Informatica MDM migration: the short answer
An Informatica MDM migration means moving an on-premises Multidomain MDM (MDM Hub) installation either to Informatica's own MDM SaaS on its Intelligent Data Management Cloud (IDMC), or off Informatica to another MDM platform. For most mid-sized companies the right first step is neither: it is a four to six week inventory of the match rules, survivorship rules, stewardship workflows and integrations you actually run, because that inventory, not the target platform, decides the cost, the timeline and the risk.
The question is urgent now for three reasons. Salesforce completed its acquisition of Informatica on 18 November 2025, for an equity value of about $8 billion, so the roadmap for every Informatica product now sits inside Salesforce. Informatica is steering on-premises customers to IDMC, and sells a packaged MDM SaaS Modernization service to do it. And the on-premises estate around MDM is ageing out: standard support for PowerCenter 10.5, the integration tool that feeds many MDM Hubs, ended on 31 March 2026, according to Informatica's published lifecycle data as tracked by Licenseware.
We work on Informatica, Reltio and Veeva Network implementations as an independent consultant, without a vendor partnership or licence to resell; our Informatica MDM consulting page describes that work. This guide covers the three paths open to you, what has to move whichever you choose, a phased plan that avoids a big-bang cutover, what drives cost and timeline, the extra work pharma and life sciences teams face with HCP and HCO masters, and when it is better to wait.
What changed, and why on-prem MDM customers are being asked to move
Informatica built its MDM business on an on-premises product, Multidomain MDM, with the MDM Hub at its core, Informatica Data Director (later Customer 360 and Supplier 360 interfaces) for stewards, and PowerCenter or Data Quality jobs feeding it. Over the last few years Informatica moved its development effort to IDMC, its cloud platform, where MDM runs as a SaaS service alongside cloud data integration and data quality.
Two things follow for an on-prem customer. First, each on-prem version has an end of standard support date in Informatica's Product Lifecycle Guide, after which you pay for extended support or run unsupported. Dates differ by version and hotfix level, so check the current guide on the Informatica Knowledge Base for your exact build rather than trusting a date quoted second-hand. Second, new capability, including the AI features Informatica now markets as CLAIRE, lands in the cloud product first or only.
The Salesforce acquisition added a third consideration. Informatica's MDM is now owned by a CRM vendor. If Salesforce is central to your estate that may be a reason to stay; if your CRM is Veeva Vault CRM, Microsoft Dynamics or something else, it is a reasonable moment to test whether an independent MDM platform fits better. Competing vendors have noticed: Semarchy, Reltio and Syncari all published Informatica migration pages in 2026, each recommending its own platform. None of them is a neutral source, and neither is Informatica's modernization offer.
The three paths for an Informatica MDM migration
Modernize to MDM SaaS
Informatica's documented route converts on-prem MDM Hub objects into MDM SaaS configuration, then migrates the data. Its own documentation says the modernized objects, such as data models, reference data and data quality rules, must be configured and available in MDM SaaS before data is migrated, and directs customers to Informatica Global Customer Support for the data migration step itself. Informatica also sells an MDM SaaS Modernization Pack, a professional services package that translates on-prem functionality into native MDM SaaS components and covers the initial data load plus the daily ingress and egress integrations.
Re-platform
The usual alternatives are Reltio, Semarchy, Profisee, Stibo Systems, or Veeva Network for life sciences customer data. A re-platform is a new implementation that reuses your knowledge: the match logic, survivorship decisions, source trust rankings and steward procedures are portable as documented rules even though the configuration is not. Vendors pitch short timelines (Semarchy, for example, advertises go-live in under 12 weeks), but those figures come from the vendors and depend on how much of your current logic is written down.
Stabilise and wait
If your version is still in standard support, or extended support is affordable, staying put for 12 to 18 months while you document rules, clean sources and decide properly is a legitimate choice. It is the wrong choice only if it becomes a way of not deciding.
