Higher Education ERP Failure Rates and Causes: A Research Review

The first honest thing to say: Bettera does not do ERP replacement work. We do orchestration on ServiceNow. So why publish a research review on ERP failure rates in higher education? Because the research is the single most important context for the platform decision a CIO actually faces. If the published evidence shows that ERP replacement misses budget, schedule, or scope at the rates the data suggests, the right question is no longer "which ERP do we choose." It is "is replacement the right pattern for this category of system to begin with." That question has an answer. It is in the research.

What the published research actually says

A compilation of sources where the institutions involved have consented to be referenced through published academic work, sponsored EDUCAUSE studies, or named industry research.

The most current higher-education-specific data comes from a 2023 EDUCAUSE study More than "Going Live": Achieving Institutional Transformation through ERP Implementation. Key findings: ERP systems in higher education are reported as "misaligned, overly customized, and costly," and nearly half of surveyed institutions had either recently upgraded or planned to do so within five years. The study describes ERP implementations as "often more about the people than the technology," with success depending on change management and stakeholder engagement rather than vendor selection.

A Walden University dissertation grounded in Gartner cost containment research found that ERP implementations in higher education result in budget overruns 50 percent of the time.

The Panorama Consulting 2025 ERP Report puts the average ERP implementation cost at approximately $450,000 for mid-sized organizations across all sectors. The top causes of budget overruns: underestimated project staffing (38 percent), scope expansion (35 percent), and technical or data issues (34 percent). More than a quarter of organizations exceeded their budgets. The report also notes that only 7 percent of organizations use their ERP systems as-is, with 93 percent requiring customization.

The Workday published implementation methodology puts a combined HCM and Financials implementation at 12 to 24 months. That is the vendor's own number, before institutional change management cost layers on top.

The 2023 EDUCAUSE Technology Solutions Market Dashboard, drawing on data from 871 institutions since 2019, shows that finance ERP market share is fragmented: Ellucian 44 percent, Oracle 21 percent, Workday Financial Management 9 percent, Microsoft Dynamics 4 percent, with the remainder spread across more than 20 additional products. Doctoral institutions use a wider variety of finance ERP systems than associates or bachelors institutions, which suggests that consolidation through replacement is hardest at exactly the institutions where it is most often promised.

The pattern across sources is consistent.

Cost overruns are common.

Schedule overruns are common.

The institutions that succeed cite change management and stakeholder engagement, not technology, as the determining factor.

The five structural conditions that make higher ed harder

The research describes the failure rate. It does not always explain why higher education specifically produces it. Five structural conditions account for the pattern.

Faculty governance. Every major system change is a senate vote, not a CTO decision. Decision rights are distributed by design, and the platform that goes in this year has to survive the leadership turnover, the strategic plan revision, and the curriculum reform that come next year.

Federated decision rights. The registrar, the deans, the bursar, advancement, athletics, the medical school, the library, and the student affairs office each have a stake. Each can slow the project. Most do, somewhere between scope definition and go-live.

Decades of customization. Banner installations are often 20 years old. PeopleSoft instances inherit a generation of homegrown forms, custom reports, and workflows that the people who built them have retired. The institutional knowledge buried in those customizations is rarely fully recoverable, and reverse-engineering it consumes a meaningful share of every replacement project's budget.

Cycle-time constraints. Registrar deadlines do not move. Payroll runs do not move. Federal compliance reporting does not move. Higher education does not have the option to take a quarter offline for migration. Every replacement must produce parity with the existing system before each deadline cycle, which compresses timelines and inflates budgets in ways the corporate ERP literature does not account for.

Identity attachment. Current systems are tied to the offices and people that run them. Replacing the SIS is not just a technology project. It is a political project that asks the registrar, who has run the SIS for fifteen years, to trust a new platform with her professional reputation. The 2023 EDUCAUSE study made this the central finding for a reason.

These five conditions compound. Each makes the next worse. They are why corporate ERP replacement timelines do not transfer to higher education, and why the Panorama cross-industry numbers understate the higher ed reality.

The recurring shape of higher ed ERP project outcomes

The pattern across two decades of higher ed ERP projects is not a technology failure. It is a sequence.

In year one, scope is set with optimism. The vendor reference implementations look promising. Leadership signs the contract.

In year two, scope expands as offices realize the new system does not natively do what their custom Banner or PeopleSoft code does. Customization debt accumulates. The transformation narrative starts to fray.

In year three, governance fatigue sets in. Faculty senate has voted on the project a dozen times. The CIO is in front of the board explaining why the timeline has slipped. Internal staff who started the project leave, often for vendors or for institutions that are not in year three of an ERP replacement.

In year four, the project pivots from "transformation" to "replacement of what we had." The vision narrows. The leadership team that signed the contract has, in many cases, changed.

In year five, scaled-back go-live. The institution has the new system. It does most of what the old system did. The board is told the project is a success.

The CIO who recognizes this sequence has the question: is this the program we are signing up for, and is there an alternative pattern that does not produce it.

