Connect your Legacy ERP to modern systems — without replacing it.

The default response to an aging ERP is to replace it. Boards approve the budget, system integrators pitch the roadmap, and two years later the organization is over budget, behind schedule, and wondering what went wrong. The data on ERP replacement outcomes should give every IT director pause before signing that contract.

Independent research consistently places ERP failure or significant overrun rates at 50–75%, depending on how failure is defined. Panorama Consulting and Gartner survey findings across multiple years show that a majority of projects exceed their original budget, timeline, or expected benefits. In discrete manufacturing, where bill-of-materials complexity and job costing amplify every integration risk, the failure rate climbs even higher. Real-world cases underscore the point: Birmingham City Council's Oracle Fusion implementation began with an initial £19 million budget that grew to approximately £90 million by 2024, with total projected costs through 2026 reaching an estimated £216 million. These are not anomalies — they are the statistical baseline.

The alternative that experienced CIOs are choosing is an API middleware layer: a purpose-built integration architecture that connects the existing ERP to modern CRM, WMS, e-commerce, BI, and AI tooling without touching the system of record. The economics are straightforwardly different. An iPaaS platform from MuleSoft, Boomi, Workato, or Azure Integration Services typically runs $80,000 to $350,000 annually in licensing for a mid-market deployment — a predictable operating expense, not a multi-year capital commitment. A well-scoped integration layer project runs $300,000 to $1.2 million in services depending on endpoint count and data complexity, which is an order of magnitude cheaper than a full replatform in most scenarios.

The market is already moving in this direction. Kyndryl's 2025 State of Mainframe Modernization Survey of 500 senior IT leaders found that 80% of organizations changed their modernization strategy over the previous year, with 50% moving toward a hybrid approach rather than full replacement. Only one respondent out of the 500 surveyed planned to move off their core platform entirely. The survey also found that modernization project costs dropped from $9.1 million in 2024 to $7.2 million in 2025, driven largely by more targeted, integration-first scoping.

The operational upside of an API layer is also concrete. In one documented case in the financial sector, a firm that built an API and data layer first — before adding AI capabilities — achieved a 35% improvement in return on marketing investment and a 22% reduction in cost per acquisition. A global logistics firm that modernized its legacy ERP using an API-first architecture reduced data processing latency by 55% with asynchronous task execution and enabled the rapid rollout of new modules without full system redeployments. An Australian manufacturer used the same approach to connect its ERP to CRM, accounting software, shipping providers, and Power BI without rewriting the ERP core — eliminating manual data re-entry across four separate systems.

None of this means replacement is never right. Tool sprawl, end-of-support deadlines, or a genuinely broken process model can make replatforming the correct call. But the burden of proof should sit with replacement, not with integration. The question for an IT director or operations leader is not whether the legacy ERP is perfect. It is whether a full replacement is the right lever — or whether an API middleware layer can deliver the connectivity, agility, and AI-readiness the business needs at a fraction of the cost and risk. For most enterprises right now, the data strongly favors the layer.

If you are planning a full ERP replacement, the odds are not in your favor — and the budget your vendor quoted almost certainly is not either. Research consistently shows that between 55% and 75% of ERP projects fail to meet their stated objectives, according to Gartner. Panorama Consulting's 2025 research puts the overall failure rate even higher, at 68%. Only 32% of implementations are completed on time, and 49% of organizations report measurable operational disruption after go-live. These are not outlier outcomes. They are the norm.

The financial damage starts well before a single user logs in. Average implementation costs run 189% over budget across industries, rising to 215% in manufacturing, according to Panorama Consulting data. The Standish Group's CHAOS Report found that challenged projects — those that stumbled but were not abandoned — cost 189% of their original estimates. On-time delivery is equally elusive: 55% of ERP projects exceed their planned timeline by more than six months, and data migration issues alone cause an average 20% schedule delay on 70% of rollouts.

What makes full replacements particularly dangerous is the gap between the quoted price and the actual price. Software, hardware, and basic implementation services typically represent only 30 to 40 percent of what a full ERP program ultimately costs. The rest comes from categories that rarely appear on a vendor proposal: internal staff productivity loss (subject-matter experts routinely give 20–40% of their working time to the project), data cleansing costs ranging from $15,000 to $75,000 depending on data volume, custom report development adding $30,000–$80,000 for a programme with 20–30 reports, and third-party add-on licensing running $5,000–$30,000 per add-on annually. A project requiring four to six add-ons can easily add $50,000–$120,000 to annual cost of ownership alone. Scope creep, which affects 62% of ERP projects, then compounds each of these line items.

The case record for large full replacements is instructive and should be required reading before any board approves a rip-and-replace. Hershey's $112 million ERP project compressed a planned four-year rollout into roughly 30 months to meet a Y2K deadline, then went live right before Halloween — its highest-volume season. Order-to-cash broke down entirely, leaving more than $100 million in confirmed candy orders undeliverable despite inventory sitting in the warehouse, and quarterly profits fell approximately 19%. Nike committed approximately $400 million to modernize its supply chain in 2000, deploying demand-planning software before its core SAP platform was ready. The systems operated with incompatible data formats, generating flawed forecasts that ordered too many low-demand products and too few high-demand ones — resulting in $100 million in lost sales and a 20% drop in stock value. Revlon's 2018 SAP S/4HANA cutover after its Elizabeth Arden acquisition disrupted its North Carolina manufacturing facility so severely that the company could not fulfill $64 million in orders, sending its stock down 6.9% in a single day.

The root cause pattern across failures is consistent. Internal organizational factors — poor requirements definition, insufficient change management, and executive disengagement — drive 60–70% of ERP project collapses, according to Panorama Consulting. Employee resistance to change is the leading implementation inhibitor, cited in 82% of troubled projects. A further 35% of failures involve inexperienced project teams, typically from the implementation partner rather than the client organization. Poor project governance alone increases failure rates by 50%.

For IT directors and operations leaders, the strategic implication is direct: the risk in a full ERP replacement is not primarily technical. It is organizational, financial, and temporal. Before committing to a rip-and-replace, pressure-test the vendor quote against the full lifecycle cost, stress-test the go-live timeline against your peak operational periods, and treat change management not as a line item to trim but as the project's central deliverable. The companies that recovered from these failures — including Hershey, which completed a subsequent mySAP upgrade 20% under budget — did so only after redesigning the entire process, appointing new leadership, and abandoning the big-bang approach entirely.