IFS Cloud implementation · Data migration · Independent IFS Cloud practice
Settle your data before IFS Cloud goes live, not in acceptance testing
A data-first IFS Cloud implementation settles which legacy data moves, cleans it in the source system and tests the business on full-volume migrated data before go-live. The aim is to take the process decisions that bad data hides out of user acceptance testing, where they are most expensive, and into the months before it.
Key takeaways
- Migrate active data, not the whole history: a written rule for "active" (for example, activity in the last 24 months) keeps obsolete parts and inactive suppliers out of IFS Cloud.
- Clean in the legacy system before extraction. A missing customer ID or payment term is a process decision, and it is cheaper to take before the load than during testing.
- Change the process to fit IFS Cloud standard instead of customising IFS Cloud to copy the legacy system. The IFS Cloud documentation says: "Minimize tailorings and in particular customizations." Each customisation you keep has to be reassessed at every update.
- Test on full-volume migrated data, end to end. Hand-picked sample data proves the software works; only real volume shows whether tax codes and postings come out right.
- Go-live does not end the data work. A master data board with named owners keeps the clean state from decaying.
An ERP project exposes every inconsistent workflow and every gap in master data, because IFS Cloud enforces rules that the legacy system let users bypass. Whether that exposure becomes a list of fixes or a stalled go-live depends on when the data is dealt with. The five steps below put data first in the plan. If your project has already stalled, start with IFS Cloud Implementation Rescue: Diagnose, Decide, Recover instead.
- Scope. Define what counts as active data and migrate only that.
- Cleanse. Fix missing and inconsistent data in the source system before extraction.
- Align. Move the process onto IFS Cloud standard and turn person-dependent checks into configuration.
- Validate. Load full volumes repeatedly and run end-to-end business flows on migrated data.
- Sustain. Give master data named owners after go-live.
1.Which legacy data should move to IFS Cloud?
Only data the business will use after go-live should move to IFS Cloud: open balances, open orders and the master data those records and the next transactions need. Migrating 15 years of history "just in case" fills search results and lists of values with obsolete parts and suppliers nobody buys from, and every extra record has to pass the same validation as the active ones.
The scope rule has to be written down and agreed by the business, not decided by the person writing the extraction query. A typical rule says a part, customer or supplier is active if it had a transaction in the last 24 months or appears on an open document. The rule is tested against the source early, because the first count of "active" records usually starts the real discussion about what the business still sells and buys.
History that stays behind still has to be readable for audits and customer questions. How to keep it accessible and switch the old system off is described in Legacy ERP After IFS Cloud Go-Live: Archive and Switch-Off.
2.Why clean data in the source system before extraction?
Data cleaned in the source system before extraction keeps process decisions out of user acceptance testing. The plan "we will clean the data after we load it into IFS Cloud" moves every unanswered question, such as which payment term a customer really has or which of two supplier records is the right one, into the phase where the business should be signing off scenarios.
A data gap is usually a process gap. A missing customer ID means nobody owns customer creation; three spellings of the same supplier mean nobody checks for duplicates. Asking the business to fill these fields before extraction forces the owner question early, while there is still time to answer it.
IFS Cloud provides Data Migration Manager for the load itself. The IFS technical documentation describes it as "a standalone Tool that store, harmonizes, cleans data, flags for duplicates, converts and deploys all migration data", with "inbuilt data cleansing and validation" (Data Migration Manager, IFS Cloud 25R1 technical documentation). Validation against basic data catches records that reference a code not yet defined in IFS Cloud; in a thread on the IFS Community, a customer record failing this check shows the data status "Error, Basic data". The tool reports the problem; deciding the right value is still a job for the business owner of that data. The tool mechanics are covered in IFS Cloud Data Migration: FndMig & DMM Field Guide.
3.Should IFS Cloud be customised to match the old process?
A legacy process was built around the limits of the legacy system, so copying it into IFS Cloud with customisations carries those limits forward and adds update cost. IFS Cloud ships twice-yearly Release Updates, which the IFS documentation describes as "twice-yearly updates containing new functionality as well as cumulative fixes of all severities/priorities", and its advice for staying current is to "Minimize tailorings and in particular customizations" (IFS Cloud 25R1 Fundamentals). Every customisation kept has to be assessed again at each update.
The harder part of alignment is the knowledge that lives in people rather than in the system. A rule such as "the warehouse lead checks the label by hand before shipping to this customer" works until that person is on holiday. In a data-first project each such rule is written down during design and moved into basic data or configuration in IFS Cloud, so the system enforces it and the migration loads the values it needs.
Where standard really does not fit, the deviation needs a recorded owner and reason. The reasoning per object is set out in Clean Core in IFS Cloud: Purchasing & Inventory.
4.How should migrated data be tested before go-live?
