I was on a briefing with a vendor that is a well-known integration vendor. They offer a platform that is cloud-based which supports application integration (operational data), data integration (analytical data), along with various other hybrid methods, approaches and use-cases. It’s a broad, well known and respected platform.
I asked a question of the VP Marketing: “Do you help clients’ handle the issues with the quality, trust and consistency of the data flowing through the pipes? Or do you focus just on the technical aspects of moving data from point “a” to point “b”?”
The VP of Marketing handed the question over to the VP of product management. I was advised that this vendor does NOT touch information governance, stewardship, MDM, trust, or anything to do with what moves through the pipe. They just ensure data is moved from one place to the next. So if you were moving dodgy data before you use their platform, you can now move dodgy data much faster, and save IT some money at the same time. In fact, based on the marketing of this vendor, you can digitally transform your business (without any reference to what it is that you are sharing/moving/integrating between people, systems and organization).
I question that. I think I question the statement, “You can digitize your business” this way. If any of our businesses evolve from a batch/ETL form of integration to a real-time message-based (dare I say API) integration, without any change to what is being integration, what have I done for the business? What has changed? I do accept that the act of re-designing the integration approach might lead to a re-design of the business process itself, but that is not the same as saying that data integration leads to business process improvement. They tend to be very different conversations had by very different people.
The final part of the briefing was interesting for a different reason. AI and machine learning in general is popping up all over the place. It is popping up in the integration space. Actually it would be true to say that it is popping up in the whole IT space. There are MDM vendors who are using AI/ML to “discover” good sources and good uses of data; there are integration vendors who are using AI/ML to “discover” successful ways to create connectors between files, systems, users and organizations. This is all good stuff.
Though I focus more on the business value/semantic side of the equation, I accept that the technical efforts employed – by business or IT roles – can improve results. So I accept, perhaps grudgingly, that “it depends”. For some organizations, a better (technical) mousetrap for integration might improve business performance. For another organization, the mousetrap needs to focus on semantics. It seems that it is “horses for courses”. I think I am leaning away from seeking the golden nugget, the “one”, the answer that will fix everything.
I think I am leaning toward “levers”. I am thinking that data and analytics leaders (as in Chief Data Officer, or CIO’s for that matter) should look through a lens that offers up levers:
- Which lever should YOU pull/push for YOUR business at THIS time for THAT outcome?
As such there may not be one answer that ever fits since there is not one lever, one outcome, one rime, or one chance. But that is not how vendors’ sell their stuff. To a vendor, every answer can be found in their box of tricks. As such, digital transformation is not one thing – and not one technology or approach can lead to digital transformation. I guess that makes sense…