Master Data Management Strategy; Balancing Control with Self-Service
This is the continuation of the transcript of DM Radio’s program titled “The Consumerization of Business Intelligence: How and Why.”
Byron Igoe: Yeah, actually I think I was on another DM Radio segment last year, talking about master data management and some other topics around governance and controls in a world where you need to still enable self-service. InetSoft has their own ways of addressing this.
We can support the IT layer, where they have got tight controls on things, but enable self-service for the users. But I am sure that various vendors are going to come out with their own solutions to the problem as well.
Eric Kavanagh: Right. And Tracie, do you want to chime in?
Tracie Kambies: Yeah. I was going to say that I was on that same DM Radio last year as well, and I think that the same concepts apply to the Big Data space and making things more consumable, the data more consumable. I think that what we are going to be seeing is different models for how you leverage it and different organizations are going to create these hybrid models.
Some organizations might be able to be more centralized, in that the data is served up through Big Data solutions and then consumed from that centralized organization and shifted out. But then there is this decentralized, more categorical type of management, governance, and ownership of that data, there can be more of the federated models.
Collaboration and Ownership Models in Master Data Management
So I think that it's just going to depend on the business needs. What does the business really need and what are they trying to do with it? What are the fashions and trends that they are going to need to stay in front of? And the speed of their business is going to have a lot to say about how they structure their ownership models of the data.
I think the other big point I would say about this is that it's becoming an environment in the business world and across our industry where collaboration is happening more frequently. And so ownership is going to be about who is collaborating together and how they determine it at that point in time. One business problem that people are trying to solve is offering health care. Health care providers and insurance companies are working together better so that patients have better care and better preventative care.
Well, there is a lot of information ownership going on there over who owns the patient data. Well, I think that if they do decide to collaborate and our data solutions can offer the information faster, better, and in a more consumable way, their ownership is now joint. So it's an interesting concept. I think that there is a lot of growth and opportunity in the space for people to define it based on what they are trying to do and what they are really trying to deliver to their end consumers.
What Is StyleBI's Role in Master Data Management?
StyleBI is not a traditional Master Data Management system — it’s a cloud-native business intelligence and semantic-layer platform that complements MDM by providing a governed, business-friendly view of master entities, surfacing data quality and lineage, enforcing consistent definitions, and enabling controlled self-service access for analytics and downstream consumers.
Quick clarification (important)
- MDM systems (Informatica, Profisee, Stibo, Pimcore, etc.) are built to be authoritative systems of record for golden master data, provide workflows for stewardship, reconciliation, and push/pull synchronization between systems.
- StyleBI is a BI/semantic layer and analytics platform that connects to many source systems and data warehouses, then exposes curated models, metrics, and governed dashboards to users and applications. It helps you use master data reliably in analytics and decisions, but it generally doesn’t replace an MDM hub responsible for creating and synchronizing golden records.
The practical roles StyleBI plays in an MDM ecosystem
- Semantic standardization & one source of truth for analytics: StyleBI lets organizations define canonical, business-friendly metric and dimension definitions (customer, product, location) in a semantic layer. When analytics, dashboards, or embedded reports consume those definitions, the business sees consistent values and calculations — reducing the “which ROAS is correct?” fights. This is functionally vital to MDM’s goal of consistency across consumers.
- Surface data quality issues to stewards: By comparing source values, showing nulls, duplicates, conflicting attributes or sudden distribution changes in dashboards, StyleBI helps data stewards spot MDM problems early. Dashboards can highlight candidate duplicate customers, mismatched product hierarchies, or missing mandatory attributes so the MDM team can remediate in the master registry.
- Data lineage & provenance for trust & audits: StyleBI can show where a metric or attribute came from (source table/field + transform) and which transformations were applied in the semantic model. That lineage helps auditors and stewards understand whether discrepancies are modeling artifacts or real master-data errors.
- Row-level security & governance boundaries: MDM often requires that certain users only see specific slices of master records. StyleBI enforces RLS and role policies at query time so downstream analytics respect governance rules without copying or exposing golden records unnecessarily.
- Integration & connectivity: StyleBI connects to numerous operational systems, warehouses, and lakes (connectors and query endpoints). That makes it a practical place to join master records from the MDM hub with transactional and behavioral data for downstream reporting, cohorting, and ML feature validation.
- Serving the golden record to consumers: Where MDM exposes golden records to a data warehouse (consolidation/coexistence styles), StyleBI acts as the consumer and the consistent access layer that presents those golden attributes in business terms to analysts, apps, and embedded UIs.
Where StyleBI is not a substitute for MDM
- No canonical master reconciliation workflows: Typical MDM chores — merging records, governance workflows for steward approval, survivorship rules, multi-system synchronization — are functions of an MDM platform, not a BI semantic layer.
- Not the authoritative system of record: StyleBI should not be the SOR for master attributes; it should surface and present the SOR’s golden data, or point to it in the warehouse. Treat StyleBI as the “trusted consumption layer,” not the source of truth for writing master data back into operational systems.
Example architectures: how StyleBI and MDM can work together
Consolidation/Hub style: An MDM hub consolidates customer/product masters into a golden table in the data warehouse. StyleBI connects to the warehouse, builds semantic models on top of the golden tables, enforces metric definitions and RLS, and exposes dashboards to business users. Stewards monitor dashboards for anomalies and trigger MDM workflows when issues appear.
Coexistence style: Multiple systems continue to hold operational master attributes; the MDM hub coordinates survivorship and synchronization. StyleBI connects to both the MDM outputs and operational sources to compare values, expose lineage, and support reconciliation reports that feed stewardship processes.
Operational recommendations — what teams should do if they have StyleBI + MDM
- Define the business glossary inside the semantic layer (canonical names, attribute meanings, allowed values). Use that glossary as the single vocabulary the analytics team and MDM stewards reference.
- Publish an MDM→Warehouse contract: Agree on field names, data types, refresh cadence, and confidence metadata so StyleBI models can rely on stable inputs.
- Create stewardship dashboards: Build dashboards that surface duplicates, schema shifts, cardinality spikes, and key missing attributes for stewards to action. StyleBI is excellent at making those problems visible.
- Version metrics & record provenance: When a master attribute changes (e.g., product category taxonomy), version it and annotate dashboards—so historical reports remain interpretable. StyleBI’s semantic layer is the natural place for those versions.
- Don’t write golden records from BI: Avoid write-backs to operational systems from dashboards except through controlled MDM processes; use BI annotations or work-item links to the stewardship workflow instead.
| Previous: Consumerization of BI for More Intuitive Data Analysis | Next: Enabling Data Drilldown with Information Architecture |