By Tracey Halvorsen

When I built Blanding, I wanted to make something visible that I had been talking about for years: how much higher education websites look and sound alike. Put enough of them next to each other and the vocabulary starts to feel awfully familiar. Everyone has a vibrant community. Everyone is preparing the leaders of tomorrow. Apparently, tomorrow is going to be extremely well supervised.

An audit that flags that language gives people a way into the conversation. But the more time I spend with the questions underneath Blanding, the less satisfied I am with stopping at the words. A familiar phrase can conceal something genuinely valuable that hasn't found its way into the explanation. It can also be the approved substitute for a decision nobody wants to make. Counting the phrase won't tell us which problem we have.

That's where my thinking about the next version has been going. Before asking how to make a claim sound more distinctive, I want to know what the institution has chosen, what experience that choice produces, and why it would matter to the person considering it. Sometimes the work is better writing. Sometimes we need to go find the people doing something worth writing about. And sometimes we need to get the people with the authority to make a decision back into the room.

That last possibility gets expensive when we keep treating it as a website problem. It becomes more urgent when someone can get an AI-generated answer assembled from several sources without ever following the path we built for them.

Bobby Gillespie recently posted that we need a new name for brand strategy. His argument places it within corporate strategy, connecting direction, positioning, audiences and decisions to the reputation a business earns. He uses a river to explain how far upstream that work sits from visual identity.

In the comments, Boris Alexandrov describes a practical problem: clients will fund tangible identity work while leaving strategy unfunded, so designers end up filling the gap. He argues for strategists and creatives working together from the beginning.

I find that last point particularly useful. Design can expose a strategic problem long before anyone agrees what to call it. Trying to organize a website, explain a program or make a promise tangible forces questions that can remain comfortably abstract in a presentation. Creative work is part of how we discover what we mean.

The question is who can act on what we discover. A designer can notice that a promise of personal attention isn't supported by the experience being described. They probably can't change advising capacity, staff schedules or the budget. If the strategy requires a particular experience, somebody has to decide what providing it will take and whether the organization is willing to do that.

When the issue is an unmade choice, we should acknowledge what choosing involves. Giving one audience priority means accepting that another is less central in that context. A position worth pursuing may require funding one experience before another and deciding what the institution can sustain. The people in those conversations have real responsibilities and real stakes. AI can lay out options and challenge assumptions. It can't accept those tradeoffs on the institution's behalf.

I'd rather have that conversation than spend another six months making an unresolved promise look more convincing.

One of the things I've had to keep checking in this work is the assumption that difference is always desirable. It isn't. Please keep calling the admissions section Admissions. Apply Now is doing a perfectly reasonable job. Nobody needs to solve a riddle to find the application.

Think about choosing a car. You expect familiar controls, and you need comparable information about price, range, space and safety. Then you start weighing the combination against your own life. A feature doesn't have to exist in only one car to matter to you. You also don't need the manufacturer to reinvent the steering wheel to prove it has a point of view.

The same distinction matters on a university website. Tuition, requirements, schedules and qualifications need to be easy to find and compare. Familiarity helps. The harder question is what the institution wants someone to understand about the combination of experiences it offers, and whether that combination is valuable relative to their actual alternatives.

Take personal attention. Suppose a program gives every student an adviser, offers evening appointments and builds advising into specific points in the semester. Those are things we can describe and investigate. They might matter a great deal to a student working during the day. They might also be available at every other program that student is considering. Establishing that a promise is true doesn't automatically make it a reason to choose you.

And we can't settle that by swapping logos in our heads. That's a useful first provocation, but eventually we need to understand what people are choosing between and what they care about. An audit can identify what is visible on the pages it reads. It can't establish what students experience, how they perceive the institution, or why an internal decision was made.

This is where AI makes the brand-strategy conversation more consequential for me. We have spent years building websites that introduce an organization, establish a feeling, guide people toward relevant information and help them act. Visitors have always arrived halfway through that experience through search, a shared link or a recommendation. AI adds another way for the information to be encountered, selected and recombined outside the experience we designed.

Google describes its AI search features as sometimes making multiple related searches across topics and sources to assemble a response. A person can ask a complicated comparison question and receive an answer drawn from several places, with supporting links. Pew's analysis of U.S. browsing activity in March 2025 found that visits with AI summaries were associated with fewer clicks on traditional results. That doesn't tell us what every student will do, but it gives us a reason to question how much of our carefully planned journey someone will ever see.

Imagine the working student asking which programs they can realistically complete around a full-time job. Their answer might draw on a program page, course schedules, tuition information and something said elsewhere about the experience. They may never see the homepage video or the paragraph that explains what the institution believes about supporting students.

Does the promise still make sense in the pieces? Can someone find the conditions that matter alongside the attractive claim? If one page promises flexibility and another describes attendance requirements, will a person be able to understand how those fit together?

I don't think the answer is to flatten a website into a pile of facts for machines. A place can have a character worth experiencing. Design helps people feel that, explore it, understand information and decide whether they trust it. A website still has work to do when someone wants to verify an answer, talk to a person, visit or apply. Accessibility and clear interactions matter throughout. Declaring the website dead would let us avoid some fairly interesting design problems.

This is about the relationship between the institution as a whole and its individual expressions. A coherent brand should give different programs room to explain their value to different people. The program for working adults needs different evidence than the residential undergraduate experience. Those differences can make the institution easier to understand, provided they connect to choices it actually supports. Repeating the same institutional paragraph across every page doesn't accomplish that.

There are no guarantees here. AI can miss information, misread it or assemble a misleading answer, and it uses sources beyond the ones an institution controls. Google explicitly says meeting its guidance doesn't guarantee inclusion. Clear evidence gives people and systems something more useful to work with; it doesn't purchase a recommendation. We still have to look at what is actually being said about us.

For teams trying to figure out what to do with all of this, I'd start with one promise in one program, process or student experience. Choose something a prospective student might genuinely use to make a decision. Find out what the institution does to make it true, who can experience it, what limits apply and what evidence is available. Then look at how that information appears across the relevant pages.

AI can help with that investigation. Give it the source material, ask it to find patterns or contradictions, and make it show where each observation came from. Check its work. Something absent from the supplied pages may be happening every day on campus. Something described beautifully may be happening much less reliably than the copy suggests.

Bring the result to people who can judge it. Current students can help us understand whether the description rings true. Prospective students can tell us what they understand, why it might matter and what they still need to know. If we want to learn whether they recognize or remember the institution, we need to ask questions designed for that. A better click rate answers a different question.

Then decide what needs to change and who has the ability to change it. That might be a page. It might be the way departments share information. It might be the experience itself.