Understanding the value of data integration

To understand the value of data integration, one has to first understand the changing data landscape. In the past few years, more data has been created than existed in all of time prior to that. In 2014, I penned a post asking ‘Are enterprises prepared for the data tsunami’? When it comes to data, enterprises of all sizes and maturity face two core issues: 1) How to effectively manage the sheer volume of data in a meaningful way and 2) How to extract insights from the data. Unfortunately, the traditional ways to manage data start to break down when considering these new challenges.

DIVERSE DATA SETS

In the above-mentioned post, there was reference to an IDC report suggesting that by 2020, the total amount of data will equate to 40,000 exabytes or 40 trillion gigabytes. That is more than 5,200 gigabytes for every man, woman and child in 2020.

However, unlike data in the past, this new data will come from an increasingly varied list of sources. Some of the data will be structured. Other data will be unstructured. And then there is meta data that is derived through analysis of these varied data sets. All of which needs to be leveraged by the transformational enterprise.

In the past, one might have pooled all of this data into a classic data warehouse. Unfortunately, many of the new data types do not fit nicely into this approach. Then came the data lake as a solution to simply pool all of this data. Unfortunately, this approach is also met with challenges as many enterprises are seeing their data lakes turn into data swamps.

Even beyond data generated internally enterprises are increasing their reliance on externally sourced data. Since this data is not created by the enterprise, there are limits on how the data is leveraged. In addition, simply bringing all of this data into the enterprise is not that simple. Nor is it feasible.

Beyond the concept of different data sets, these new data sets create ‘data gravity’ as they grow in size. Essentially, creating a stronger bond between the data set and the application that leverages it. As the size of the data set grows, so does its ‘gravity’ which prevents movement. All of these reasons create significant friction to considering any movement of data.

VALUE OF INTEGRATING DATA

The solution rests with data integration. Essentially, leave data where it resides and leverage integration methods to the various data sets in order to create insights. There are actually two components when considering how to integrate data.

There is a physical need for data integration and one that is more logical in nature. The physical component is how to physically connect the different data sources together. This is easier said than done. It was already challenging when we managed all of the data within the enterprise. Today, the data resides in the hands of many other players and approaches. This can add complexity to the integration efforts. Modern data integration methods rely on Application Programming Interfaces (APIs) to create these integration points. In addition, there are security ramifications to consider too.

The logical integration of data often centers around the customer. One of the core objectives for enterprises today is customer engagement. Enterprises are finding ways to learn more about their customer in an effort to build a more holistic profile that ultimately leads to a stronger relationship. Not all of that data is sourced internally. This really is a case of 1+1=3 where even smaller insights can lead to a larger impact when combined.

THE INTERSECTION OF DATA INTEGRATION AND ADVANCED FUNCTIONS

Data integration is a deep and complicated subject that is evolving quickly. Newer advancements in the Artificial Intelligence (AI) space are leading enterprises to gain greater insights that even they didn’t think about. Imagine a situation where you thought you knew your customer, but the system suggested other aspects that weren’t considered. AI has the opportunity to significantly augment the human capability to create more accurate insights and faster.

Beyond AI, other newer functions such as Machine Learning (ML) and Internet of Things (IoT) present new sources of data to further enhance insights. It should be noted that nether ML nor IoT are able to function in a meaningful way without leveraging data integration.

DATA INTEGRATION LEADS TO SPEED AND INSIGHTS…AND CHALLENGES

Enterprises that leverage AI and ML to augment their efforts find increased value from both the insights and the speed in which they respond. In today’s world where speed and accuracy are becoming a strong differentiation for competitors, leveraging as much data as possible is key. In order to leverage the sheer amount of data, enterprises must leverage data integration to remain competitive.

At the same time, enterprises are facing challenges from new regulations such as the General Data Protection Regulation (GDPR). There are many facets and complexities to GDPR that will only further the complexities for data integration and management.

While enterprises may have leveraged custom approaches to solve the data integration problem in the past, today’s complexities demand a different approach. The combination of these challenges push enterprises to leverage advanced tools to assist in the integration of data to gain greater insights.