for. Not “to inform people” (that’s not a purpose, that’s a tautology). What specific outcome? What decision is it helping someone make? What happens if this content works?
Then ask who. Not “who internally wants this”—that’s usually where things start, and it’s backwards. Who externally needs this? Prospective students? Current students? Parents? Alumni? If you can’t name a real audience with a real need, the page probably shouldn’t exist.
Then figure out how. Once you know the purpose and the audience, structure follows naturally. Where does this logically live? What template fits? Who needs to review it? How does it connect to related content? Who’s responsible for keeping it accurate?
The problem in higher ed is that this sequence almost always runs in reverse. It starts with WHO (a stakeholder with more authority than you demanding a page) → HOW (whatever workaround gets them off your back fastest) → WHY (never articulated because by then the page already exists and everyone’s moved on).
That’s not a criticism of the people stuck in that pattern. It’s a description of what happens when web teams don’t have the institutional authority to do things differently.
A Tool That Might Help
If you want to start shifting the pattern, even a little, here’s something concrete you can try.
Before any new page gets approved, see if your team can answer five questions in under a minute:
What’s the goal? Specific outcome, not “provide information.”
Who’s the primary audience? Be specific. “Prospective students” isn’t enough. Which ones?
What should they do after reading this? Apply? Register? Call someone? Download something?
Where does this fit in their journey? Are they just learning about you, or ready to make a decision?
Who owns this going forward? Actual name, actual office. Who updates it when things change?
If you can’t answer all five clearly, that’s a signal. Maybe the page shouldn’t exist yet. Maybe it needs more thinking. Maybe there’s existing content that should be updated instead.
This won’t solve everything. You’ll still face pressure to publish things that don’t pass the test. But having a simple framework gives you something to point to. “We started using this process and this is why” is an easier conversation than just “No.”
The Stuff That Piles Up
You probably already know what’s lurking in your site. Pages get almost no traffic, not because they’re hidden, but because nobody’s looking for them. They existed to serve a very specific purpose, and were forgotten. The same degree program is described three different ways on three different pages, each created by someone trying to help, none of them coordinated. Bits of scholarship information are scattered across a dozen URLs, admissions, financial aid, and in various PDFs. When the numbers change, updating all of them takes half a day.
This stuff accumulates because there’s never enough time to clean it up. Everyone’s too busy responding to the next request. The content debt builds up.
Moving the Needle
We’re not going to pretend there are easy fixes, but here are some things that have helped other teams in similar situations:
Change how requests come in. If your intake process starts with “what do you need and when,” try shifting it to “what problem are you trying to solve and for whom.” It’s a small change, but it creates space for different conversations. Sometimes requesters realize they don’t actually need a new page, they need an update to something that already exists.
Document not just what exists, but what belongs where. Most site documentation consists only of a site map and content entry instructions. That’s not enough. If you can write down the rules, what kind of content goes in each section, when it’s okay to create something new versus updating what’s there, what level of duplication is acceptable (usually: almost none)—you have something to point to when requests come in that would break the structure.
Build constraints into the CMS where you can. Heavy governance policies fail because they require political capital most web teams don’t have. But if the system itself enforces patterns, templates that don’t allow certain kinds of drift, fields that pull from a single source of truth, compliance becomes the default. People will work around anything that requires extra effort. Make the right thing the easy thing.
Pick one category of information and fix it completely. You can’t overhaul everything. But you might be able to consolidate all tuition information into one authoritative source. Or all application deadlines. Or all program requirements for one college. When something changes, you update one place. When AI retrieves that information, it finds one consistent answer. Small wins build credibility for bigger conversations.
Cut pages before you redesign. If there’s a redesign on the horizon, push for a serious content audit first. Archiving pages that serve no user need, consolidating duplicates, redirecting outdated URLs—that work is less glamorous than a visual refresh, but it matters more. Redesigning a cluttered site just gives you prettier clutter.
Communicate and educate. Working at an intitution of higher learning means that you are surrounded by a lot of brilliant people. Many of these people do a lot of writing and communicating as part of their work.
What This Is Really About
This is about publishing with purpose.
When content has a clear reason to exist, when everyone knows what it’s for and who it serves, the downstream problems get smaller. Navigation makes more sense because things have logical homes. Governance becomes possible because people understand what they’re being asked to maintain. Volume shrinks naturally because purposeless pages don’t get created in the first place.
And when AI systems crawl your site looking for answers to give prospective students, they find consistent information instead of contradictions you never intended to publish.
None of this is easy. The structural pressures in higher ed—the diffuse authority, the competing stakeholders, the chronic under-resourcing of marketing and web teams—those are real, and they’re not going away.
But if you can start asking “why does this need to exist?” before things get created—even some of the time, even imperfectly—you’re moving in a better direction.
The question is whether you can get to a place where every page on your site has an answer to: Why is this here, and who does it serve?
If you can answer that in 60 seconds, you’re doing better than most..
