So far, we’ve considered the rivalry of data fabric versus mesh. We’ve explored what is a data fabric, what is a data mesh, and how a data catalog supports these architectures. We’ve even made the case that this is less a rivalry than a love story in the making.
That’s right: You can leverage a meshy fabric architecture. But how do you get started?
How to Leverage a Meshy Fabric Architecture
This white paper makes this information actionable with a methodology, so you can learn how to implement a meshy fabric with your data catalog. In this blog, we’ll tease three high-level tips and takeaways. For the full story, download the white paper here!
3 Tips in Building a Meshy Fabric Methodology
Tip 1: Start with the people
The good news: You’re moving to a more efficient, transparent process. It will offload pressure from IT, improve your data supply chain, and lead to smarter decision making. The bad news: You’ll need to prepare your people to change their habits and behaviors. This begins with a fundamental shift around how people see data as well as their own relationship to it.
First thing: Data is a powerful asset. Consider it now, too, as a product, which, if used correctly and by the right people, can provide real ROI. So where do people factor in? And how do you communicate this to them?
Start with your domain experts. These are your new data producers. Encourage these folks to think of themselves as shopkeepers. How will they welcome people to great data and guide best-in-class usage? How will they promote satisfaction for their customer base? How will they ensure they improve this system – both the data products they deliver, and how they deliver them – over time?
Data consumers, too, should consider their new role. We’re all consumers in one sense or another. We know what makes for a good shopping experience, from click-to-buy to delivery day and beyond. Educate your new data “shoppers” on how they can be exceptional customers to your new group of shopkeepers or distributors. Communication and kindness are key!
Tip 2: Let the data catalog do the heavy lifting
What data is useful? Odds are, that which is most used by people. A data catalog with a Behavioral Analysis Engine will measure human behavior around data to locate your most valuable and actionable data. This data about data, AKA metadata, is an essential layer of your new meshy fabric.
Once it’s set up, a data catalog will ingest metadata automatically from built-in connectors. Applying a layer of behavioral intelligence enables the catalog to showcase the most valuable data products, which are likely fit for wider consumption and usage.
But how can these new assets be used? Leaders must use the data catalog to establish governance principles that clarify best practices for data usage across the enterprise. Documenting key guidance, like ownership rights and consequences, aligns all people on usage expectations.
Leaders must also build policies about data into the catalog, from both an internal perspective (“How do we define data quality?”) to an external, compliance-focused perspective (“How do we ensure analysts use private data legally?”)
Again, a data catalog can automate the tedium of compliance at scale. For example, once a new rule is created, a feature like Policy Center will propagate that rule to all data of a certain category throughout the organization.
Tip 3: Never, ever skip data governance
Data governance isn’t exactly a trip to Disneyland. For this reason, many data and analytics projects often skip this crucial step. But if your new, shiny AI-powered project is like a trip to Disneyland, think of data governance as the healthy breakfast that fuels your fun. It’s an obligation, sure, but you’ll be thankful late in the day that you laid this crucial foundation.
Similarly, your data catalog is the foundation to a painless data governance program that scales across the business. As you develop that program, a data catalog offers a place to house crucial definitions in its business glossary, as well as a place to define data domains, which enable consumers to quickly find what they seek.
Set up and organize your data domains with input from the team that will use it most. Launch one domain at a time, using your organizational chart as a guide (go by department: sales, marketing, HR, etc).
Be warned! Questions will arise. For example: “How do we define the term customer?” Some words will be relevant to all domains, and must be assigned dual domain ownership. Encourage your domain owners to collaborate and align on key shared definitions all can agree upon.
Learn how to build a meshy fabric
Aligning your people, process, and technology around a new system like a meshy fabric can feel daunting. But with the Alation Data Catalog, you can surface what’s already being done and organically build your meshy fabric around these informal structures.
Keen to get started? This white paper, The Meshy Fabric Methodology, walks you through the 7 key steps in more detail. Download it today.