Migrated data should be tested by loading full volumes repeatedly and running end-to-end business flows on the result. Testing with perfect, hand-picked sample data proves that IFS Cloud works; it does not prove that the business can run on its own data.
Each trial load is reconciled against the source with written rules, and then the business runs its scenarios on the loaded data: an order from entry to invoice, a purchase from requisition to supplier payment, a stock movement through to posting. These runs find the defects that sample data hides, such as a tax code mapped wrongly for one country or a posting that lands on the wrong account for one product group. The same full-volume logic applies to performance, which is the subject of Go-Live Is a Load Test. The checks per trial load are listed in The Data Migration Governance Checklist.
The data migration strategy and the readiness checklist are the two documents that hold this together. The Data Migration Strategy Template (IDK-14) records scope rules, cleansing ownership and the trial-load sequence, and the Data Migration Plan and Readiness Checklist (IDK-15) turns them into dated tasks. Both are part of the IFS Cloud Implementation Documentation Kit.
5.What keeps data clean after go-live?
Data stays clean after go-live only when each master data area has a named owner and changes go through an agreed route. Treating go-live as the finish line lets duplicates, free-text workarounds and unofficial codes return after the project team leaves.
A small master data board with a fixed meeting rhythm approves new codes and structural changes, reviews duplicate reports and decides on requests for new fields. Its owners are business people, not the project team that leaves after hypercare. How processes drift once the project team has gone is described in Process Drift After Go-Live in IFS Cloud.
6.The five steps: typical trap and the fix
Each step of a data-first IFS Cloud implementation replaces one common shortcut with a decision that is taken earlier and written down.
| Step | Typical trap | Fix |
|---|---|---|
| Scope | All history migrated "just in case" | Written rule for active data, agreed by the business and tested against the source early |
| Cleanse | "We will clean it in IFS Cloud after the load" | Business owners fill missing IDs and terms in the source before extraction |
| Align | IFS Cloud customised to copy legacy habits; checks done by one person from memory | Process moved onto standard; person-dependent rules turned into basic data or configuration |
| Validate | Tests on hand-picked sample data | Repeated full-volume loads, reconciliation and end-to-end flows on migrated data |
| Sustain | Data ownership ends with the project | Master data board with named business owners after go-live |
7.Frequently asked questions
What is a data-first IFS Cloud implementation?
A data-first IFS Cloud implementation decides which legacy data moves, cleans that data in the source system and tests the business on full-volume migrated data before go-live. It moves process decisions that bad data hides out of user acceptance testing and into the design and migration phases.
Why not migrate all historical data to IFS Cloud?
Migrating all historical data to IFS Cloud, such as 10 years of closed orders, fills the production environment with records nobody works on and adds every one of them to validation and reconciliation. The usual approach is to migrate open balances and active master data, and to keep closed history in an archive or reporting repository where it stays readable.
What is the biggest risk in IFS Cloud data migration?
The biggest risk in IFS Cloud data migration is finding bad data only after it has been loaded. When master records reference basic data that does not exist in IFS Cloud, every record that depends on it fails validation, and the transactions that need those records cannot be entered. Cleaning in the source system before extraction finds these gaps before the load instead of after it.
Should we customise IFS Cloud to match our old process?
Customising IFS Cloud to match an old process carries the limits of the legacy system into the new one and adds work at every twice-yearly Release Update. The IFS Cloud documentation says: "Minimize tailorings and in particular customizations." Every customisation kept has to be reassessed at each update, so adopting IFS Cloud standard and changing the business process is the default; a customisation needs a recorded owner and business reason.
What does IFS Data Migration Manager do in an IFS Cloud migration?
IFS Data Migration Manager is the IFS Cloud tool for preparing migration data and deploying it between environments. It has built-in data cleansing and validation and flags potential duplicates. It reports records that fail validation against basic data; choosing the correct value remains a decision for the business owner of that data.
8.About the author
Dariusz Myśliwiec: 18 years hands-on with IFS (Apps 7.5 to 10 and IFS Cloud). IFS Certified Associate Consultant. PRINCE2® 7. Independent practice based in Kraków, delivered remotely across Europe and globally. You talk to the consultant who does the work, not a sales layer.
Selected clients from my IFS work: Fugro · LGC · BVI Medical · Betafence (PRÆSIDIAD) · Barlinek · NGK Ceramics · Newag · Oleofarm.
IFS is a registered trademark of IFS AB; this practice is not affiliated with IFS AB.
Put the data decisions on paper before the first trial load
The Implementation Documentation Kit gives you 24 templates for an IFS Cloud project, including the data migration strategy, the migration plan with readiness checklist, the test strategy and the cutover plan.
See the Implementation Documentation Kit Read: The Data Migration Governance Checklist