MDM without data quality? A recipe for failure.
For more data quality info: Subscribe to our newsletter on Linkedin
Increasing volume, speed, and types of data: the amount of data is ever multiplying, and with it, pressure for personalisation and compliance. As well, to maintain control of data, businesses are turning to Master Data Management to break open silos and automate governance processes. And this is not just an outside concern when you consider that almost 50% of organisations manage their reference data separately, in 11 systems or more (BI-Insider). In some cases, data cleansing and preparation takes up more than 60% of the time devoted to data science (Datamation). However, MDM projects can create more problems than they solve if they rely on flawed data. To avoid this and make MDM a true tool for supporting performance, data quality has to be considered in the design of the MDM repository, then throughout its entire life cycle.
1. MDM, DQM: what they mean
When a company launches a data governance project, two areas are at play: Master Data Management (MDM) and Data Quality Management (DQM). These two elements, which complement one another perfectly, play specific roles.
MDM: giving data structure to foster sharing
MDM provides the foundation for effective governance by structuring data and organising data’s access, traceability, security, and its update over time. In order to do this, MDM implements an essential connection point: a unique repository of reference data – the UCR. This repository centralises data that is stable and shared by several services, including customer information, products, suppliers, and employees. It allows the entire organisation to speak the same language, regardless of the system being used (CRM, ERP, e-commerce, BI).
DQM: making data reliable in order to use it better
Alongside this, Data Quality Management takes care of making data reliable, making sure data is exact, complete, that there are no duplicates, and that the data is up to date. DQM operates both upstream (at input, in batch mode) and in continuous mode (automatic cleansing, scoring, and providing indicators), to ensure that data is usable at any given time. By providing automatic data qualification, DQM avoids feeding incorrect information to the MDM, that could then distort results.
Two tools to use together
A well-structured repository has no value if the data it contains is unusable.
Conversely, clearing up silos without an overall structure limits how far the efforts can reach. This is why MDM and DQM have to be combined so that data can flow, support decision-making, and generate value at every level of the company.
- Almost 50% of organisations manage their reference information separately, in 11 or more systems (BI-Insider).
- +60%: extra time spent on data science involving cleansing and preparation of data (Datamation).
- 60%: the portion of companies that report that incorrect data brings them significant additional costs every year (WifiTalents).
- 95% of the interactions with customers could be handled by sophisticated chatbots over the next ten years (Templeton-Recruitment).
- Up to a 20% improvement in data quality in companies that deploy MDM solutions - an essential factor for guaranteeing well-informed decision-making (Gartner).
- Up to a 10% reduction in operational costs for companies that use MDM through the increased precision and effectiveness of data (McKinsey) and up to a 40% reduction in costs related to data management (Worldmetrics).
- 87% of companies report a reduction in data-related errors after the MDM is put in place (Worldmetrics).
- +70%: the number of companies in the financial sector who have or will put in place software for managing reference data, a sign that MDM is essential for guaranteeing the exactness of financial data, as well as regulatory compliance (Deloitte).
2. How data quality truly changes an MDM project
In an MDM project, data quality works at two key levels of the data life cycle: prior to integration into the UCR, and throughout its entire use cycle.
Upstream: making incoming data reliable
Providing an MDM with clean data involves cleansing and remediation of data from different pre-existing databases, and real-time qualification for newly-collected data. Data quality takes care of:
- Removing duplicate records from different systems (including CRM, ERP, and e-commerce) to avoid duplicates from the start,
- Validating key information (including email addresses, telephone numbers, and postal addresses) and standardising them according to consistent reference formats,
- Completing fields that are incomplete, enriching essential fields, detecting inconsistencies.
This step ensures that the data entering the repository is reliable, clean, and ready for use. Without this, the MDM only amplifies errors.
In real time: maintaining quality on all collection channels
Once the data is centralised in the MDM, its quality needs to be maintained over the long term as information is updated, added, or merged. Data quality has an approach that includes the automatic processing that is needed:
- automated quality control, whether in real-time or in batch mode,
- business-line rules that are shared with MDM governance,
- indicators of completeness, uniqueness, and validity, integrated into the steering process,
- automatic alerts or corrections should errors be detected.
Through this continuous approach, the UCR remains dynamic, up to date, and ready for use. Instead of just being a method of storage, it becomes a true source for team confidence – and the single correct reference for customer information.
3. How the MDM and DQM tandem benefits operations
The benefits provided by properly cleansed and managed data contribute to all company business lines. This includes specific use cases in certain sectors.
Business Line Focus: Data, CRM, Marketing, IT
- Data & IT teams
The MDM + DQM tandem provides data managers with consistency, reliability, and interoperability. Gone are the silos and the ambiguous choices between divergent sources. Data becomes traceable and standardised, with governance ensured, which fosters the creation of reliable data pipelines, effective access management, and GDPR compliance.
- Marketing & CRM
Marketing and customer relation teams can now confidently rely on a 360° customer view that is consolidated and ready to use. With duplicates eliminated and contact data validated, they improve targeting, deliverability, and personalisation of their campaigns, across all channels. This provides higher conversion rates and customer relations that are consistent and enriched with the right information.
- Sales management and customer service
Having reliable customer data makes it easier to determine opportunities, set priorities on the actions to take, and to avoid friction points (incomplete records, errors in contact information, etc.). This saves time, builds trust, and improves performance.
Sector Focus: Retail, Banks, Industry
- Retail & e-commerce
In an omnichannel context, having a unique customer repository and a reliable customer history is essential. Recognising a customer across all touch points is no longer guesswork. The customer experience can at last be personalised, whether it be in the shop, on the web, or through a mobile application.
- Banks & Telecommunications
Since the banking and telecommunications sectors are constrained by regulations and complex offers, they can benefit from the combination of MDM and DQM on many levels. Data quality and governance helps companies bring together verified customer data from multiple systems (KYC, CRM, composite offers, etc.), ensure compliance, and produce accurate reporting.
- Industry & supply chains
In industry, management of products, suppliers, or contracts relies on solid data that is free of errors or duplicates. MDM ensures structure, while DQM guarantees the precision of data that is critical to production, logistics, or contract monitoring, between different systems.
In a context where business systems are drowning in data, where personalisation depends on multiple sources, and where regulation is increasingly restrictive, automating data governance is no longer optional. The data has to be reliable, consistent, and usable. This is why integrating data quality into MDM is critical for shoring up relevant use of data in the long term. This is even more important when considering that generative AI, advanced analytics engines, and omnichannel platforms are going to rely more and more on the UCR. Their performance will depend on the quality of the data being processed.
About DQE
Because data quality is essential to customer knowledge and the construction of a lasting relationship, since 2008, DQE has provided its clients with innovative and comprehensive solutions that facilitate the collection of reliable data.
17
Years of
expertise
800
Clients in all
sectors
3Bn
Queries per
year
