Application vs. Data Integration: Which is better?

February 7 2012

1 comment. 0
motif

This question boils down to one thing: are you part of the transaction or not?

Data Integration grew out of the adoption of relational databases and the need to move information between them. Typically, data integration is batch-orientated and deals with data at rest. In other words, the process that created the data has already completed.

Application Integration, on the other hand, deals with integrating live operational data in real-time between two or more applications. Typically an “event” will occur. For instance, when a customer places an order, this triggers an integration flow that updates and enriches data in other applications in real-time.

Which is better?

Both application and data integration have a place in the world. Hollis Tibbetts of ebizQ does a great job of positioning the two.  However, application integration is where the action will be over the next few years. Here’s why:

  1. Instant datafication” — Data integration happens after the fact and typically in either interday or intraday batches.  The results of this data integration are used to understand how well our businesses are performing, spot problems and anomalies, and to make decisions about how to maximize resources. To be competitive, companies can no longer wait for this data. Time is now measured in milliseconds, not minutes, hours, or days.
  2. SaaS has changed the nature of application data — Previously, all our data was stored behind the firewall in databases. This data still exists, but now there is data fragmented among SaaS applications as well.  The way to access SaaS data is through the application API, not the database.  This is application integration, not data integration.
  3. Agility — Applications need to respond to new data feeds, changes in logic, and new applications at a much faster rate to stay agile. It can no longer take weeks or months to on-board a new partner, change application logic, or add new applications to the data mix.
  4. Cloud and Open APIs are a great leveller —  Applications become more consumable and easier to manage with Cloud and Open APIs. SaaS applications usually have APIs that greatly reduce the complexity of integrating applications in real-time.  Cloud platforms like iPaaS enable a new level of simplicity for integrating applications for real-time exchanges.

These factors represent an evolution of the way business gets done, especially the need to react immediately. To compete, companies of all sizes will need to think in real-time. The right platform is thus needed to make it possible to react to change in real-time.

Will Data Integration go away?

The short answer is no. There will still be the same demand for ETL, data cleansing, and MDM, but application integration is the only way to deal with the changing requirements of integration. Capturing the data as part of the transaction means it can be acted on immediately, whether that is synchronization with other applications, corrective action, or escalation.

The future of Integration

The future of integration is application-centric and real-time. Furthermore, it won’t feel like integration, which has typically been hard to do and expensive. We’re betting that our iON platform will change how integration is done. We’re not the only ones that see the massive opportunity,  here’s what Hollis Tibbetts has to say:

“Much like Salesforce with its $25 per month entry-level cost upset the Siebel Systems (whose competing product was often hundreds of thousands of dollars) applecart, MuleSoft’s iON will do the same to the EAI marketplace.”

 


We'd love to hear your opinion on this post

One Response to “Application vs. Data Integration: Which is better?”

  1. In my opinion, data integration unfortunately is misunderstood. It has typically to do with the transfer of data in a bulk fashion and not much with realtime vs batch. Based on the current understanding of application integration, bulk transfer of data through application API, data services should still be considered part of data integration and not necessarily application integration.

    Agree(3)Disagree(0)Comment