Data Centric: When Data Becomes the Business of Every Business

Data Centric: When Data Becomes the Business of Every Business

Enterprises globally are evolving. Much of it has got to do with data. Already, some of the world’s big Internet and IT companies like Amazon, Google, and Microsoft have become data-centric businesses. Such a business model has data at the core.

At the start of 2019, data-centricity was on the list of the top 5 trends to watch out for in the world of business. We are in the last quarter of the year, and that observation looks justified.

In the last decade or so, with the advancement in technology, we saw a movement in the corporate world that started from “Big Data” and then moved on to the deployment of data analytics. Simply collecting huge tracts of data was not enough; it was the means to an end. Tools, including applications, were built around organizational data to help companies do business more scientifically.

We are now in the next phase of the process – transforming an organization into a data-centric one. What does that mean?

Data-centricity Can Be Explained, thus:

In this way of doing business, data is considered an asset as tangible as a company’s hardware or its headquarters building. It is at the heart of the Enterprise’s operations; in fact, the entire IT and business architecture is built keeping in mind the fact that data is a prime and permanent asset.

A simple, non-technical explanation of data-centricity can be – moving away from an application-centric to a data-centric way of doing business.

So far, we’ve seen that those businesses that decided to deploy data analytics worked in this manner: data was extracted from the business practices and fed into a data warehouse or data lakes. Data analytics was a result of a business process using an application built with that specific functionality. Which meant the application owned the data.

Data-Centric Example: For example, in this way of doing business, individual sections buy products or services to respond to a certain commerce need. In all this, however, the potential value represented by that data was rarely inputted back into the Enterprise’s operations. In a nutshell, data analysis was being done in silos, something that almost all Enterprises that have deployed data analytics models will agree on. In an application-centric approach, each app comes with its data model, storing only the data it needs. As the years go by, each business unit develops a segmented vision of its own data, not shared through the rest of the business.

In such a data-driven organization, data analysis flows from the top down. In a data-centric design, however, data analysis flows in the exact opposite manner but feeds into operations. Data is no longer being used to extract answers for daily business questions, but it is the data that’s feeding the units the information needed to run the business more efficiently.

But there’s more for an Enterprise to become truly data-centric. Its human resources, processes, and technologies must be geared towards firmly understanding that data is at the heart of the Enterprise, and must be used collaboratively to advance the business goals.

What Is a Great Advantage To Be Derived From Being Data-Centric?

An Enterprise can accrue many benefits by becoming data-centric:

• First and foremost, a data-centric model gives a massive financial lead to Enterprises adopting this approach, starting with savings on infrastructure and reduction in other recurring costs
• Makes the entire Enterprise “data-smart”, and not just a few team leaders
• Can be used to digitally disrupt the market a business is in
• No more data silos within the organization

How Can My Enterprise Become Data-Centric?

As a starting point, it helps if your organization is already a data-driven one because at least some business units are familiar with the positives of data analytics. It’s about making organizations start walking on the path where all operation and analytical processes are based on ALL available data, and without having the prior need to know exactly what questions they are going to ask.

But the transformation from data-driven to data-centric can’t happen overnight.

Any Enterprise wishing to become a data-centric one, first and foremost, needs to invest in:

1. Its people, which also includes the company culture
2. Its infrastructure
3. Its processes

The investment has to be of mind, matter and money”.

A crucial first step in this transformation is for IT and business divisions to come together and brainstorm on several issues, starting with a postmortem of the prevailing in-house technology.

It is evident that once the business goals are defined, and the future growth mapped out, the technology required to become a data-centric Enterprise needs to be identified. In this process, they will face this question – do I need to re-invent my entire tech stack?

As the huge amounts of data start flowing in, it becomes a challenge for Enterprises to store, retrieve and process it using the traditional application architectures. For a data-centric business, what is needed is an agile, data-centric architecture that responds to constant change.

There’s obviously no single system available that can address the large and ever-increasing scale of data and its computation. But there are new infrastructure and application concepts, largely based on artificial intelligence, emerging to address these problems of data-centric computing.

Here’s what the biggies have done. Microsoft Azure or Amazon Web Services, for example, use distributed architectures that work in tandem. Each also retains the ability to scale as the data load goes up.

But this may not necessarily be what your business may have to also do while turning data-centric. Data team leaders need to sit with business heads and discuss their critical business priorities, before deciding on the tech solution.

Data leaders in the Enterprise need to automate the data source and distribution of analytics to various interested parties internally. For this, they may deploy an architecture that prefers the shared data models over individual data applications having their own separate models.