What CIOs are hearing, and what is actually broken

Replacement conversations rarely start with the technology. They start with a meeting where a board member, a peer CIO, a vendor account executive, or a frustrated VP says some version of "this feels old."

What CIOs are actually hearing on a typical week:

• "Peer Institution X just moved to Workday."

• "The vendor is sunsetting our version."

• "We cannot hire engineers who know this system anymore."

• "Faculty say it takes 47 clicks to submit a leave request."

• "Why does it look like it is from 2003?"

• "Our reporting is broken."

• "Mobile is unusable."

• "We have 14 different login screens."

Underneath every one of these is a different problem. Some genuinely point to the limits of the system of record. Most do not.

The question worth asking before any replacement decision: is the institution outgrowing the system, or is the experience layered on top of it broken?

Outgrowing the system means the data model genuinely cannot support what the institution needs. The vendor has truly sunset. Performance ceilings have been hit. Compliance requires a record-level change. These cases exist. They are rarer than they sound.

The experience being broken means that the system of record is doing its job, which is holding data correctly, but everything around it is making people miserable. The portal is fragmented. The forms are ugly. The workflows route through email. There is no self-service for the things people want to self-serve. Reports are static. Mobile is missing. Integrations are brittle.

The second pattern is the far more common diagnosis. And it is the one ERP replacement does not solve. A new ERP gives the institution a different system of record, with a different UI, that requires a different set of integrations, and produces a different set of complaints starting in year two.

A different pattern fixes the experience without replacing the underlying record.

ServiceNow Employee Workflows orchestrate the cross-functional work. Employee Center provides the unified front door. The existing ERP, whether Banner, Workday, PeopleSoft, or Oracle, keeps doing what it is good at, which is being the source of truth for employee, financial, and student data.

The faculty member who used to navigate seven portals sees one. The HR analyst who used to copy data between systems sees one queue. The CIO who used to defend a five-year replacement program defends a one-quarter expansion of a platform that already works.

The outgrown-versus-experience diagnosis is the single most useful question a CIO can ask before signing a replacement contract.

Three questions to answer before committing to replacement

The first is the outgrown-versus-experience question above. The other two diagnostics:

What specific institutional outcome is the replacement promised to deliver that no other pattern can deliver?

"Modernization" is not an outcome. "Reducing the time it takes a faculty member to onboard from three weeks to three days" is an outcome. If the answer is the former, the project is a wishlist, not a program. The Panorama data shows that scope expansion is the second-largest driver of budget overruns precisely because wishlist programs cannot resist scope creep.

What is the political and operational reality of the migration cycle?

Five years of disruption is the median, not the worst case. The structural conditions documented above are not avoidable through better project management. They are properties of the institution.

Have we tested the non-replacement pattern?

The orchestration alternative does not require replacing the system of record. It requires connecting the existing record to a workflow layer. The test is small, fast, and reversible. The replacement decision is none of those.

A CIO who can answer all three with specifics has done the work the published research calls for. A CIO who cannot has not yet done the diagnostic.

The pattern that sidesteps the failure mode

The research consistently shows that systems-of-record replacement is the high-failure pattern in higher education. The structural conditions are not avoidable. They are properties of the institution.

The orchestration pattern does not fight those conditions. It works with them.

Banner stays. Workday stays. Salesforce stays. PeopleSoft stays. The institution adds a workflow layer that orchestrates across the systems already in place, owns the user experience, and routes the work to wherever the data actually lives. ServiceNow is the platform purpose-built for this layer in higher education. Bettera is the consulting partner that runs the orchestration sequence as a quarterly proof rather than a five-year program.

This is the framing of our Pillar 1 piece on orchestration versus consolidation. It is also why we publish a research review on ERP replacement despite not doing ERP work. The research does not argue against ERP. It argues for matching the platform decision to the actual problem. When the problem is the experience, orchestration is the smaller, faster, and more durable answer.

The full case is in our Beyond the Help Desk white paper.

Frequently asked questions

Is the research saying ERP replacement never works in higher ed?

No. The research shows it fails at high rates in this sector, and that the institutions that succeed cite change management and stakeholder engagement as the determining factor, not vendor selection. Replacement can work. It is rarely the lowest-cost path to the underlying outcome.

What if our existing ERP truly cannot be sustained?

That is the outgrown case. Run the diagnostic in the section above. If the answer holds, replacement may be the right call. The research review is meant to disqualify replacement projects that do not need to be replacement projects, not the cases where replacement is genuinely required.

What does Bettera actually do if not ERP work?

Orchestration on ServiceNow. We help R1 institutions connect the systems of record they already have, through ServiceNow workflows, Employee Center, and case management.

The full answer is at bettera.co.

Where this leaves the CIO

Most replacement conversations would benefit from the diagnostic in this post. Most institutions would discover, partway through, that what they need is the experience pattern fix, not the system of record pattern fix. The research has been saying this for two decades. The platform pattern that responds to it has only existed for a few years.

For the orchestration framing this post sits on top of, our Pillar 1 piece is the foundation.

The full case is in our Beyond the Help Desk white paper. If you want a working session on the diagnostic for your institution, that is what we do.