Independent IFS Cloud practice · Master data
The golden record: one trusted version of each part, customer and supplier, and where IFS Cloud fits
A golden record is the single, agreed version of a master data entity, such as a part, customer or supplier, built by matching duplicate records from several systems and keeping the best value for each attribute. IFS Cloud can hold that record and protect it once it exists, but building it from several source systems is a master data management process with rules that the data owner decides.
Key takeaways
- The record is created in five stages: ingest, clean and standardise, match, merge with survivorship rules, publish.
- Survivorship rules decide which source wins for each attribute, for example the finance system for bank details and the CRM for contact data. The data owner signs them.
- For a company that runs purchasing and inventory in IFS Cloud, IFS Cloud is the natural system of record for parts and suppliers, so the merged record is loaded into it and maintained there.
- Data migration to IFS Cloud is the cheapest moment to merge master data, because duplicates are merged before they collect orders and invoices.
- After go-live, the IFS Cloud duplicate check flags likely duplicates when leads, accounts and supplier requests are created, and permission sets limit who may create the master records themselves.
1.What is a golden record in master data?
A golden record is the version of an entity that every process and report is allowed to trust. It sits at the centre of master data management (MDM): when the same supplier exists in the ERP, in a CRM and in an old purchasing database, under three spellings and with two bank accounts, it is the one supplier that remains after matching and merging, with each attribute taken from the source the business trusts most for it.
Such a record is a governance outcome, not a feature. It needs a data owner who decides what “the same supplier” means and which source wins, which is why it belongs to the wider topic described in the definition of data governance.
2.How is a golden record created?
The record is created in five stages, and the decisions sit in stages three and four.
- Ingest. Extract the records for one domain from every source that holds them: ERP, CRM, PLM, spreadsheets kept by a department.
- Clean and standardise. Fix formats, casing, country codes, units of measure and obvious errors, so that matching compares like with like.
- Match. Identify records that describe the same entity, first by deterministic keys (VAT number, supplier part number, DUNS number where it is kept), then by similarity of names and addresses. Every uncertain match goes to a steward.
- Merge with survivorship rules. For each attribute, apply the rule that decides which value survives: a trusted source, the most recent update, or the most complete value.
- Publish. Load the merged record into the system of record and send it to the systems that consume it, together with a cross-reference from old keys to the new one.
The cross-reference in stage five is the part projects forget. Open orders and historical reports still carry the old keys, as do interfaces, and without a mapping table nobody can trace a legacy document to the surviving record.
3.What are survivorship rules, and who sets them?
Survivorship rules state, attribute by attribute, which value survives when sources disagree, and the data owner for the domain sets them. Writing them per attribute, not per record, avoids the trap of choosing one “best” source that is right for addresses and wrong for payment terms.
| Attribute (supplier example) | Illustrative rule | Reason |
|---|---|---|
| Legal name, VAT number | Trust the finance system | Checked when invoices are posted |
| Bank account | Trust the finance system, and require a second person to confirm any change | Payment fraud targets this field |
| Contact person, phone | Most recent update wins | Changes often; the newest value is most likely correct |
| Payment terms | Trust the signed contract, entered by purchasing | Commercial term, owned by purchasing |
| Supplier category | Set by the master data team from a fixed list | Free text breaks reporting |
The table is illustrative, not a client record. The rules themselves are project decisions and belong in a decision log with their reason, as described in the IFS Cloud decision register.
4.Where does IFS Cloud fit in a golden record strategy?
IFS Cloud is a transactional system, so for parts, suppliers and, where sales run in IFS Cloud, customers it acts as the system of record that holds the trusted record and uses it in every order. It does not merge records from Salesforce, a legacy ERP and a PLM tool into one on its own; that matching and survivorship work happens in an MDM tool or in the migration process before load.
Data migration is the natural moment to do this merge. In the IFS Data Migration Manager, a migration project imports legacy source data and maps it to target tables before migration jobs load it. Matching and merging happen between import and load, so each entity arrives in IFS Cloud once. How the Data Migration Manager is set up is explained in IFS Cloud Data Migration Manager: containers, mapping, scope and IFS Data Migration Manager step by step.
After go-live, two standard functions protect the record. The IFS Cloud duplicate check runs when a new lead, account or supplier request is created and compares it with existing records using configured rules; the action can block the save, warn with a list of possible duplicates, or notify. Permission sets limit who may create the record itself, so that new suppliers enter through a request that the duplicate check can see.
5.What goes wrong, and how do you fix it?
Master data consolidation fails on four predictable problems, each with a specific counter-move.
| Problem | Counter-move |
|---|---|
| Poor source quality | Clean and standardise per source before matching, and send source errors back to the system that created them |
| Conflicting values | Survivorship rules per attribute, signed by the data owner, with uncertain matches reviewed by a steward |
| Too many systems at once | Start with one domain and one or two sources; suppliers carry payment risk and parts drive every order line, so either is a good first domain |
| Duplicates return after go-live | Creation only through a request, the duplicate check on that request, and a monthly measure of new duplicates reviewed by the owner |
A master record maintained as a product, with an owner, a published definition and quality measures, is the idea behind implementing IFS Cloud master data as data contracts. The steps to put the owners and rules in place across a project are in implementing data governance in IFS Cloud. If your migration is the moment to clean master data, my professional data migration services cover matching and survivorship decisions, then reconciliation before load.
6.Frequently asked questions
What is a golden record in master data management?
A golden record in master data management is the single, agreed version of an entity such as a part, customer or supplier. It is built by matching duplicate records from several systems and keeping, for each attribute, the value chosen by survivorship rules. Every process and report then uses that one record.
What is the difference between a golden record and a master record?
Golden record and master record can mean the same thing. Where they are distinguished, the master record is the container for an entity in an MDM system, and the golden record is the result of cleaning, matching and merging, which is the version published to the business systems.
How are conflicting values handled when building a golden record?
Conflicting values are resolved by survivorship rules written per attribute, such as trusting the finance system for bank details or taking the most recent update for a phone number. The data owner for the domain approves the rules, and uncertain matches go to a data steward for a manual decision.
Does IFS Cloud create golden records automatically?
IFS Cloud does not merge records from several external systems into golden records on its own. It is a transactional system that holds and protects master data once it exists: the duplicate check catches similar leads, accounts and supplier requests at creation, and permission sets limit who may create records. Building golden records from several sources needs an MDM process, and one place to run it is data migration with the IFS Data Migration Manager.
Which master data domain should a golden record project start with?
A golden record project should start with one master data domain that has a clear owner and a visible cost of duplicates, such as suppliers or parts in an IFS Cloud company. Suppliers carry payment risk and parts drive every order line, so a single domain with one or two sources shows results before the method is extended.
7.About the author
Dariusz Myśliwiec: 18 years hands-on with IFS (Apps 7.5–10 and IFS Cloud), IFS Certified Associate Consultant, PRINCE2® 7. Based in Kraków, delivering remotely across Europe and globally through an independent practice. You talk to the consultant who does the work, not a sales layer.
Selected clients: 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.
Build the golden record before you migrate
Tell me which domain causes the most duplicates in your systems today. On a call we sketch the matching keys, the survivorship rules your data owner has to sign, and where the work fits in your IFS Cloud migration plan.
Book a call about your master data Read the definition of data governance
