Making sense of application data management

Business solutions start with clear definitions of data

co-workers using a laptop and having a discussion
Thinkstock

With all the information out there—2.5 quintillion bytes a day, by one count—it’s no surprise that today’s businesses struggle with classifying, organizing and governing data. Whether they need the data or just end up with it (digital exhaust), they must have it handy. Shrewd data management is the basis for turning information into revenue.

Recently, businesses have been retooling their data management strategy by focusing on the larger architecture of the data hub. The data hub connects all data in an enterprise, ultimately giving all business users the 360-degree view of the data they need to do their job. Ideally, this would happen in the context of the business applications they already use; making this transparent and efficient, while enabling data stewardship on a collaborative basis across the enterprise.

In my last column, I wrote about taking the data hub one step further to make it intelligent. This time I want to do a deeper dive on a critical component of the data hub: application data management (ADM).

Defining and mastering application data management

As analyst and research VP Andrew White at Gartner has pointed out, ADM is a sort of new subfield that exists both alongside and within master data management (MDM). Application data management (ADM) masters data that is shared (common) among multiple applications, but not necessarily the entire enterprise.

For instance, a typical business today might have supply chain management, a customer relation management (CRM) system, and billing software. Each system runs a different part of the business. Yet all these systems have data that is common across them, such as customer names, addresses, billing and shipping addresses, and invoices.

Each system also has other data. In the supply chain system, there is logistics information, drop shipping details, taxes, and duties. The CRM has leads and opportunities, additional contacts, past orders, and negotiations, while accounting software contains bank account and routing numbers—information that needs high security, to be seen only by few staff members in the entire organization.

The common data is different. It’s what’s often referred to as “slowly changing dimensions.” Over the course of your life, very slowly, your address, phone, and email change, but you’re still the same person. The same thing is true if you work for one company but get promoted or transfer offices; some numbers and letters attributed to you will change, but others won’t.

Information that slowly changes dimensions is considered master data and is kept in a separate database with information about these small, slow changes over time. The more rapidly changing application data is transactional—information like a person’s income or a business’s revenue. It changes all the time (like every quarter) and is kept alongside customer information. Though it’s not master data, a business still wants to master it.

Application data management in practice

Throughout a business day, various individuals in an organization will update these groups of information. Depending on their role and permissions, they may update, or approve or submit for approval to a data steward bit parts of application data. They’ll update at different speeds, with different levels of specificity and accuracy. As the changes are enacted, the shared data is immediately reflected across all applications. So, ADM does everything MDM does, but ultimately serves a different case: shared across multiple applications.

What connects everything together? That’s the data hub. The data hub includes data governance, data quality and enrichment, as well as workflows (such as approvals, and iterative processes) They reflect how data changes over time and bring crystal clarity for traceability, lineage, and audibility.

Artificial intelligence: the key component

Until recently, the ability to use a data hub strategy has been hindered by the encumbering need of integration and the requirement to cobble together multiple software platforms and services to a functional system. Artificial intelligence and machine learning technologies bring the “last mile” of automation and correlation to make the data hub feasible.

This final layer is the “intelligent” data hub—which contemplates the above-referenced data capabilities, including AI and machine learning which leads to an intuitive business user-friendly interface that makes data processes easily consumable for any staff member in the organization.

Business end users are the ones that must ultimately be empowered to build customer loyalty and explore cross-sell and upsell opportunities. Data can help them, but only if it’s stored in the right place and stewarded from the right application to the right person at the right time.

Bringing it together

The data industry has done itself a disservice by having many componentized pieces of software for segmented parts of the larger requirement. This was born out of a desire to own a niche within a crowded market. Increasingly, the way to deliver the value so desperately needed is by bringing it together in a single platform and streamlining complexity with an intuitive design. Watch this space.

This article is published as part of the IDG Contributor Network. Want to Join?