What ERP Data Migration Costs When You Consolidate Three Legacy Systems
Most data migration quotes cover the integrator's hours and leave out the 4,000-plus hours your own staff will spend. Here's a stage-by-stage cost model for merging three legacy ERPs.
Why the migration line in your implementation quote is usually half the real number
If you're consolidating acquired businesses onto one cloud ERP, the data migration line in the integrator's statement of work is probably the most underestimated number in your business case. Integrators don't lowball on purpose. The SOW prices what they will do: extract scripts, transformation logic, load templates, a set number of mock conversions. It rarely prices the work only your people can do, like deciding which of three customer records for the same account is correct, or signing off that opening inventory valuation ties to the ledger.
In multi-system consolidations, that internal effort can match or exceed the external fees. So when a CFO asks for the ERP data migration cost, the honest answer has two parts: cash paid to vendors and tools, and staff hours taken away from running the business. This piece costs both, stage by stage, using one hypothetical company so the numbers add up and can be checked.
All prices below are typical 2026 market ranges. They vary by contract, region and vendor, so use them to sanity-check quotes, not replace them.
The hypothetical: a 250-user distributor merging three acquired systems
This is a hypothetical example built to show how costs behave. It doesn't describe a real company.
Say a wholesale distributor of industrial parts has grown by acquisition and now runs three systems:
- System A: an on-premise mid-market ERP at the parent company, about 15 years of history, reasonably well maintained.
- System B: an older accounting-plus-inventory package at an acquired regional business, with heavy use of free-text fields and side spreadsheets.
- System C: a heavily customized legacy system at a second acquisition, with its own item numbering and a separate warehouse module.
The target is one cloud ERP for 250 named users across finance, purchasing, sales and three warehouses. Combined, the systems hold roughly 18,000 customer records (assume a fifth are duplicates across entities), 4,500 vendors, 65,000 item records with overlapping SKUs, open AR and AP, open purchase and sales orders, and inventory balances by location and lot.
The company decides to migrate master data, open transactions and two years of summarized GL balances. Everything older goes to an archive. Migrating less history is the biggest single lever on cost, and we'll come back to it.
For internal time, we use a blended, fully loaded $70 an hour. Adjust it to your market: perhaps $25 to $45 for finance and operations staff in India or Southeast Asia, and $80 to $120 in Western Europe, the Gulf states or North America. External consultant rates are assumed at $120 to $220 an hour onshore and $35 to $80 for offshore delivery centers.
Stage-by-stage costs, from profiling to archiving
| Stage | External spend (vendors, tools) | Internal hours | Internal cost at $70/hr |
|---|---|---|---|
| Profiling | $25,000–$40,000 | 200 | $14,000 |
| Cleansing | $45,000–$80,000 | 1,800 | $126,000 |
| Mapping and transformation | $90,000–$150,000 | 450 | $31,500 |
| Mock loads (four cycles) | $70,000–$120,000 | 900 | $63,000 |
| Reconciliation and cutover | $30,000–$50,000 | 550 | $38,500 |
| Legacy data archiving | $40,000–$110,000 | 200 | $14,000 |
| Total | $300,000–$550,000 | 4,100 | $287,000 |
Profiling
Profiling means running each source through a tool or scripts to count nulls, duplicates, invalid codes and orphaned records. External cost covers a few weeks of a data analyst plus a profiling tool. Internal time goes mostly to IT staff granting access and explaining what the custom tables in System C mean. Skimp here and every later stage costs more, because the surprises show up during mock loads.
Cleansing
This is where data cleansing cost lands, and most of it is internal. A consultant can flag the roughly 3,600 probable duplicate customers. Only your credit controllers and sales managers can decide which record survives, which payment terms apply and which ship-to addresses are still valid. Add vendor and item rationalization, unit-of-measure rework and harmonizing three item-numbering schemes, and 1,800 hours is our planning assumption for this scope. It could run higher if System B's free-text fields hide pricing agreements.
Mapping and transformation
Mapping is the core of the ERP data conversion: field-by-field rules from three sources into one target, including chart-of-accounts consolidation, tax code mapping by country, and new item and location structures. It's the largest external line because it's skilled build work. Internal hours go to finance and operations leads approving the mapping specs. Make sure they read them before signing.
Mock loads
Plan for at least three full mock conversions, and four with three sources. Each cycle needs business users to validate loaded data: spot-checking customer balances, running test orders, confirming inventory by bin. Those 900 hours usually fall to people who are also doing user acceptance testing and their day jobs.
Reconciliation and cutover
Finance has to prove that opening balances, open items and inventory valuation in the new ERP tie to the legacy ledgers. Auditors in most jurisdictions will want to see that evidence. Cutover weekends add overtime and often temporary warehouse staff.
Archiving
The history you don't migrate still has to be retrievable for audits, tax inquiries, warranty claims and disputes. The cost range is wide because the options differ so much, as covered below.
Where internal staff time hides, and how to put it in the budget
In the example, internal effort is roughly 34% to 49% of the total, depending on where external fees land. The all-in total is about $587,000 to $837,000. Budget only the SOW and you're short by about $287,000.
The hours matter more than the dollars. Divide 4,100 hours by roughly 1,700 productive hours a year and you need about 2.4 full-time people for a year. In practice the work lands on a handful of people: the AR lead, the purchasing manager, the inventory controller and the financial controller. They can't give up a large share of their week for months without something slipping, whether that's collections, supplier follow-up or month-end close.
