Data migration services: from the old system to the new one, complete and verified
Data migration services move your business data from old systems into new ones: an ERP replacement, a new CRM, a move from spreadsheets or a legacy database, or a switch between SaaS providers. We profile the data, clean it, map it to the new structure, test the migration in full before go-live and reconcile every total, so the new system starts with data people trust. Everything is processed in the EU, and the scripts and documentation are yours.
- You see the result before it goes live
- Fixed price for the agreed scope, in writing before work starts
- Source code, domain and accounts in your company's name
Why data migration decides whether a new system works
A new ERP or CRM is only as good as the data in it on day one. When customers are duplicated, product codes do not match, open invoices are missing or balances do not agree, users stop trusting the system in the first week and go back to their spreadsheets. Data migration is often planned as a technical step at the end of a project; in practice it is the part with the most business decisions in it, and it should start as early as the system selection.
Migrations we handle
- ERP replacement: customers, suppliers, items, price lists, bills of materials, open orders, stock, open receivables and payables, opening balances and history. See ERP systems.
- CRM migration: accounts, contacts, activities, opportunities, consents and attachments. See custom CRM.
- Spreadsheets and access databases into a proper system with users and history.
- Legacy databases from old in-house or discontinued software, often without documentation.
- Document migration from file servers and shared drives into a document management system, with metadata.
- SaaS to SaaS and SaaS to your own platform, through exports and APIs.
- Database and platform moves as part of a cloud migration.
How a data migration runs
- Inventory and scope. Which sources, which data objects, how much history, what goes to an archive instead.
- Profiling. We read the source data and measure it: volumes, empty fields, duplicates, invalid values, broken links between records. This replaces guesses with facts and is the basis for the plan.
- Mapping. Field by field and value by value from source to target, with the transformation rules, agreed and signed off with the data owners in your team.
- Cleaning. Duplicates merged, codes harmonised, outdated records flagged for a decision. Rules are applied automatically where possible; your team decides the rest.
- Build the migration. Scripts that extract, transform, validate and load, repeatable at any time, with every rejected record logged and explained.
- Trial migrations. At least one full run into a test copy of the new system, checked by your users with real tasks.
- Reconciliation. Record counts, totals, balances and samples compared between source and target, documented.
- Cutover. A planned freeze of the old system, the final run, reconciliation and go-live, with a rollback plan.
Our guide to data migration cost explains how a migration like this is estimated, step by step.
Reconciliation: how you know nothing is lost
| Check | Example |
|---|---|
| Counts | Number of active customers, items and open orders in source and target |
| Totals | Sum of open receivables and payables, stock value, order value |
| Balances | Opening balances per account agree with the closing balances of the old system |
| Samples | A set of records checked field by field by the people who know them |
| Rejects | Every record not loaded, with the reason and the decision taken |
| Usability | Users complete real tasks in the test system with migrated data |
Personal data in a migration
Most migrations move personal data about customers, contacts and employees, so the GDPR applies to the migration itself. Article 5(1)(d) requires personal data to be accurate and kept up to date, and Article 5(1)(e) not to be kept longer than necessary, which makes a migration the right moment to delete what you no longer need rather than carry it over. Article 32 requires appropriate security, so migration files are encrypted, access is limited to named people, test systems use only the data they must, and temporary copies are deleted after cutover, with a record of when. As your processor, we sign a data processing agreement and keep the data in EU regions throughout.
Getting your data out of a SaaS provider
Leaving a cloud or SaaS service is easier in the EU since the Data Act applied on 12 September 2025. Read on 12 November 2026, Regulation (EU) 2023/2854 requires providers of data processing services to let customers port their exportable data and digital assets, with a contract that sets a maximum notice period of two months and a transitional period of up to 30 calendar days for the switch, during which the provider must give reasonable assistance (Article 25). From 12 January 2027, providers may not impose switching charges (Article 29). We plan the export with you, use the provider's APIs where the export function is not enough, and check that what came out is complete. Our EU Data Act guide explains the rules in full.
Archive instead of migrate
Not all history belongs in the new system. Old closed orders, inactive customers and years of transactions slow the new system down, make the migration longer and keep personal data longer than needed. A common pattern is to migrate master data, open items and a defined period of history, and to move the rest into a read-only archive that people can search and that is kept for exactly the retention period the law requires for accounting or other records, then deleted. The archive costs little to run, is easy to audit, and lets you switch the old system off.
What you get
- A data profile of every source: volumes, quality issues and decisions needed.
- Signed-off mapping and cleaning rules, field by field.
- Repeatable migration scripts, with logs of every rejected record.
- Trial migration reports and a final reconciliation report.
- A cutover plan with freeze times, responsibilities and rollback.
- Where agreed, a searchable archive of the history that was not migrated.
Common data migration risks, and what we do about them
- Poor data quality found late: profiling at the start, not in the week before go-live.
- Undocumented old systems: we read the database structure and the data itself, and confirm meanings with the people who used it.
- Business decisions left to the technical team: mapping and cleaning rules are signed off by your data owners.
- One-shot migrations: repeatable scripts and trial runs, so the final run is a rehearsed routine.
- Long downtime at cutover: history migrated in advance, only the final delta during the freeze.
Where data migration fits in a project
Data migration is part of most of our system projects: a new ERP, a CRM, a web app that replaces spreadsheets, or a move to the cloud. We also take on migrations alone, when another supplier implements the new system and you want the data in safe hands. The plan always starts from the processes the data supports; our process mapping guide shows how we map them.
You own it
As a standard term, the migration scripts, mapping documents and reconciliation reports are delivered to a repository in your company's own account, so the migration can be rerun or audited later. The data stays in EU regions, in systems in your company's name. You can keep us for the next migration, run the scripts yourself or hand them to another team. See prices.
Questions we get
What is data migration?
Data migration is moving data from one system to another: from an old ERP to a new one, from spreadsheets into a CRM, from an on-premises database to the cloud, or from one SaaS product to another. It includes extracting the data, cleaning it, transforming it to the structure of the target system, loading it and checking that everything arrived correctly.
What is the difference between data migration and cloud migration?
Cloud migration moves whole systems, servers and applications to a cloud platform. Data migration moves the data itself into a different system or structure. A cloud migration may need little data migration if the system stays the same; replacing an ERP always needs a full data migration, wherever the new one runs.
How long does a data migration take?
It depends on the number of sources, the volume and above all the quality of the data. A migration from one well-kept system can be done in weeks; one from several old systems with duplicates and gaps takes longer, mostly because of cleaning and decisions on what to keep. We give a plan after profiling the data, not before.
Do we need to migrate all historical data?
Rarely. Most companies migrate master data, open items and a defined period of history into the new system, and keep the rest in a read-only archive that meets the legal retention periods. This makes the new system faster and the migration cheaper, and is better for data protection.
How do you make sure no data is lost?
By reconciliation. Before go-live we compare record counts, totals and balances between source and target, check samples record by record with your team, and run at least one full trial migration. Every rejected record is logged with the reason, so nothing disappears silently.
Can you migrate data out of a SaaS provider?
Usually yes, through the provider's export functions and APIs. In the EU, the Data Act gives customers of data processing services the right to port their exportable data, with a transitional period of up to 30 calendar days for the switch under the contract, and from 12 January 2027 providers may not charge for switching.
Get a written quote
Tell us in a few lines what you need. We reply within one working day with a written quote: scope, timeline and a fixed price or an hour estimate. Asking is free and commits you to nothing.