Data Fabrics and the Importance of Data Verification and Enrichment Solutions
Scott Taylor and Robert Dickson discuss how data fabrics work to weave all data sources into one location.
Blog: Modern Data Architectures and the Age-Old Role of Location Data
Although data fabrics have been in the technology sector’s vernacular for some years now, they are finally receiving increased focus as a critical approach to managing data to fuel innovations in technology that underpin digital transformation. Industry-analyst firm Forrester defines data fabrics as “a unified, integrated, and intelligent end-to-end data platform to support new and emerging use cases.”
On an episode of Ask the Experts, my co-cost Robert Dickson, Loqate’s VP of Professional Services, and I explored the topic of data fabrics within the changing data architecture landscape and the consistent role of location information with Jessie Snyder, Program Director of Product Management at IBM Data Integration.
“We’re seeing a massive proliferation of data,” Jessie told us. “A lot of organizations have fracturing within their different business units, and there are many new personas trying to access data. The data fabric helps organizations operate more effectively with distributed data sources.”
In fact, IBM was recently confirmed as a leader in the Forrester Wave: Enterprise Data Fabric Scorecard. The Analyst firm observed that IBM is “in a strong strategic position, with a class-leading partner ecosystem and a demonstrated commitment to investing in data fabric.”
There’s no one-size-fits-all data fabric model to suit every business, but the benefits are clear:
An agile and flexible structure
Enhanced data visibility and visualization
Improved access to data across all entities
Better all-round data governance
Data fabrics help standardize data management practices in the cloud, on-premises, and for edge devices. As we move into an ever-digitized business, organizations are looking for newer, more sophisticated and streamlined ways of managing their data. The defining factor of a data fabric is its ability to weave together all data sources into one, easy-to-access location. See what I did there? Weave - that’s the fabric part!
THE location for location data
To stand at the top of the data-rich world, businesses must ensure all their analytical data flows effortlessly throughout the entire enterprise, creating an agile, insight-driven organization that supports digital-first consumers.
Many enterprises use address data, a vital component of a customer’s digital identity, to gather consumer insights and develop strategies. However, doing so without an added layer of verification in their data fabric allows out-of-date, mistyped, duplicate, or incorrectly allocated metadata to obstruct accurate analyses, leading to loss of revenue or worse, customer trust.
“With location data, the sky is the limit in terms of use cases with a data fabric,” Jessie enthused. “The IBM Cloud Pak has our address verification capability built right into it. That helps enterprises assess and remediate quality problems, especially as it pertains to addresses. The address verification portfolio is an incredibly important part of what we’re doing at IBM today. We embed address verification and data quality steps right into the process so data users can’t avoid them.” Establishing a foundational architecture to ensure that data is trustworthy reduces manual efforts and frustration for data consumers down the line.
Spending quality time with data quality
Historically, executive stakeholders have not paid enough attention to data quality, but that seems to be changing. “The light bulb is starting to turn on,” said Jessie. “We’re seeing data and analytics leaders start to focus more on data quality. When we start to bring artificial intelligence into the picture, a lot of organizations are realizing that they don’t have the quality data they need. Having the right addresses as part of your customer data set is just so incredibly important because often it’s the only record that you have.”
Getting the business more involved is a critical step toward data-driven success in any organization. New personas are popping up on the business side. Data analysts, while focusing on traditional use cases like reporting and dashboarding, are expanding their scope. “A new kind of citizen data integrator is trying to do all sorts of interesting things with data,” said Jessie.
“I’ve seen the data-driven conversation elevate to be more business and solution-focused,” Robert added. “I’ve also seen continued recognition that high-quality verified, enhanced and governed data is critical to delivering trusted results.”
It always helps when you can tie everything back to the outcomes. Data quality, especially for location information, is critical for organizations trying to achieve, for example, a comprehensive customer 360 view. Finding strategic use cases for location data seems as difficult as looking for grains of sand on the beach – they are literally everywhere.
For example, in the healthcare space, quality patient records are often dependent on an accurate address. “We are seeing new techniques in financial services to combat fraud,” said Jessie. “We see applications for location data in artificial intelligence and driving machine learning models. Anywhere you want a quality output, we have to ensure quality data, especially quality addresses, are part of the input.” To rework the old GIGO cliché – goodness in will increase the likelihood of goodness out.
The WRONG 80/20 rule!
Great data upfront can have significant downstream effects on improving efficiency. “At IBM, we did a study to determine where analysts spend the majority of their time,” continued Jessie. “We found that the data preparation steps, getting data ready for doing, whether it’s analysis, data science or, the normal kind of BI, consumes about 80% of the time of data and analytics journey.” That leaves only 20% of the time left to create the value the business is looking for. “It’s often the data engineer or data quality specialist who will be involved in remediating poor addresses, verifying those addresses, and correcting those addresses. So, it’s just so important that we focus on the quality as early in the process as possible,” concluded Jessie.
Data Fabric for everyone
Speaking of clichés, many enterprises are suffering from classic data siloes. “It is tough to get a comprehensive picture of your data landscape,” explained Jessie. “Organizations are struggling to operate across a distributed data landscape and want different workloads deployed in different places. A lot of mission-critical data is currently locked in on-prem transactional or operational systems.” CIOs and CDOs realize the strategic nature of looking at their organization’s data holistically. Although this is driving a shift to the cloud, it is getting more complex,” says Jessie. “A lot of organizations are not removing some of the dead weight they had before. So, we’re also talking about how we solve those problems with architectural concepts, like the data fabric, and want to bring in things like address verification to help remediate that data quality.”
Gartner claims that by 2024 “data fabric deployments will quadruple efficiency in data utilization while cutting human-driven data management tasks in half.” Uncovering those types of results, stitched with the integration of high-quality, verified location information, suggests data fabrics are tailor-made for today’s results-oriented enterprises.