Which path fits which situation
Salesforce is your CRM and you license other IDMC services
Modernizing to MDM SaaS is usually the shortest path. Your integration and data quality work stays on one platform and Informatica's tooling does part of the conversion.
- Lean towards: MDM SaaS
- Validate: steward UI with real users
Pharma company mastering mainly HCPs and HCOs on Veeva CRM
Veeva Network was built for this domain and connects natively to Vault CRM and OpenData. An enterprise MDM may still be needed for products, suppliers or patients.
- Lean towards: Veeva Network for customers
- Validate: what other domains the hub masters today
Several domains, heavy relationships and hierarchies, non-Salesforce CRM
This is where a structured comparison of Informatica MDM SaaS against Reltio, Semarchy or Profisee earns its cost. Run each against your own sample data, not a demo set.
- Lean towards: competitive evaluation
- Validate: three-year cost, not year-one
Three or four sources, batch updates, one main consumer
You may not need an enterprise platform at all. A master built in your warehouse with documented match rules and a review queue can replace a lightly used MDM Hub.
- Lean towards: warehouse-based master
- Validate: real-time and audit needs
Version still supported and a CRM change is coming
Wait. Migrating the master and the CRM at the same time doubles the reconciliation work and makes every defect harder to trace.
- Lean towards: stabilise and document
- Validate: support end date for your exact build
Nobody can explain the current match rules
Do not choose a platform yet. Reverse-engineer and document the rules first, because every path depends on them.
- Lean towards: rules discovery
- Validate: match results on a labelled sample
What has to move, whichever platform you choose
Data model
Base objects, child tables, lookups and relationship types in the MDM Hub, mapped to the target model. Expect to drop unused columns and rename what was never clear.
- Entities and attributes
- Hierarchies and relationship types
- Reference data and lookups
Match rules
Exact and fuzzy match columns, match rule sets, search levels and thresholds. These rarely convert one to one, so they must be re-tested on a labelled sample in the new platform.
- Rule sets and thresholds
- Auto-merge versus manual-review splits
- Measured false and missed match rates
Survivorship and trust
Source trust settings, decay, validation rules and any custom survivorship logic. Write them down attribute by attribute before you configure anything.
- Source rankings per attribute
- Recency and completeness rules
- Validation rules that downgrade trust
Stewardship
Data Director or Customer 360 task workflows, queues, roles and approval steps, plus the audit history your compliance team relies on.
- Tasks, queues and roles
- Approval and escalation steps
- History you must retain
Integrations
Inbound landing and staging jobs, real-time calls through the Services Integration Framework or Business Entity Services, and outbound feeds to CRM, ERP and the warehouse.
- Batch loads and their schedules
- Real-time APIs and their consumers
- Outbound publish and message queues
Cross-references and IDs
The cross-reference table that maps every source key to the golden record ID. Downstream systems store those IDs, so losing or reissuing them breaks reports and CRM links.
- Source-to-golden key mapping
- Merge and unmerge history
- ID strategy in the target
Customisations
User exits, custom Java, cleanse functions and hand-written SQL. These are the items vendors' automated converters do not cover, and they are where hidden effort sits.
- User exits and Java code
- Custom cleanse functions
- Scheduled scripts outside the Hub
A phased Informatica MDM migration plan
Inventory and assess (4 to 6 weeks)
Export the Hub configuration, list every source, consumer, rule and customisation, and profile the data in each source. Confirm the support end date for your exact version and hotfix from Informatica's lifecycle guide. The output is a written inventory and a short list of paths worth pricing.
Decide with your own data
Give each shortlisted platform the same anonymised or de-identified sample and the same match test. Compare match quality, steward experience and three-year cost, including integration licences and internal admin headcount.
Rebuild model and rules
Configure the target model, then rebuild match and survivorship rules and test them on a labelled sample until false and missed matches sit within thresholds the business has signed off.
Rebuild integrations
Re-point inbound sources and rebuild outbound feeds, preserving or remapping cross-reference IDs so CRM, ERP and warehouse links survive.
