Contents
Key takeawaysWhich vendors drive spendHow metrics inflate the billWorked exampleCampus licensing trapsWhat we have seenCheck your own positionCutting spend at renewalSmall college or large systemRenewal timelineWhat to do nextFAQUniversities license Oracle, Workday, Microsoft and Adobe on headcount metrics that overstate real use. Reclaim alumni accounts, align counts to active users, separate research entitlements and time renewals against the academic year.
- Five vendors carry the bill. Oracle, Workday, Ellucian, Microsoft and Adobe account for most university software spend.
- Metrics count bystanders. Vendors price on total enrollment, total employees or FTE counts that include people who never use the system.
- The overcount is large. In our higher education work, licensed counts ran 15 to 35 percent above active users on headcount based metrics.
- Alumni seats pile up. Alumni and inactive accounts inflate named user counts unless they are reclaimed every term.
- Research belongs apart. Grant funded use often carries its own entitlements and should come off the central agreement.
- Read the definition. Oracle's Enterprise FTE Student metric counts part time students at 25 percent, which a total enrollment quote ignores.
- Discount comes second. A large academic discount on a headcount metric still produces an inflated bill, so fix the count first.
Which vendors drive higher education software pricing?
Five vendors carry most of a university's software bill: Oracle, Workday, Microsoft, Ellucian and Adobe. Student information, finance and human resources systems sit alongside campus wide productivity and creative software, and each group is priced on a different count of people.
In our work, universities lose less on unit price than on the count. The number the unit price is multiplied by usually includes people who never use the system.
The administrative core
Oracle PeopleSoft Campus Solutions and Workday Student run the student and finance side at many institutions. Ellucian Banner is the other major student system. Microsoft and Adobe supply the seats every student and employee sees.
- Oracle and Workday. Student records, finance and human resources. See the Workday education portfolio, and our note on how Workday counts Full Service Equivalent workers.
- Ellucian. Banner and its surrounding modules, which fall in the student information row of the table below.
- Microsoft. Identity, email and productivity, often under academic agreements through the Microsoft education program.
- Adobe and lab software. Creative and specialist tools, priced per named user, or per device in labs and classrooms.
How do FTE and student metrics inflate the bill?
University vendors price on full time equivalent staff, total student headcount or total employee count. Those counts include people who never touch the system, so the metric overstates real use before any discount is applied.
The headcount mismatch
A finance system priced on total employees charges for faculty who never open it. A student system tied to enrollment charges for students whose only contact with campus IT is the learning platform. The table shows where each type of system tends to overcount.
| System | Typical metric | Inflation risk |
|---|---|---|
| Student information | Total enrollment | Counts students who never log in |
| Finance and HR | Total employees or FTE | Counts staff who never use it |
| Productivity | Per named user | Alumni and inactive accounts retained |
| Lab and creative | Per device | Idle lab machines licensed full time |
What the contract definitions actually say
The vendors' own definitions are often narrower than the number in the quote. Read them before you accept a count, because each one gives you something to hold the vendor to.
- Oracle Enterprise FTE Student. Every full time student counts as one, and every part time student counts as 25 percent of one. Full time and part time follow your own classification policy, and the licensed quantity must at least equal the count on the order date.
- Oracle Enterprise FTE Student, reporting. You must report the count to Oracle every year, 90 days before the order anniversary. If the count rises you buy more; if it falls you get no refund or credit.
- Microsoft Enrollment for Education Solutions (EES). EES now counts Education Qualified Users: employees and contractors, excluding students, who use the covered products. The organization wide count must be at least 1,000, and for products such as Microsoft 365 A3 and A5, students can be licensed at a ratio of 1:40 per qualified user at no extra cost.
- Oracle Java SE Universal Subscription. The Employee metric counts all full time, part time and temporary employees, plus the staff of contractors and outsourcers who support your operations. Student workers on payroll count. Oracle lists pricing from $15 per employee per month.
- Workday HCM and Financials. Priced on Full Service Equivalent workers, where part time and contingent staff are weighted into one figure that you and Workday agree in the order.
What does an inflated count cost a university?
A worked example shows the size of the gap. Say a university has 18,000 full time and 6,000 part time students, 5,200 Microsoft staff seats in its tenant, and 900 Adobe Creative Cloud named users. All figures are hypothetical.
| System | Count in the proposal | Correct count | Proposal above correct count |
|---|---|---|---|
| Oracle student system, Enterprise FTE Student | 24,000 (total enrollment) | 18,000 + (6,000 x 25%) = 19,500 | 4,500 FTE, 23 percent |
| Microsoft 365 A3, staff seats | 5,200 accounts | 5,200 less 700 student workers who can move to student licenses and 500 departed staff = 4,000 | 1,200 seats, 30 percent |
| Adobe Creative Cloud, named users | 900 named users | 900 less 180 with no sign in for a full term = 720 | 180 users, 25 percent |
Now compare that with the discount. If Oracle offers a further 10 percent off the price on the 24,000 count, you pay for the equivalent of 21,600 FTE at the old rate. The correct 19,500 count with no extra discount is still about 10 percent cheaper, every year of the term.
