Extracting value from data and fostering a data-driven organization is paramount to keeping pace with the speed of change and meeting customer demands. This article focuses on how business leaders can unite their IT departments and business users with data fabric designs that increase efficiency and create value.
Companies of all sizes, across all industries, are dealing with economic uncertainty. In a recent survey by The Conference Board, 98% of CEOs said they’re preparing for a US recession over the next 12–18 months. For this reason, it’s no surprise that business decision makers are pivoting their strategies and using data-driven decisions to cut costs and increase efficiency. It’s time to work smarter—not harder—and make the most of what you already have.
So how can companies do more with less while maintaining efficient internal operations and meeting consumer demands? The answer starts with data and ends with informed decision making.
Companies need to collect data to know who their customers are and what they need, the status of their suppliers, and where there are weak spots in the supply chain. But beyond collecting data, it’s paramount that companies extract the value from their data to make better and faster business decisions. This is where data fabrics—an emerging data management design that allows companies to securely and seamlessly access, integrate, model, analyze, and provision data—come into play. Implementing an effective data fabric increases an organization’s ability to harness its data to create business value.
In my role as Vice President, Product Management at Tableau, I lead teams that are responsible for the tools and services that our customers rely on to connect, manage, analyze, and govern their data. In this article, I’ll discuss the business benefits of uniting your tools to design a data fabric and how you can start building one at your organization.
The Current Data Challenges
If your organization is anything like those of our customers, you’re currently facing two major challenges: 1) the huge—and growing—amount of data and 2) an inability to access the data you need.
Statista estimates that the world produced 79 zettabytes of data in 2021—meaning that in every hour that passed that year, the world made more bytes of data than grains of sand on earth. Only a few years ago, most companies stored their data on-premises or in a single cloud. Today, as the volume of data continues to grow, companies are moving to multi-cloud environments to spread their risk and take advantage of different clouds’ capabilities. When data is distributed across multiple clouds—each with its own governance structure, catalog, and user and identity management systems—and stored in on-premises data sites and individual desktops, it’s challenging to govern, manage, and interpret. Without organized and accessible data, employees don’t have the information they need to make critical decisions.
According to research by Forrester Consulting, commissioned by Tableau, 70% of employees will be expected to use data heavily by 2025, up from 40% in 20181. As more employees rely on using data in their day-to-day roles, organizations must find ways to make data and insights easily accessible.
The challenge isn’t collecting enough data. It’s having robust analytics tools and infrastructure in place so individuals can get the right data at the right time and act on real-time insights.
How Data Fabrics Cut Costs And Increase Efficiency
A data fabric creates a single, interoperable environment that connects all the different data and analytics elements in an organization’s ecosystem. Unlike data integration—a one-dimensional element of a data fabric focused on getting users access to a unified view of their data—governance, cataloging, discoverability, automation, and analysis are the remaining key ingredients needed for a data fabric.
Data fabric designs optimize internal operations without a big, upfront financial investment. Here’s how they can help you cut costs while boosting efficiency:
1. Data fabrics reduce duplication. By design, a data fabric ensures that the same data isn’t stored in multiple places and multiple systems aren’t working to achieve the same task.
For example, to share a project update, you wouldn’t send individual messages to each stakeholder. Such repetition is time-consuming, clutters communication channels, and invites manual errors. You’d use a shared project management software to communicate once to everyone. Similarly, companies don’t want five analytics solutions—they want a single solution that fits within their existing ecosystem. Instead of four different catalogs, they want one catalog that enables them to discover and understand data. And when gaps and security issues arise, companies don’t want unique governance protocols for each solution—they want one that spans all of them.
2. Data fabrics connect every part of your data and analytics ecosystem. If your data resides on-premises, but your analytics solution is in the cloud, hauling all your data into the cloud to conduct analysis is inefficient. Instead, tighten your data fabric using virtualization services that allow you to conduct analysis wherever your data lives. Virtualization services like Tableau Bridge enable you to complete your analysis and pull insights without moving data back and forth between catalogs in different clouds and on-premises instances.
