For some, it is simply a matter of having multiple dashboards on one computer screen, and each dashboard having its own direct connection to a specific data source.
Others believe that data integration means having a single view of the world where all data sources are joined together and presented in one dashboard.
As with most things in life, there are pros and cons of these two types of data integration.
From what I have seen and heard over many years, the vast majority of organizations have taken the first approach because it’s the easiest to accomplish. For the most part, all you need to do is download a connector and link each visualization to the relevant source.
It’s worth pointing out that integrating social media feeds (such as Twitter) is an increasingly popular type of integration source because it’s not a complex data set. Not surprisingly, there are literally thousands of vendors that offer data integration services designed to help customers analyze consumer sentiment.
However, the lesser-travelled road is the most exciting area of data integration; and this involves providing organizations with a single dashboard of data pulled from multiple internal and external data sources.
The enterprise data sources here can be incredibly complex - such as marrying spend data from 10 disparate instances of multiple ERP systems and being expected to cleanse, geocode and risk profile the data using such data providers as Dun & Bradstreet.
The primary business benefit of this type of integration is having your complete dataset ready for reporting and analysis without the mental gymnastics of having to make the connections while you look at multiple disconnected dashboards.
If you’re asking yourself why organizations haven’t embraced this second approach, it’s simply down to two facts:
To put this into layman’s terms, data integration in a single dashboard is like being in a building where all floors are connected and all rooms on each floor are themselves inter connected. Each room contains vast amounts of data which can easily be accessed and analyzed.
This boundary-less style of data discovery isn’t possible in the first approach because in that scenario you are in single floor building with a series of rooms that are completely isolated from each other and where analysis is dependent on the unique availability of the data in each room. The tools available are therefore generic, blunt and deliver poor information.
The level and depth of data integration is increasingly becoming a competitive differentiator for data-centric organizations. It’s a race to give users as much relevant information as possible in a timely manner so they can make faster, smarter decisions based on accurate and connected data
If you are not combining your internal data sources and analysing these with relevant and timely external data, you’re probably playing catch-up. The good news, you’re not alone.
According to Forrester Research, more than two-thirds of organizations expect to increase spending on data management services.
So, the next time you’re looking at how to integrate data, first understand what you mean by data integration. Your answer will define the type of technology you need to support your organization’s information management requirements.
You’ll soon realise that your BI strategy will change from being technology-based to one that is data-centred; an evolution that will empower your colleagues to exploit data to accelerate business objectives.