So the question isn't whether to pay for internal time. It's how:
- Backfill the day job, not the project role. Bring in temporary staff or a finance outsourcing provider for routine work like cash application, invoice processing and cycle counts, so your experts can make the data decisions. Backfill commonly costs $40,000 to $120,000 per role-year, depending on country.
- Name data owners per object. One accountable person each for customers, vendors, items, the chart of accounts and inventory. Without that, cleansing decisions bounce around and the hours grow.
- Track internal hours weekly. If cleansing has burned 60% of its hours with only 30% of records resolved, you have a schedule problem, and you can see it months before go-live.
- Set a cleansing deadline before mock load two. Records not cleansed by then either migrate as-is or get archived. Open-ended cleansing is how projects run late.
Our position: for a consolidation of this size, put internal hours in the business case as a costed line with funded backfill. Companies that treat them as "absorbed" usually pay anyway, through a delayed go-live and integrator change orders.
Data migration tools and archiving choices that move the total
Tool choice shifts the split between external and internal effort more than it shifts the total. Three broad approaches are available internationally:
- Native ERP import tooling, such as SAP's Migration Cockpit, the Data Management Framework in Microsoft Dynamics 365, Oracle's file-based data import, or NetSuite's CSV import and SuiteCloud tools. Usually included in the subscription. Good for loading, weaker for cleansing and cross-source matching.
- General ETL and integration platforms, such as Informatica, Boomi, Qlik Talend or Azure Data Factory. Licenses often run $20,000 to $100,000 a year at this scale. Worth it if you'll reuse the platform for integrations after go-live.
- Specialist migration and data quality platforms, such as Syniti and similar products, with built-in matching, rules and reconciliation reporting. Often $50,000 to $200,000 per project including services. Worth considering when three or more sources make deduplication the bottleneck.
For this distributor, native tools plus a mid-priced data quality or matching tool is the sensible default. A specialist platform starts to pay off if profiling shows duplicate rates well above the assumed 20%, or if more acquisitions are coming.
Legacy data archiving options
You have three realistic choices for the history you leave behind:
- Keep the legacy systems running read-only. Cheap at first glance, but three sets of licenses, hosting and security patching can add up to $30,000 to $80,000 a year. It also leaves unsupported software on your network.
- Extract to a data warehouse or data lake with basic reporting. Often $40,000 to $90,000 to build, plus modest ongoing storage costs. The best fit for most mid-market companies.
- A dedicated archiving product with retention policies and legal hold. Higher cost, but easier to defend in an audit in tightly regulated sectors.
How long you keep the archive depends on where you operate. Retention periods for accounting and tax records commonly run from about five to ten years, and they differ by record type: the UK, India and Germany, for example, all set different periods for different books and tax records. Data-protection laws such as the GDPR, India's DPDP Act or Brazil's LGPD push the other way, toward deleting personal data you no longer need. Confirm the rules for each entity with a qualified tax advisor and legal counsel before you choose the archive design.
Red flags in migration quotes, and a scoping rule to apply first
When you compare integrator proposals, these signal that cost is being pushed onto you or into later change orders:
- "Client responsible for data cleansing" with no estimate of hours. That's your 1,800-hour line, unpriced.
- Only one or two mock loads for three source systems.
- Fixed fees built on a narrow record-count assumption. Ask what happens if item counts come in 30% higher after profiling.
- No reconciliation deliverable. If the vendor isn't producing tie-out reports, your finance team will build them by hand.
- Archiving left out entirely, or quietly assumed to mean keeping the old systems switched on.
- Profiling scheduled after the fixed price is set. Ask for a paid profiling phase first, then a re-quote.
Before you sign anything, apply one scoping rule. Migrate master data, open transactions and opening balances, plus only as much history as the business will actually query in the new system, which is usually one to two years of summarized balances. Archive the rest. Every extra year of detailed transactions across three systems adds mapping, mock load and reconciliation effort on both sides of the budget.
Next step: commission a fixed-price profiling exercise across all three systems, typically four to six weeks. Use its record counts and duplicate rates to rebuild the table above with your own volumes and local labor rates. Then ask every shortlisted integrator to quote against that same baseline.
Quick answers
What does ERP data migration typically cost for a mid-sized company?
For a single-source migration at a 100–300 user company, external fees often run $100,000 to $300,000. Consolidating several legacy systems can push that to $300,000 to $550,000 or more. Internal staff time can add roughly 50% to 100% on top of the external spend. These are typical 2026 market ranges that vary by region and contract, so confirm with vendors against your own record counts.
Should we migrate all historical transactions into the new ERP?
Usually not. Most companies migrate master data, open transactions and opening balances, plus one to two years of summarized history, and archive the rest in a searchable store. Migrating full transaction history multiplies mapping, testing and reconciliation effort, and the new system rarely needs that detail day to day.
Can data cleansing be fully outsourced?
Partly. Consultants and data quality tools can find duplicates, invalid codes and formatting problems at scale. Your own staff still have to decide which records survive, which commercial terms are correct and which items are obsolete. Make business owners accountable for those decisions, and backfill their routine work so the decisions get made on time.
How long must we keep legacy ERP data after migration?
It depends on each country you operate in and on your industry. Retention periods for accounting and tax records commonly fall between five and ten years. Data-protection laws may require you to delete personal data sooner. Have a qualified tax advisor and legal counsel in each jurisdiction confirm the periods before you design or decommission an archive.
Written by the Freight Base System editorial team. Prices, features and rules change often - confirm details with the vendor or a licensed professional before you decide.