Maintaining a ten-year-old on-premise ERP costs more every year for less and less value. According to Gartner, by 2025 companies will spend 40% of their IT budget on technical debt. Organizations maintaining legacy systems spend an average of 60 to 80% of their IT budget keeping the past alive — at the expense of any transformation initiative.
Yet launching an ERP replatforming project without a solid business case means facing two failures: the project won’t get funded, or it will be launched with a miscalibrated budget and die at the first overrun.
This article gives CIOs and CFOs a structured methodology to build a decision document that withstands scrutiny from senior leadership.
Legacy ERP: What Does That Actually Mean in 2026?
Defining a Legacy ERP: Technical and Commercial Criteria
An ERP becomes “legacy” when it combines at least three of the following characteristics:
- Outdated technical architecture: proprietary on-premise server, unsupported database, Win32 or Java 6 client-server interface.
- Version outside the vendor’s mainstream support cycle: no new security patches, no regulatory update included.
- No native REST API: integrations rely on flat-file exchange or direct database access.
- Dependency on scarce skills: ABAP, Progress 4GL, COBOL, RPG/AS400 development — profiles whose market shrinks every year.
- Growing functional gap: business requests pile up in the backlog with no delivery timeline.
The most objective criterion remains vendor support. SAP ECC 6.0’s mainstream maintenance end date (Enhancement Packages 6 to 8) is set for 31 December 2027 — and SAP has repeatedly confirmed this date will not be extended (Rimini Street). Yet according to Gartner and CIO research, only 39% of SAP ECC customers had licensed S/4HANA by end of 2024 (SAVIC Technologies): more than 60% of the installed base has not yet started its migration.
The Grey Zone: The Maintained but No-Longer-Developed ERP
The most common situation in mid-market companies is not an ERP that is clearly “end of life” — it is an ERP that runs, that is maintained by a long-standing service provider, but that receives no new features. The vendor has stopped active development (or was acquired), the provider applies security patches, and business teams live with functional gaps that they paper over with Excel.
This scenario — which can be called “zombification of the IT estate” — is harder to objectify than a vendor end-of-support date, because there is no hard deadline. The focus must shift to operating costs to build the business case.
ERP Platforms Most Commonly Affected
- Industry-specific on-premise ERP from the 2000s: Sage 100, older Epicor versions, Infor legacy platforms, Syspro older releases.
- AS/400 and RPG applications: still present in industrial mid-market companies.
- SAP ECC 6.0: the 2027 deadline makes it the most widely discussed case.
- Microsoft Dynamics NAV: migrated since to Business Central, but frozen on-premise NAV 2009–2015 installations still exist.
- Internally developed business software on COBOL, Progress, or Clipper from the 1990s–2000s.
The 7 Warning Signals That Demand Action
1. Maintenance Cost (AMS) Exceeds the Cost of a New ERP Over 5 Years
This is the clearest financial signal. Compile your AMS invoices for the past three years, include internal costs (system administrators, DBAs, key users mobilised for incidents), and project over five years using a conservative 5% annual growth assumption. Compare with a SaaS TCO over the same period. In many cases, the status quo cost already exceeds the migration cost.
2. Internal Skills Are Scarce and Ageing
If your operational stability depends on one or two people mastering a language that junior developers no longer learn (legacy ABAP, Progress 4GL, RPG), you face a critical “bus factor” risk. Losing a COBOL developer in 2026 can cost 6 to 18 months of recruitment and $200 to $350 per hour for a specialist freelancer.
3. The Vendor Has Announced End of Support
This is the hardest trigger to dismiss in a board meeting. SAP ECC 2027 is the canonical example. But there are also unmigrated Dynamics NAV branches, Oracle E-Business Suite configurations losing extended support, and regional ERP vendors acquired by larger players who progressively wind down development of the legacy product line.
4. Integrating New Regulations Requires Costly Custom Development
E-invoicing mandates rolling out across the EU, CSRD for companies with more than 250 employees, real-time VAT in Spain, e-reporting in Germany — every new regulatory obligation is a custom development project on a system that was never designed for it. If your legacy ERP generates a quote of $100,000 to $250,000 per regulatory requirement, the replatforming justification becomes arithmetic.
5. User Experience Impairs Productivity and Hinders Talent Retention
An ERP with a 2005-era Win32 interface creates friction for recruiting junior profiles and generates measurable data-entry error costs. According to a study cited by RecordPoint, 48% of workers surveyed waste more than 3 hours per day due to inefficient systems — a productivity loss estimated at £28,000 per employee per year in a UK context. Translating this to a mid-market company context produces numbers that are easy to argue before a CHRO or commercial director.
6. The ERP Cannot Interface with Modern Tools
The absence of a native REST API prevents natural integration with e-commerce platforms (Shopify, Magento), CRM tools (HubSpot, Salesforce), accounting automation, or analytics (Power BI, Tableau). Every integration is a custom development costing $25,000 to $100,000 per connector — and breaks with every version upgrade of the third-party system.
