Much like witches, ghosts, or werewolves, many consider getting to a single source of truth with your data to be a myth. We've heard many people say creating a single source of truth is too challenging, too costly, and not worth the investment. But unlike the aforementioned Halloween folklore, data and a single source of truth can haunt organizations year-round.
What feeds the problem? Multiple platforms, applications, departments, people - and of course data - are fragmented throughout the enterprise.
One of the demons - data silos - continues to lurk throughout large organizations. Harvard Business Review says data silos can make it prohibitively costly to access data and put it to other uses. Data silos are a collection of information in an organization that are isolated and many times inaccessible to other parts of the business. Think of them as castles of treasured information that could help standardize data and provide a comprehensive view of products, people, companies and more. Technology plays a role in that applications used by separate departments may capture information that isn't shared across the organization. The organizational structure and company culture may also inhibit the sharing of information. This can lead to:
To break down data silos and move towards a single source of truth, organizations need a complete data management and data governance program. Collecting, integrating and managing volumes of data coming in from many new sources can create complex systems of truth. By incorporating accurate, consistent and transparent data into enterprise processes, you can complete important tasks faster and confidently understand relationships among entities. Let's look at a couple of ways to tear down data silos and build a single source of truth.
Implement the Correct Software
Legal entities, natural people and products are distinctly different data domains and the data representing each is distinctly different as well. Yet, we continue to see organizations attempt to force different types of data into the same database schema. Data management software that supports distinctive data domains ensures that the required attributes are stored and managed for the type of data your organization tracks.
Additionally, we see cloud-hosted databases as a necessity when breaking down data silos in large enterprises as it makes sharing data in real time much more affordable and achievable. We recommend using rules and workflows to streamline data management actions and improve consistency within your organization. Proper implementation of high quality software is essential to a successful data management program.
Process and Procedure
While people are a valuable resource, they introduce inconsistency and variance which can be detrimental to a data management program. But with proper foresight, organization, and process documentation, you can limit this variability. Think about how your organization will use the data when collecting the data. What fields are important, what information indicates something unique about an entity (e.g., legal entity identifier), or what information provides context for data quality management and data usage about products and people? Document this information and share it throughout your organization to ensure consistent data usage.
As data continues to flow faster and enter the enterprise from many channels, it will be important to use a reliable, yet flexible process to ensure records are updated as close to real time as possible. And because data is always changing, that adds to the challenge. We've only touched upon two examples above, but we're interested to understand how other organizations have approached this.
For example, do the business organizations and executive teams have an enterprise level view of your data? Can you see and understand the detail needed to make confident business decisions? Let's discuss in the comments below and if you enjoyed reading this blog, get the Kingland blog delivered to your inbox.