How Telcos Gain Faster Data Insights with Data Mesh and Data Fabric

How data mesh and data fabric can enable faster data insights for telcos.

Last Updated: November 15, 2022

Telecommunications is probably the most data-rich industry in the world. Telco providers gather data about call details, mobile phone minutes, network equipment, server logs, billing, social networks, and more. Vinay Samuel, founder and CEO of Zetaris, examines how data mesh and data fabric can help telcos gain relevant data insights faster.

While telcos are rich in data, they are poor in data insight because the vast volumes of data are scattered across business units and departments, from product development to network management. The challenge isn’t in obtaining the data. It’s making that data accessible to gain customer insights quickly and cost-effectively.

Telecommunications data is typically stored across multiple systems, such as billing systems, product management platforms, and data warehouses. Data also resides in critical applications that support everything from customer service to network availability. The only way to gain a single, accurate view of the customer and customer experience is by aggregating all those data sources. Telecom companies continue to struggle to find the best strategy to break down data silos.

Using ETL for Data Access

A major obstacle to frictionless data access is the flawed assumption that it must be stored in one central location before you can analyze the data. Creating an additional data warehouse or repository for analytics is complex and costly, especially when you must access data from multiple sources. The data must be extracted, transformed for use in a single database, and loaded into the database (ETL).

Even the best ETL tools and strategies fail to deliver what telcos really want – a single, cohesive view of their data. Implementing an ETL strategy is costly and complex. For example, adding data sets that may have been omitted can result in expensive projects that take so long that the data is no longer needed or out of date by the time the project is completed.

Data Mesh Brings a Decentralized Approach

To reduce data warehousing costs and improve data access, telcos are embracing a new approach that is proving to be a game changer for analytics. Rather than relying on a central data repository, telcos are adopting a data mesh and data fabric approach.

Data mesh is an approach to data access, not a technology. Data mesh uses a decentralized data architecture that shifts responsibility for analytical data from a central data location to domain teams via a domain-agnostic data platform. Rather than moving data around the enterprise, business experts determine the best way to share data, including the data sources, and provide access in a timely and secure manner.

For example, one data owner could determine that they access the data they need through a secure API. The owner of a separate system could provide data access by extracting that data to a data warehouse or other data repository. With data mesh, each data owner chooses the means to share data and imposes business rules, such as access control.

See More: Why You Need a Data Fabric for Effective AI

Data Fabric Controls Access

The data mesh provides data access using a data fabric. Data fabric is a design concept that provides an integrated layer or fabric of data and connecting processes. The data fabric enables data owners to impose rules for access through the data mesh. Using metadata to identify data sources, the data fabric dictates how data from different data in different sources across the enterprise can be joined and queried in real time. Software to turn the data fabric into a decentralized analytical data platform (and query engine) is called the networked data platform (from Zetaris).

For example, when a data scientist is analyzing service representative call behavior, they may need to compare data from different disparate systems to build a data product that describes the representative’s performance. Comparing this data in different places using a traditional ETL (or centralization) process will be slow, costly and error-prone. So, the data scientist may use the data mesh leveraging the Networked Data Platform to connect to and cross-query the data in different places. The answer to this query can then be displayed with all the information relevant to that customer and representative in the data scientist’s tool of choice – such as PowerBI or Tableau. That data could include network usage, billing history, contact information, and more.

Rather than using an ETL model to load the data into a single system, the Networked Data Platform (from Zetaris) software works with the data fabric to initiate a query and organize the necessary data into a coherent view. This approach is much faster than ETL, is far less complex, and is more cost-effective.

The data mesh/networked data platform model can work with existing analytics tools. Any software application sees the data fabric as a regular database when in fact, it is a virtualized data warehouse (a data warehouse that contains only rules about the data that exists elsewhere across the network). And unlike the ETL data model, the data fabric makes it easy to add new data sources in hours or days, so data is still timely and relevant.

Telecommunications have become more complex, with new service offerings resulting in even more data. Telcos need a new approach that eschews the ETL process to make that data useful. Data mesh and data fabric can deliver faster customer insights, unlock competitive advantages, and enable innovation and digital transformation at lower costs. Every aspect of telecommunications operations, from network planning to customer service, can benefit from faster access to information.

Are you leveraging data mesh and data fabric for faster data insights? Tell us on FacebookOpens a new window , TwitterOpens a new window , and LinkedInOpens a new window .

MORE ON TELCO: 

Image Source: Shutterstock

Vinay Samuel
Vinay is a forward-thinking entrepreneur and a pioneer in "event-based analytics" and parallel database technologies. Vinay's more than 20 years of experience has been cultivated through developing leading Big Data and information management companies. His vast knowledge of business management, marketing, data warehousing, and big data has been a driving factor in his success in utilizing data-driven strategies to produce sustainable solutions for worldwide companies.
Take me to Community
Do you still have questions? Head over to the Spiceworks Community to find answers.