What licensing traps hit universities specifically?
Four traps recur on campus: research use mixed into administrative agreements, alumni and inactive accounts, lab device counts, and an academic discount that hides a steep renewal uplift. The first quote rarely separates them out.
Research and grant funded use
Research computing often runs on different terms from administrative use. Grant funded projects can carry their own entitlements, bought with grant money for a named project, that the central agreement then counts again. Ask research offices for their license purchases before the renewal starts.
Alumni and inactive accounts
Universities keep accounts long after a student leaves. Email and productivity seats for alumni inflate the named user count unless they are reclaimed each term, and graduates who still hold a paid license are the easiest saving on the list.
Microsoft adds a storage angle. Education tenants get a base of 100 TB of pooled storage from the first renewal on or after August 1, 2024. Moving graduates to the free Exchange Online for Alumni license does not delete their OneDrive, so those files count against the pool until an administrator runs Remove-SPOSite and Remove-SPODeletedSite.
Lab device counts
Labs are licensed per device, and machines that sit idle for a term still carry a full license. Adobe offers shared device licensing for labs, libraries and classrooms, so a lab machine can run Creative Cloud without a named user license for everyone who sits at it. Match the device list to machines that are actually booked.
The academic discount and the renewal uplift
Academic pricing starts low and can climb steeply at the first renewal. A generous first term discount on an inflated count sets the base that every later uplift is calculated from, so the count you accept in year one follows you for the life of the relationship.
What have we seen in recent university renewals?
Across roughly 15 to 20 higher education engagements we advised in 2024 and 2025, the headline academic discount hid a metric that counted total enrollment or total staff. Three patterns recurred.
- Licensed counts ran high. On headcount based metrics, licensed counts ran 15 to 35 percent above active users.
- Alumni seats piled up. Alumni and inactive accounts made up 10 to 20 percent of named productivity seats.
- Research was counted twice. Research and grant funded use was double counted against the central agreement in most of the institutions we reviewed.
We also found four different metric models running on a single campus: enrollment, employee or FTE, named user and per device. Each one needs its own count, its own owner and its own cleanup date, which is why a single central renewal review tends to miss at least one of them.
Why the academic discount does not settle the question
The usual advice is that the academic discount makes higher education software a solved problem, so universities should simply renew. We disagree.
In the engagements above, the gap between the licensed count and active users was larger than the extra discount on offer. Negotiate on active users and reclaim alumni accounts every term, then judge the discount on the corrected count.
A large academic discount on a total headcount metric is still an inflated bill. Get the count right first, then argue about the percentage.
How do you check your own license position on campus?
Start from the systems that record who actually signs in and who is actually enrolled or employed. Each vendor count has a matching internal source, and most universities already run the reports for other purposes.
- Microsoft. Sign in activity in Microsoft Entra ID and the usage reports in the Microsoft 365 admin center show which licensed accounts have gone quiet. Cross check them against HR status to find departed staff and student workers on staff SKUs.
- Student systems. Your institutional research office holds the official census count, split into full time and part time. The same enrollment is published through IPEDS, so assume the account team has already looked it up.
- Workday. A headcount report by worker type and time type gives the inputs to the FSE figure. Compare it with the number in your order form.
- Adobe. The Admin Console lists license assignments by user and by shared device package. For last sign in dates, your single sign on logs are usually the more complete source.
- Oracle Java. Payroll headcount plus contractor and outsourcer staff, including student employees. Check it before Oracle asks, because the Employee metric does not care how many people run Java.
- Research. Grant accounting and the research office's purchase records show which projects bought their own licenses.
How do universities cut software spend at renewal?
Reclaim inactive and alumni accounts, align the metric to real users, separate research from administrative entitlements, and time renewals against the academic year. Start with the reclaim, because it pays back before the vendor is even involved.
Reclaim every term
Run a reclaim against graduates and inactive staff before each renewal. The academic calendar gives a natural cleanup point after every term, and most universities skip it.
- Reclaim. Drop alumni and inactive accounts, a saving that needs no vendor agreement.
- Metric alignment. Negotiate counts based on active users, and push back on total enrollment or total headcount.
- Sector benchmarks. Compare terms against peer institutions using Educause sector data, including its Core Data Service on IT spending, before signing. Consortium agreements such as Internet2 NET+ in the US or Jisc in the UK give another reference point.
Set the count date after the cleanup
Set the count date so it falls just after a reclaim. For enrollment metrics, that means a count drawn from the fall census once withdrawals are processed. Oracle's Enterprise FTE Student quantity is fixed at the order's effective date and reported again ahead of each anniversary, so pick an effective date that puts both after your cleanup.