7. Technical Debt Blocks Digital Transformation Projects
An AI initiative, an RPA automation project, or a supplier portal consistently runs into the constraints of the core IT system. Technical debt becomes a quantifiable strategic barrier: how many projects have been deferred, scaled back, or cancelled because of legacy ERP constraints? This inventory, built with business unit heads, produces a list of missed opportunities that can be valued.
Replacement, Upgrade, or Coexistence: Mapping Your Options
Before building the replatforming business case, you need to put all four real options on the table. A strong decision document does not present a single recommendation: it articulates all four scenarios with their respective costs, timelines, and risks.
Option 1 — Major Upgrade (Stay with the Current Vendor)
You remain in the vendor’s ecosystem but upgrade the version. This is the SAP ECC → S/4HANA scenario, or Dynamics NAV → Business Central. The advantage: continuity of processes, partial preservation of existing configuration, and an established commercial relationship. The downside: a major upgrade with more than 40% customisations typically ends up as a disguised reimplementation — at equivalent cost to replatforming, but with less process rethinking. See our analysis of major ERP upgrades for detail on the hidden costs.
Option 2 — Full Replatforming (Change Vendor)
You choose a new ERP platform and reimplement from scratch. This is the most structurally impactful and most expensive option in the short term, but it also delivers the greatest functional gains and the best 5-year total cost of ownership reduction. It forces you to redesign processes rather than reproduce them — which is often the opportunity to fix inefficiencies that the legacy system had calcified.
Option 3 — Temporary Coexistence (Bimodal IT)
For large enterprises or multi-entity mid-market companies, one approach is to deploy the new ERP on new entities or new lines of business while maintaining the legacy system for stable perimeters. This is the “two-speed IT” approach. It reduces the cutover risk but generates interfacing costs and IT governance complexity that accumulates over several years.
Option 4 — Composable Architecture (Module-by-Module Replacement)
You keep the legacy ERP core and progressively replace peripheral modules with specialised SaaS solutions: a cloud WMS, a modern CRM, an e-procurement tool, a business intelligence platform. This “best-of-breed” approach minimises cutover risk but multiplies the number of systems to integrate and maintain.
Decision Matrix
| Option | Initial Cost | Operational Risk | Functional Gain | Recommended When… |
|---|---|---|---|---|
| Major upgrade | Medium | Medium | Limited | < 20% customisations, active vendor |
| Full replatforming | High | High (during project) | Maximum | > 40% customisations, inactive vendor |
| Coexistence | Low short-term | Low | Partial | Multi-entity, different paces |
| Composable | Medium | Low | Partial | Solid ERP core, peripheral gaps |
Building the Business Case: A 5-Step Methodology
Step 1 — Quantify the True Cost of the Status Quo
This is the most overlooked and most decisive step. The apparent cost of the status quo (annual licences + AMS) is systematically underestimated. To objectify it, build a 5-year table that integrates:
Measurable direct costs:
- Maintenance and AMS (contract + out-of-contract interventions)
- Licences and updates
- Server infrastructure (hardware, hosting, backups)
- Custom development costs for new regulatory obligations
Indirect costs to estimate:
- Internal hours mobilised on incidents (number of tickets × average resolution time × internal hourly rate)
- Cost of Excel workarounds: inventory the parallel spreadsheets and estimate double-entry time, reconciliation errors, delayed month-end close
- Missed opportunities: deferred projects, harder recruitment, commercial limitations linked to the absence of e-commerce integration
Representative example (fictional): A $200M-revenue industrial mid-market company was paying $420,000 per year in AMS and infrastructure for a 2009 ERP. Adding hidden costs (12 FTEs partially mobilised on incident management, 3 blocked transformation projects) brought the true status quo cost to $680,000 per year. Over 5 years with 7% annual maintenance cost growth, the total reached $4 million.
Step 2 — Estimate the Full Cost of Migration
The TCO of an ERP migration contains several line items that integrator quotes systematically understate. For a 100–500-employee mid-market company, here are the reference ranges:
| Line Item | Range |
|---|---|
| Licences and subscriptions (first 3 years) | $170,000 – $700,000 |
| Integration and configuration (integrator) | $230,000 – $1,000,000 |
| Data migration and cleansing | $35,000 – $170,000 |
| Training and change management | $45,000 – $135,000 |
| Internal resources mobilised (key users) | $55,000 – $230,000 |
| Dual-run period (old + new platform, 3–6 months) | $25,000 – $90,000 |
| Peripheral reintegrations | $35,000 – $230,000 |
| Indicative total — 100–500-employee company | $595,000 – $2,555,000 |
These ranges are wide because the cost depends heavily on the volume of existing customisations, the quality of master data, and the number of peripheral systems to reintegrate. Our 5-year ERP TCO guide breaks these items down for the major vendors.
