Click to learn more about author Farnaz Erfan.
Data Profiling has traditionally been regarded as an IT tool. This is because Data Profiling tools were in the purview of technical people when first introduced, and because data was primarily sourced from IT-implemented ERP or CRM systems.
Today things have changed. Upwards of 80 percent of enterprise data is unstructured. This is data from external sources, from customers’ digital footprints, second- or third-party data, or public sources. Additionally, and business users are now closest to the data’s context, not IT.
Consequently, new, modern, interactive Data Profiling is surfacing as a reliable guide. Similar to the North Star pointing towards the right direction, interactive Data Profiling is providing emerging business opportunities which include:
Data Commercialization / Information-as-a-Service
Companies are sitting on massive amounts of valuable, but underutilized, data and searching for opportunities to capitalize on it. Conditions for data commercialization are ripe: enormous data volumes, decreased storage costs, enhanced user experience principals, and improved data and analytics tools. Monetizing data assets involves integrating outside data from adjacent industries, new partners, or public sources with internal information to develop rich information products or services.
According to Accenture’s Data Monetization framework, the value of data grows as its correlation with other insights increases. Therefore, many companies are using Data Profiling tools to onboard and add second- and third-party data to their first-party information in order to mix, enrich and create context.
Interactive Data Profiling removes some of the most time-consuming part of data-intensive projects by providing an intuitive, visual and interactive application for business users to onboard, profile, and create quality information.
This is important because second- and third-party data sources are typically dark data with unknown values and often come in varied forms, whereby some data files may be slightly -or drastically- different from others. The interactive nature of Data Profiling tools provides visibility to all data values, regardless of its original shape. These tools flatten and harmonize the data structure into a tabular, Excel-like format, enabling business consumers to access, investigate, and inspect data content before determining how to enrich, improve, and commercialize it.
IoT Product Testing and Collecting Market Intelligence
Although IoT device testing covers a broad spectrum, ranging from usability testing, performance testing, device compatibility testing, and more, one highly data-centric area is functional testing. This normally entails function validation, device-user interactions, operability (i.e. start/stop/restart/interrupt), and error handling. Additionally, to analyze usage, customer experience, and adoption, many manufacturers conduct field tests with actual users to collect usage data.
Both use cases – functional testing pre-launch and product usage testing post-launch – require quick access to device data. However, the process of leveraging IT and combining Excel and SQL databases to inspect device data is much too lengthy.
Interactive Data Profiling can help here because IoT data is often created in a semi-structured, flat file format carrying very rudimentary information such as a timestamp, latitude, longitude, temperature, pressure, connected/disconnected, on/off, or paused mode.
To augment the context and generate real business insights, an analyst often creates derivatives of the data. For example, in a connected car scenario, an analyst might examine things such as braking frequency to predict scheduled maintenance, or variety of speed and number of stops to associate data with driving habits.
An interactive Data Profiling tool allows analysts to examine all data values, hypothesizing exceptions, and discovering patterns of interesting behavior. And since testing an IoT product pre-launch or assessing its adoption post-launch are both time-sensitive, eliminating any IT backlogs directly impacts revenue.
Capitalizing on Mergers and Acquisitions
Two macro factors that are weighing heavily in today’s market are low growth in mature economies and cheap money. With few exceptions, organic growth is hard to come by, so shareholders attempt to grow inorganically via mergers and acquisitions (M&A), which are easily and inexpensively financed at current borrowing rates.
This is commonplace in the technology sector, where countless high-tech companies have historically consolidated. Today, M&A is a fundamental instrument utilized to boost innovation, streamline operations, and elevate customer experiences. This practice has expanded across industries, as demonstrated in acquisitions of Yahoo and AOL by Verizon, Whole Foods by Amazon, and most recently, the CVS-Aetna merger.
With most mergers, financial and supply chain systems, customer information databases, and employee/HR records were often constructed decades before, so tribal knowledge of the data content and how it is structured is lost. Moreover, the heightened economic pressures instigating these mergers necessitate the acceleration of data consolidation and migration processes. In many cases, line of business employees responsible for data operations require access to the new, merged data to drive business objectives and manage daily operations.
Data consolidation and data migration requires knowledge of the data’s current state to properly design the new system. As business people best understand relationships within the data, they are ideally suited to decide what should or should not be migrated or archived. An interactive, visual, and self-service Data Profiling solution can help them immensely with this task, rather than leaving it to IT specialists who invariably view data relationships simply in terms of tables in a database, resulting in numerous iterations and unnecessary delays.
Why Modern Environments Require New Interactive Profiling Capabilities
Because data is generated and consumed primarily by business people, Data Profiling is now rightfully a major consideration for many line of business owners. Today business users and Data Analysts across all markets and functional domains are looking to effectively and efficiently understand the vast and disparate range of their corporate, customer and business information.
While traditional Data Profiling tools previously hindered data access to business users, interactive Data Profiling tools empower business people to use modern techniques such as self-service, built-in Machine Learning, and one-click profiling to identify new opportunities with the data and to change the direction, strategy, and outcomes for their companies.