Fiscal years matter too. Microsoft closes its fiscal year on June 30, as do many universities, which puts both sides under budget pressure in the same weeks. See our note on the Microsoft June 30 year end before you plan the calendar.
Contract wording to ask for
- A written metric definition. Name the count source, such as the fall census file or an HR extract, so the number cannot be swapped for a public total later.
- A right to reduce. Oracle's standard definitions give no credit when the count falls, so negotiate a reset of the count at each renewal at minimum.
- A cap on renewal uplift. Put the maximum increase for the next term in the order, because the academic discount often disappears at the first renewal.
- A research carve out. Exclude grant funded users already licensed elsewhere, or list them on a separate schedule with its own term.
- Student employee treatment. State in writing whether student workers count. Microsoft's Education Qualified User definition excludes students, while Oracle's Java Employee metric covers anyone on payroll.
What the account team will say, and how to answer
- "Every institution is priced on enrollment." Ask for the metric definition in your contract. If it is Oracle's Enterprise FTE Student, part time students are weighted down and your own policy decides who is part time.
- "This is our best academic rate." Ask for the unit price and the count as separate lines. Then negotiate the count on its own.
- "Research groups need to sit on the central agreement." Ask which grant funded licenses are already paid for, and require that they come off the central count.
How does the approach change for a small college or a large university system?
A small college has fewer contracts but less room to move. A multi campus research university has more money at stake in each count, and more places for double counting to hide.
Small and teaching focused institutions
An institution with fewer than 1,000 qualified staff may not meet the Microsoft EES minimum and often buys academic licenses through a Cloud Solution Provider instead. Here the reclaim and the part time weighting do most of the work, and consortium pricing is often the best available benchmark.
Research universities and multi campus systems
Large systems run central agreements next to campus, hospital and research purchases. The biggest savings usually come from mapping who bought what, removing grant funded users from central counts, and making sure each campus reports against the same definition.
When should a university start preparing for a software renewal?
Start 12 months before the renewal date. Counts, reclaims and research carve outs take a full term to land, and the vendor's proposal usually arrives too late to leave time for that work.
| Before renewal | What to do |
|---|---|
| 12 months | Map each system to its metric, its count date and the office that owns the count. Request sign in and census extracts. |
| 6 months | Run the term end reclaim of graduates and inactive staff. Collect research and grant funded license purchases. |
| 3 months | Send the vendor your corrected count with its source. Benchmark the proposal against peer and consortium terms. |
| 1 month | Confirm the metric definition, reduction rights and uplift cap in the order form before signature. |
What to do next
- Inventory. List every university system and the metric each one licenses on.
- Measure. Pull active users against entitlement for student, finance and productivity systems.
- Reclaim. Remove alumni and inactive accounts ahead of the renewal, including their stored files.
- Separate. Split research and grant funded entitlements from the central agreement.
- Negotiate the count. Send each vendor your corrected count with its source before you discuss discount.
- Benchmark and sign. Check the offer against Educause peer data, then confirm the definition, reduction rights and uplift cap in the order.
Frequently asked questions
Which vendors drive higher education software spend?
Oracle PeopleSoft, Workday, Ellucian, Microsoft and Adobe carry most of it, split across student information, finance and HR, productivity and creative software. Specialist lab, research and learning platform tools add a long tail, but those five contracts are usually where a renewal review pays back first.
Why do university software metrics inflate the bill?
Vendors price on total enrollment, total employees or full time equivalent counts that include people who never use the system. A student system sees every enrolled student, and a finance system sees every faculty member on payroll, even though only a small office works in it daily.
Does the academic discount mean a fair deal?
Not on its own. The discount is a percentage applied to a count, and if the count is total headcount the bill stays inflated. Ask for the unit price and the count as separate lines, correct the count first, and only then compare the discount with peer institutions.
How do alumni accounts affect licensing?
Alumni accounts inflate the named user count until they are reclaimed. With Microsoft, moving graduates to the free Exchange Online for Alumni license stops the seat cost, but their OneDrive files stay and keep counting against the tenant's pooled storage until an administrator deletes them.
Should research use be licensed separately?
Often yes. Research and grant funded computing can run on different terms, and a grant may already have paid for licenses that the central agreement counts again. List those purchases with the research office and ask the vendor to exclude them or place them on a separate schedule.
How do I cut a university software bill?
Start with the reclaim of alumni and inactive accounts, because it needs no vendor agreement. Then correct each metric to the contract definition, remove research users licensed elsewhere, and negotiate reduction rights and an uplift cap into the order before you sign.
When should universities time software renewals?
Set the renewal and the count date so they fall just after a term end reclaim and the fall census. For Oracle enrollment metrics the annual count report is due 90 days before the anniversary, so plan backward from that date.
How do I benchmark university software terms?
Compare metrics, unit prices and discounts with peer institutions using Educause resources such as its Core Data Service, and with consortium agreements where they exist. Bring the comparison to the renewal meeting in writing, so the account team has to answer it line by line.