Master Data Management can help to achieve this
Master Data Management (MDM) has been around for some time. There are many “scars” from the old implementations that leveraged rigid technology limiting to focus on one domain like customer or product. The technology has evolved, and it now enables business-driven, agile implementations with multi-domain MDMs. Now more than ever delivering incremental business value with MDM is achievable.
This document highlights some of the key business benefits MDM can help to achieve.
1. “Trusted data@scale” – Availability of trusted master data for cross-functional reuse is critical for efficient realisation of data-driven business objectives.
Different studies show that data professionals, like data scientists, spend 80% of their time in discovering and preparing data and just 20% on the actual analytics to create insights of the data.
Data reuse can remove a critical bottleneck for the realisation of data-driven business objectives as data professionals could avoid doing the same effort multiple times. Cross-functional data reuse can enable cost-efficient scaling of data consumption for a wide range of use cases.
There is a saying that data interoperability/integration grows data value exponentially. This becomes quite clear when you look at the above examples of facts. Plan or actual figures alone would have limited value, but when they are combined for a KPI that follows, how the actual business (for example sales) is developing against the plan then the value grows significantly (exponentially).
Total facts or KPIs (for example plan vs. actual) are meaningless without consistent master data that gives a business context e.g., customers & their segments, channels, products, geographies, or sales territories and over time/calendar. The facts may come from multiple data sources, but they need to share consistent master data that enables them to combine them to deliver meaningful business performance information.
Master data is the most used data. In any data platform – data lake or data warehouse – about 80% of the entities are master data entities. Transactional/interactional data is about 85% of the data volume.
Master data has the biggest impact in providing trusted data for cross-functional reuse. The most mature companies have started to realise this and that is why there is a lot of demand for Master Data Management (MDM).
Master data is an umbrella term that can be broken into the following subcategories:
- Core Master Data (often called master data) could be summarised as parties, products & places. Examples: customer, person, organisation, supplier, product, component, stock keeping unit, offering, resource, device, facility, address etc.
- Reference Data contains lists of values used to categorise or classify other data. Examples: country codes, currency codes, type codes, status codes, role codes, customer & product segments, industry classification etc.
- Relationship Data – Reference Data and Core Master Data can be built into different hierarchies which are used in analytics dimensions. Examples: customer classification, organisation, geography, channel, product, calendar etc. hierarchies.
- Transactional or interactional data represent transactions and events performed by parties or resources. They are needed to conduct and audit daily business operations. Examples: orders, shipments, invoices, payments, call centre calls, web visits or machine failures. Typically, facts and KPIs are derived based on transactions and events.
- Analytical Data is derived data (for example KPIs) needed to measure, report, and make decisions about business processes.
The biggest data quality challenges relate to master data because it is often scattered in a lot of siloed data sources. Each data source has its own data definition, validation, and store of the master data. Data quality may be perfectly fit for purpose in its own silo, but when you integrate these silos data quality problems emerge. The main challenge in integrating data silos are the differences in the data definitions and validations, which may relate to core master data attributes like customer, names, addresses etc. or reference data like different type, status, and role codes.
MDM integrates siloed master data and produces consolidated and unified master data. Master data integration includes data certification – data standardisation, cleansing, matching and cross-referencing/de-duplicating to produce unified master data or golden records, which could be for example consolidated customer records.
Golden master data is also used to harmonise transactional/interactional data. That ensures interoperability or the ability to combine data for cross-functional reuse.
Interoperability needs to go beyond corporate boundaries. Data needs to be shared with external partners across the supply chain. It is not enough to tackle master data challenges at the corporate level to achieve that.
2. MDM & ERP consolidation
Companies change continuously. These changes impact their IT landscape – for example, their ERP systems. Well managed master data is a key for achieving interoperability and therefore resiliency for the changes. Changes may occur for example with mergers & acquisitions (M&A). As a result of this kind of change, companies end up having multiple ERP systems. ERP consolidations consume a lot of resources and are a major business for many consulting companies. These programs often have a large scope, risk, and cost.
What if a company has almost reached their ERP consolidation target and the company decides to do yet another M&A. Then the ERP consolidation target moves again further into the horizon. There are companies that do company acquisitions very frequently. How do they manage ERP consolidations?
A data-driven approach with a focus on MDM can help to reduce ERP consolidation costs or even eliminate the need for ERP consolidations. Leading companies that grow via M&A use MDM and semantic standards to connect the acquired new ERPs into the common MDM. This allows them to share the products and customers of the new acquired company through their normal sales channels and rapidly incorporate the new company as a part of the mother company.
Process harmonisation can be another driver for ERP consolidation. If a data-driven approach is used this is achieved through 2 stages: data harmonisation with help of MDM and then process harmonisation by the consolidation of ERPs or parts of them – for example, centralised purchasing and finance only. Breaking the change into 2 stages reduces risks and enables continuous visibility to business while doing the consolidation.
3. MDM & organisation changes
Managing changes to organisational master data without centralised MDM may be a very difficult task to undertake. Frequent organisation changes may become killers for any other development progress because too often companies need to change, reconcile, and manage different distributed versions of the organisation hierarchies or some financial variants of these. In many places, F&C always has to “pull a break” half a year before an organisation change to fix their siloed reporting hierarchies.
Centralising the organisational master data into MDM can help to realise “plug & play” organisation change within management reporting. MDM can help to maintain organisation hierarchies and their versioning and validity over time. You can build a future organisation hierarchy already today and simulate how the organisation, customers and related transactions would be split in the future. This way you test and validate in advance management reporting based on the organisation change.
4. MDM & single version of the truth
Many companies have a customer-centric vision where they want to improve customer experience, services, and grow relationships with their customers. In order to achieve this, they need a single version of the truth of the customer across all touchpoints. It may prove to be a big challenge – not just people and process challenges. That is where MDM solutions are needed to create unified data – standardised, deduplicated & cross-referenced.
5. MDM & Data Certification
Data reuse is dependent on the ability to find, understand, access and trust data. Data must be certified for reuse. Data certification includes data definition, standardisation, interoperability, data quality, security & privacy controls – audit & monitoring. Interoperability ensures that we can combine data for cross-functional reuse. MDM is a key contributor to data certification.
“Data does not manage itself”. Data certification requires that relevant accountabilities are assigned for continuous data management. Successful MDM implementation is business-driven and it includes incremental Data Governance deployment with people & process changes that are needed to nurture the data assets, talent & experience across the organisation. This is an interesting topic, but it will be part of another blog writing.
Want to learn more? Watch our webinar recording Strategic data management: Data Mesh meets MDM.