Parallel run and reconcile
Run old and new side by side for at least one full business cycle. Reconcile golden record counts, merge decisions and downstream outputs, and investigate every difference until each one is explained.
Cut over by domain or region
Move one domain or one country at a time, keep the old Hub read-only for an agreed period, and switch off only after the reconciliation sign-off.
Pharma and life sciences: what is different about HCP and HCO masters
In pharma, the MDM Hub usually masters healthcare professionals and organisations, and those records feed everything that matters commercially: CRM call targeting, incentive compensation, specialty pharmacy and hub data matching, and transparency reporting. A migration that changes even a small share of golden record IDs breaks the joins those processes depend on.
Three things deserve extra care. First, reference data: if your Hub validates HCPs against a reference source, decide whether that stays the same, because switching from one HCP universe to another at the same time as the platform doubles the change. Our comparison of Veeva OpenData and IQVIA OneKey covers that decision. Second, CRM: if you are also moving from Veeva CRM to Vault CRM, sequence the two projects rather than running them together. Third, identifiers: NPI and state licence numbers make deterministic matching strong for US prescribers, but organisations, affiliations and non-US records still rely on fuzzy rules that need fresh testing.
Patient data adds a further constraint. If the Hub holds any patient or patient-linked identifiers, the target must be covered under your Business Associate Agreement, and the test samples should be de-identified. The issues we see most often are written up in common data quality challenges in pharma MDM.
What drives the cost and timeline of an Informatica MDM migration
There is no honest single price for this work, and any figure quoted before an inventory is a guess. Syncari, itself a competing vendor, cites 12 to 24 months for typical Informatica implementations; vendors of the alternatives quote much shorter timelines for their own platforms. The spread comes from the same few drivers in every project:
- Number of domains and sources. One customer domain with five sources is a different project from customer, product and supplier with twenty.
- Custom code. User exits, Java extensions and hand-written cleanse functions are rebuilt by hand on every path.
- Real-time integrations. Each synchronous consumer needs to be re-pointed, tested and monitored.
- Documentation of rules. If match and survivorship rules exist only in the configuration, discovery takes weeks before any build starts.
- Stewardship history and audit needs. Regulated teams may need years of change history carried across or archived in a queryable form.
- Licences. Subscription cost for the target platform, plus any integration platform it needs, over three years rather than one.
For budget ranges on MDM projects more generally, and the line items buyers forget, see our breakdown of MDM implementation cost. For how rules are tested before and after a move, see types of matching in MDM and survivorship and the golden record, and for the wider service, our master data management services.
Common mistakes in Informatica MDM migrations
Choosing the platform before the inventory
Teams pick a target from a demo, then discover custom code and integrations nobody counted. The inventory costs weeks; skipping it costs months.
Treating rules as configuration to convert
Match and survivorship rules behave differently on a new engine even when they look identical. Re-test every rule set on a labelled sample in the new platform.
Reissuing golden record IDs
New IDs break every downstream report, CRM link and data warehouse join that stored the old ones. Decide the ID strategy in week one.
Skipping the parallel run
Without a full cycle of side-by-side reconciliation you find merge differences in production, usually through a sales report that no longer adds up.
Forgetting the stewards
The steward interface changes on every path. Train stewards on real cases from your data before cutover, not after.
Moving everything
Old attributes, retired sources and unused relationship types do not need to come along. Migrating them carries their quality problems into the new platform.
When not to migrate yet
Questions about Informatica MDM migration
Is Informatica ending support for on-premises MDM?
Can Informatica MDM Hub be migrated to MDM SaaS automatically?
How long does an Informatica MDM migration take?
What are the main Informatica MDM alternatives?
Does the Salesforce acquisition mean Informatica MDM only works with Salesforce?
Should pharma companies move HCP mastering to Veeva Network instead?
Do we need an Informatica partner to migrate?
Deciding what to do with your Informatica MDM Hub?
Tell us which domains and sources your Hub masters and which version you run. We will come back with how we would run the inventory, what it would tell you, and which paths look worth pricing.
Part of our Data Management and Engineering hub. Start with our data management services.