3. Data fabrics empower employees by enabling data-driven decision making. When data and analytics services are interoperable and data duplication is removed, the fabric becomes tighter and tighter, making it easier to manage and extract value. A functioning data fabric that includes a strong analytics tool enables employees to make more informed decisions. Instead of spending time sorting and searching through rows of data and evaluating which data is relevant to their job function, a data fabric makes it easy to find the right data at the right time. Everyone in the organization can get to insights faster and make higher-quality decisions.
How To Kickstart Your Data Fabric Journey
A data fabric is not a single product or piece of technology. It’s a design system that knits together all parts of your data and analytics ecosystem. When beginning your data fabric journey, you don’t need to purchase any new data or analytics tools. Start by organizing what you have and, as the budget allows, purchase new tools to bolster your capabilities. From there, you’ll need to invest in people and change management to ensure successful rollout and adoption. The value of a data fabric will only materialize when employees are data literate and feel empowered to make decisions with data.
Data Fabric Design Solutions
If the bulk of your data solutions and products are from a single provider, explore if they have a data fabric offering that complements your existing ecosystem. Salesforce, for example, recently announced Salesforce Genie Customer Data Cloud, powered by Tableau, which integrates data across all of the Salesforce clouds. While offerings like these may enhance your data fabric, they are not necessary when you begin building. Start small and grow.
Roll Out In Phases
Before an organization-wide data fabric rollout, begin with a subset—a team, department, or org. A focused starting point enables you to figure out what works, what doesn’t, and what’s necessary to scale.
Prioritize selecting the right department to set the tone for the rollouts that follow. Choose a department that has 1) a lot to gain from a functioning data fabric and 2) a business owner invested in building a better data and analytics strategy. You’ll also need a data architect and data steward to understand your data architecture and how to manage it in the long term, as well as engineering teams committed to delivering against the ask.
ETCIO explains that data fabrics result in data expertise that “expands beyond the IT department to business groups, internal operations, and customer relationships.” For this reason, it’s critical you determine what stakeholders need to get out of the data fabric and use that as your north star while architecting the design. Everyone should be aligned on what you’re trying to achieve and how the new data fabric will better the business.
Define The Metadata
Once your team and vision are in place, it’s time to explore your metadata. Consider your data and the tools you use. What are your data residency, data storage, data management, and analytics needs? What metadata requirements will provide the information needed to solve your business problems? A high-functioning data fabric is a sign that a company knows how to manage metadata and semantics end-to-end and that each tool can build on top of what the previous tool understood from the data.
Some organizations want to have a single data definition for their entire company. And there are organizations that want three definitions just within the sales department—they might have one for outgoing sales, one for incoming, and a third for financial models. Don’t ask how many definitions you should have. Instead, ask yourself: Are we using the right definition for this specific concept? A single definition may span teams and departments, and others won’t. If you struggle to align on metadata requirements, it may be a sign you need to start with a smaller subset.
Optimize With Machine Learning
After the metadata requirements are outlined, your engineering team can begin stitching together your tools to create the data fabric. And you can continue to optimize your data fabric over time with machine learning (ML). An effective metadata semantic layer that not only looks at the data but also layers on added context magnifies the power of ML. When organizations consistently manage data across all the components of their data fabric, they can use ML to understand how different users and groups interact with it. These insights, such as how often data needs to be refreshed or whether analysis must be computed locally versus in the cloud, enable you to optimize for cost and work smarter, not harder.
There’s no one right way to do this. Data fabrics are like a fingerprint. All data fabrics unify data and analytics solutions, but each data fabric is composed of different data storage and service providers unique to your business needs.
Keeping Pace With The Speed Of Change
Building a data fabric will allow you to establish more efficient operations without making a huge financial investment. It’s not expensive from a purely monetary perspective, but it will require good change management, and the value won’t materialize immediately. Start small and connect elements with shared metadata that are easy to integrate. The more you add to the data fabric, the easier it will be to grow because the data within it will support the expansion.
With a data fabric, you can reduce the time it takes to manage your data, increase collaboration, and boost the adoption of self-service analytics so everyone in your organization can make data-driven decisions. Your ability to stay competitive and keep pace with the speed of change is only as high as your ability to maintain a data-driven organization.
ACCELERATE TIME TO VALUE WITH DATA FABRICS
Learn more about how data fabric designs can help you drive down data warehousing costs, increase time to insight, and create amazing customer experiences.
– Read this ebook Make the most of your data fabric with Tableau