“We recommend that you confirm important info with the school, like costs and requirements.” That line comes from Google’s own Search Help page on college information, and it is an admission worth reading twice. Google shows prospective students a box of facts about your institution, tells them those facts may be wrong, and sends them to you to check. Most applicants never do. They read the box and move on.
A Google knowledge panel for universities is that box: the logo, the description, the founding date, the president, the costs and the outcomes that appear when someone searches a school by name. Google decides what goes in it. Your job is to make sure every source Google reads tells the same, current story, and then to claim the panel so you can correct it directly.
This guide to the Google knowledge panel for universities covers what the panel shows, where its facts come from, a scoring model for the six records that control it, and the exact claim and feedback process.
What does Google show when applicants search a university?
Google describes knowledge panels as information boxes for entities in its Knowledge Graph: people, places, organizations and things. A university is an organization and a place at once, which is why its panel can carry a map, an address and campus photos next to a founding year and a list of notable alumni.
For four-year schools in the United States, Google added a layer in June 2018. In a post on Google’s blog, product manager Jacob Schonberg announced that searches for a college would show the average cost after financial aid, broken down by household income, along with graduation rates, typical annual income ten years after enrollment, undergraduate enrollment, student body statistics, notable alumni and similar colleges. The screenshots in that announcement used UCLA, Spelman College and Yale as examples.
Those numbers do not come from your admissions page. According to Google’s Search Help documentation, most US college data comes from two US Department of Education projects. The first is IPEDS, the Integrated Postsecondary Education Data System, whose questionnaires every institution in federal student aid programs must complete. The second is the College Scorecard, which combines that institutional reporting with federal financial aid data and Treasury tax information. Google also notes that some outcomes shown are recorded for only a subset of students.
The rest of the panel, the description, the president, the founding date and the logo, comes from what Google calls “a variety of sources that compile factual information,” plus facts supplied by verified representatives who claim the panel.
This matters to applicants because the panel is often the first thing they see, and on a phone it can fill the screen. It matters even more now that AI answers draw on the same entity understanding. When an assistant is asked which colleges in Georgia have strong engineering programs, it relies on what it knows about each school as an entity. A university that Google confuses with its own medical center, its athletics program or a school with a near-identical name starts that comparison at a disadvantage.
The common failures are easy to spot. A panel that shows a president who left two years ago. A description lifted from an old Wikipedia revision. A map pin on the hospital instead of the main campus. A logo from before the rebrand. Or no panel at all, with the business school’s listing appearing in its place.
Wikipedia, Wikidata and federal records set the facts
Google gains confidence in an entity when independent sources state the same facts in the same way. For a university, three kinds of records carry most of the weight.
The first is Wikipedia. Most established universities already have an article, and its infobox (name, motto, type, founding date, endowment, president, enrollment, location) reads like a template for a knowledge panel. You cannot write or edit your own article as if it were marketing copy. Wikipedia’s conflict of interest guideline asks people connected to a subject to propose changes on the article’s talk page rather than make them directly, and every fact needs a reliable published source. What you can do is publish the sources: a dated press release for a new president, an annual fact book, an accreditation letter. Editors need something to cite.
The second is Wikidata, the structured database behind Wikipedia’s infoboxes and a common reference for search engines and AI systems. A university’s Wikidata item holds typed statements: instance of university, inception date, country, official website, logo, head of the organization and a long list of external identifiers. One of those identifiers is property P1771, the Integrated Postsecondary Education Data System ID, which Wikidata describes as an identifier for colleges and universities in the United States. The University of Southern California’s item, for example, carries IPEDS ID 123961. That one number ties the Wikidata item to the federal record Google uses for costs and outcomes. If it is missing, the two records have to be matched by name, and names are where confusion starts.
The third is the federal record itself. Your institutional research office already submits IPEDS data every year. The name, address and website in that submission should match what is on your homepage and in Wikidata, character for character. If your official name in IPEDS is different from the name you use in marketing, list the marketing name as an alternate name everywhere else, so Google can see both belong to one school.
Score your campus with the Registrar Score
I audit every institution with a model I call the Registrar Score. A registrar’s job is to keep the official record straight, and that is the job here. The score covers six records. Each one earns 0 points if it is missing or wrong, 1 point if it exists but conflicts with another record, and 2 points if it is complete and matches the others. A perfect score is 12.
