As companies grapple with how to gain more value from their data, defining, mastering and managing data has become critical. Data that’s essential to doing business can include information such as customer records and product specs. This data may be generated and used by multiple applications across the enterprise, which makes managing - and mastering - it complex.
In a recent survey, almost half of all organizations said that they use at least ten different types of data to do business. To leverage data successfully without feeling overwhelmed, companies must be able to access it and be confident in its quality. That’s where master data management (MDM) is supposed to come in.
MDM rose in prominence as a way for companies to manage their essential data at scale, and across multiple systems and processes. The idea behind MDM is that technology can be used to help companies match, deduplicate, and centralize data entities from their different systems and processes. However, that doesn’t always work in practice. Master data management often falls short of its intended goals, leaving many companies wondering if there’s a better way to approach data management from disparate systems and processes.
Here’s how to evolve beyond traditional master data management (MDM) and take a more proactive approach to data governance.
While master data management was widely touted for its potential to solve major challenges, the reality is, it doesn’t live up to its promise. The main issue is that it’s reactive. Companies are ingesting large amounts of data, without considering the broader context or potential relationships and duplications with other processes.
MDM technology promised to be able to deduplicate, standardize and integrate this data. However, by the time these tools were being considered, it was already too late and the results were less than satisfactory. Also, MDM is a massive undertaking that requires regular engagement. It’s not a one-off project, it’s a constant game of catch-up.
Instead of managing data reactively, organizations need to start by understanding their operational processes and ensuring that their data strategy is integral and complementary. Then, they can incorporate data governance from the start.
For example, with MDM, a company may try to consolidate customer data from multiple systems, such as their CRM, ERP, project management platform, and customer portal. They might end up creating profiles for each customer in each system, and then rely on an MDM tool to clean up the records and consolidate them. However, the issue that created the problem in the first place isn’t solely a technology issue. People and processes play a role. There are no clear lines of responsibility for who is in charge of making customer records. The process of ingesting all data and cleaning it up later is also part of the problem.
To move beyond master data management and actually design for data governance, companies need to consider the full spectrum of factors involved in the data lifecycle.
By considering technology alongside people and processes, you can prevent some of the issues that MDM was supposed to solve reactively. So, a potential solution to the problem outlined above would be to implement a process where one team is responsible for creating every customer record. That record could then live in one system and feed other systems throughout the company.
Making governance a priority enables companies to get more value from their data. The chart below illustrates the data lifecycle for analytics and all the parties involved. The activities on the left (data sourcing, extraction, and loading) are necessary, but have low analytical value. Organizations need to ensure these steps are highly repeatable and take little effort to replicate and scale — this is where a well-thought-out data strategy and sound technical architecture can help.
When organizations define their data before it's ingested, it gets them out of the constant game of catch-up. Think of it like a recipe. You don’t combine a bunch of ingredients, put them in the oven, and then decide that you’ve made a cake after you eat it. Instead, you determine that you want to make a cake, then you combine the right ingredients in the right amounts, and bake it.
Apply this same approach to data. When everyone’s in agreement about the desired “dish,” or data product, and understands which “ingredients,” or data elements, it takes to create it, then you can make sure you capture and define this data properly in the first place. Business and IT need to collaborate to determine what types of data products the company is aiming for, and which ingredients are needed to create it. Then, you can implement processes and technology that support those decisions.
The activities on the right of the graphic (data modeling, training, and serving) do offer the company high value. High-cost resources, such as data scientists, should be devoted to these tasks while spending minimal time and effort on the activities on the left. This is what we at Kenway call shifting left on data governance.
Incorporating data governance earlier in the data lifecycle requires companies to take a step back and examine their entire data strategy. This can be a major undertaking, and it’s crucial to get it right from the start. Kenway has extensive experience helping organizations shift left on data governance and drive their big data transformation.
We understand how to get people, processes, and technology in alignment to ensure that you maximize the value of your data. We’ll help you move beyond master data management and incorporate a solid data governance program, along with a well-architected technical stack that supports the entire data lifecycle.
To get started, schedule a consultation with our experts at [email protected] or click here.
What is the difference between MDM and RDM?
Master data management (MDM) is the process of defining, structuring, and disseminating master data. Reference data management (RDM) is the process of cleaning and structuring data that is used to classify and categorize other data. So, MDM is applied to data like customer records, such as their name and address. RDM is applied to data like postal codes and countries.
What is master data management in CRM?
When applied to CRMs, master data management enables companies to compile customer profiles using data from multiple sources. Instead of having to manually import information throughout the customer journey, companies can leverage master data management to create a centralized hub for all customer information.