Step 3 — Project Benefits Over 5 Years
The benefits of a replatforming fall into three categories:
Directly measurable benefits:
- AMS cost reduction (typically 60–70% reduction by moving to cloud SaaS, per Forrester)
- Elimination of on-premise infrastructure costs
- Elimination of regulatory development costs (included in the SaaS subscription)
Productivity gains:
- Reduction in monthly close time (typical gains: 2 to 5 days on a 10–15-day close cycle)
- Automation of inter-module reconciliations
- Elimination of double-entry and parallel spreadsheets
Strategic gains:
- Unblocking transformation projects (e-commerce, analytics, automation)
- Improved recruitment capacity (modern interface, native hybrid working)
- Reduced regulatory compliance risk (updates included)
Quantify each benefit with a low and high range, and document the assumption. A productivity gain of 15 minutes per day for 50 users can be calculated: 50 × 0.25 hours × 220 days × $55/hour = $151,250 per year.
Step 4 — Calculate the Break-Even and Payback Period
The payback period is the central metric for a board: from what point does the investment pay for itself?
Simplified formula:
Payback = Total migration cost / (Annual savings + Annual gains)Using the representative example above: migration cost $1,000,000, annual savings (AMS + infrastructure) $385,000, estimated productivity gains $130,000 → payback in 1.9 years. Beyond that, the new ERP generates a net positive return.
Always present three scenarios: conservative, central, and optimistic. A board that sees the pessimistic scenario still returning the investment within 3 years decides more easily than when faced with a single 2-year projection nobody believes.
Step 5 — Present the Risks of the Status Quo vs the Risks of Migration
Senior leadership does not buy a ROI — it arbitrates between two risk families. Your business case must present them side by side.
| Status Quo Risk | Migration Risk |
|---|---|
| Critical failure with no vendor support | Budget overrun (+20 to +50%) |
| Regulatory non-compliance (fines, operational shutdown) | Go-live delay (6 to 18 months is common) |
| Loss of legacy skills (bus factor) | Change resistance and temporary productivity loss |
| Blocked strategic projects | Partial disruption during cutover |
| Loss of competitive ground to modernised peers | Peripheral integration overruns |
The key is to quantify status quo risks as rigorously as migration risks. A regulatory non-compliance risk potentially represents x% of revenue in fines. A 48-hour operational outage on a production ERP represents a calculable lost margin figure.
The 5 Classic Mistakes in an ERP Business Case
1. Underestimating change management. Organisations typically budget 5–10% for training. Successful projects actually spend 15–20%. A project that cuts corners on training recovers those costs as post-go-live productivity loss.
2. Forgetting peripheral integrations. For a mid-market company, the number of active inbound and outbound ERP flows regularly exceeds 30 to 60 integration points. Inventorying and costing these interfaces before finalising the budget is non-negotiable.
3. Comparing incomparable TCOs. Comparing the cost of an annual SaaS licence with the full 10-year cost of ownership for an on-premise ERP is a frequent mistake in both directions: either SaaS looks too expensive over one year, or legacy looks cheap because the hidden costs were forgotten.
4. Failing to include the cost of technical debt accumulated during the freeze. Every year without migration is a year of additional debt. Deferred projects carry an opportunity cost, new regulatory requirements generate one-off custom builds, master data debt accumulates.
5. Presenting only the benefits without quantifying the risks. A business case that only shows gains is perceived as a sales document, not a decision document. Seasoned decision-makers look for conservative assumptions and associated risks — give them those.
Board Presentation: What Decision-Makers Want to See
The board document should not reproduce the 40 pages of your internal analysis. It boils down to a 10–12-slide presentation with a binary narrative: the cost of the status quo vs the cost of action.
The “3 Options, 1 Recommendation” Summary Slide
Present the three viable scenarios (upgrade, replatforming, coexistence) on one slide with three columns: total 5-year cost, project timeline, level of operational risk. State your recommendation clearly along with the two decisive criteria that justify it.
The Cost/Risk/Timeline Comparison Table
A 3×3 table is enough. Colour high-risk cells red, medium-risk cells orange, and favourable cells green. The goal is visual: decision-makers should understand within 10 seconds why the recommended option is preferable.
The Internal Resources Question
The board will inevitably ask: “Who is going to own this project?” Anticipate it with a simplified project organisation chart: executive sponsor (CEO or CFO), internal project manager (CIO or dedicated profile), business representatives per domain (finance, supply chain, commercial, HR). Quantify person-days over 18 months and estimate the load relative to each contributor’s full-time capacity.
Where to Start: The 3 Concrete First Actions
If you have finished reading this article with the conviction that replatforming is the right call, here are the three first actions to take within the next 30 days:
-
Run the internal audit. Bring together your CIO, CFO, and 2–3 key users for a 4-hour session using the audit grid described in our ERP evaluation guide before migration. You will leave with a 100-point score and a list of priority gaps.
-
Quantify the true cost of the status quo. Ask your financial controller to consolidate AMS invoices for the past three years, infrastructure costs, and an estimate of internal time. That is the first slide of your future business case.
-
Identify 3 projects blocked by the legacy system. In each business unit, ask: “What would you have done with a modern ERP that you cannot do today?” These projects represent the opportunity cost of the status quo — the most convincing lever for a board decision.
To go further, see our 5-year ERP TCO comparison for precise cost-of-ownership figures across the major platforms, and our legacy ERP decommissioning guide to start planning the retirement of your current system now.