Record one is the Wikipedia article. Two points if it exists, its infobox is current (above all the president and the official website) and its claims cite sources you can point to.
Record two is the Wikidata item. Two points if it carries the correct instance type, the official website, the current head of the organization with a start date, the logo and the IPEDS ID. Check for duplicate items too. Two Wikidata items for the same school split Google’s confidence in half.
Record three is the federal record. Two points if your IPEDS and College Scorecard listings show the same legal name, address and website as your homepage. This is the record behind the cost and outcome numbers, so errors here reach applicants first.
Record four is your own markup. Schema.org defines a type called CollegeOrUniversity, a subtype of EducationalOrganization, for “a college, university, or other third-level educational institution.” Two points if your homepage carries CollegeOrUniversity markup with your official name, alternate names, URL, logo, address and sameAs links to your Wikipedia article, Wikidata item and official social profiles. Schema also gives you the department and parentOrganization properties, which let a medical school or law school point back to the university as its parent instead of competing with it.
Record five is the Google Business Profile for the main campus. Two points if the main campus profile is verified, uses the official name, carries the right category and pins the right building. This is also where department confusion lives. Google’s Business Profile guidelines allow departments within universities to have their own profiles, provided each department’s name and primary category differ from the main institution’s and from each other. Its own example of an acceptable department listing is “Massachusetts General Hospital Department of Dermatology.” The problem comes when a department uses the university’s bare name, or when the admissions office and the athletics department both claim the main campus address under similar names. Google then has several candidates for “the university” and may pick the wrong one.
Record six is the claimed panel. Two points if someone at the institution has claimed the panel and checks it at least once a term. One point if it is claimed but nobody looks at it. Zero if no panel exists or nobody has claimed it.
A score of 10 to 12 means Google has a clean picture and you are maintaining it. A score of 6 to 9 means the panel likely exists but contains at least one stale or conflicting fact. Below 6, expect the wrong panel, a thin panel or none at all. Fix the lowest-scoring record first, because a conflict in one source pulls down confidence in the others.
Claim the panel and send feedback that sticks
Claiming is the step most universities skip, and it is the one that gives you a direct line to Google. According to Google’s Get verified on Google help page, the process works like this. Sign in to a Google Account. Search for your institution on Google and find its panel. At the bottom of the panel, select “Claim this knowledge panel” (on mobile, it sits in the menu). Review the information shown, then sign in to one of the official sites or profiles Google lists. Google names YouTube, Search Console, Twitter and Facebook as the accounts it accepts. If Google cannot match you to any of them, it asks for more information to verify you.
For a university, the cleanest route is Search Console for the main domain or the institution’s official YouTube channel. Decide in advance which office owns the claim. Communications teams are the natural home, but write down which login holds it, because staff turnover is the most common reason a claimed panel becomes orphaned.
Once verified, you can suggest edits to the panel directly, and Google treats those suggestions as coming from the entity itself. It still checks them against other sources, so an edit that contradicts Wikipedia, Wikidata and IPEDS is less likely to stick than one that matches them. Update the sources first, then suggest the edit.
Anyone can also send feedback without claiming. Google’s college information help page says to search for the college, select Feedback at the bottom of the panel, describe the error and send it. That route is useful for fast corrections, like a wrong phone number, but it carries less weight than a verified suggestion. For errors in costs or graduation rates, the fix belongs at the source: correct the IPEDS submission through your institutional research office, because that is where Google got the number.
Spelman College shows why the work never ends. The Atlanta school, founded on April 11, 1881, was one of the three examples Google used when it launched its college panels, with cost after aid, graduation rates and earnings drawn from federal data. Its presidency has changed hands three times since then. Spelman announced in October 2024 that its 11th president, Helene Gayle, was taking a leave of absence, and alumna Rosalind Brewer became interim president. On June 5, 2026, the board named roboticist Ayanna Howard its 12th president, effective August 1, 2026. Wikipedia’s infobox now lists Howard, and Spelman’s own president’s office page names her. Every one of those changes had to travel through Wikipedia, Wikidata and the school’s own pages before a Google knowledge panel for universities like Spelman could show it. A school that publishes the change in all six records the week it happens gets an accurate panel. A school that waits gets the wrong president in front of every applicant who searches its name.