Integration is one of the most powerful trends in the modern tech industry but it often gets presented in a murky fashion. If you wish to understand the basics of integration, it helps for us to go back in time and take a look at tech in the enterprise roughly 10-15 years ago.
Back in the day
The first tech bubble brought some wariness to adoption of new tools. Organizations were strictly siloed. IT data systems were present, but the information was laid out into ad-hoc data stores that were created for the convenience of specific software vendors and not for people. Web applications and what we now refer to as “cloud portals” were very rare and many employees had to rely on desktop software that was difficult to maintain and patch. Such software also often involved a lot of IT administration to maintain data connections and license keys.
When organizations had to make the decision to make data systems interact, it was generally at the end of a long and expensive scoping effort that required weeks or even months. This is because the end result would invariably involve hours upon hours of expensive developer time - or even more expensive contractor time. The effort would involve writing mounds of glue code that likely could never be leveraged in another integration. In short, all of this was a mess.
Fast forward to today
We live in a very different world now. Organizations have made large strides to identify their cross-functional data flows and build enterprise structures that almost mimic the clean architectural abstractions of software systems. There is also a slew of solutions, such as Salesforce, NetSuite, SAP, and Microsoft Dynamics that offer powerful features that are specific to particular business functions. With these tools and with modern data integration we are able to:
develop a unified view of all data across different functional groups,
exchange data across different systems and different formats with minimal effort,
model workflows with GUI tools, identifying entire functional areas as “blocks” and wire them together as if they were part of one application,
bring together custom-made data tools into such workflows with only basic scripting or configuration file changes,
automate data storage and notification functions, taking away all of the manual effort needed to keep an organization in sync.
Data still does not integrate itself
All of this convenience is amazing and a welcome change from the headaches of the past. However, it doesn't work by magic nor can you implement a powerful modern integration overnight. It takes experience and industry expertise to spin up an integration that makes the best use of all modern tools. Once such an integration is in place, you must also decide whether or not to manage it in-house or keep it running in top form by leaving it in the hands of experts. Like all decisions in tech, it is not to be made lightly, but a good integration is usually what separates industry-leading, future-thinking organizations from ones that seems oddly stuck in the part.