Data Ownership

I'm sure that every company has some sort of data being pushed/pulled from one system to another.
One of the biggest mistakes in the integration world, in my mind, is not to take the ownership of data into account when data is being passed from one system to another.

A classic example:
At a client I'm working for we have an application that is being fed information about employees. We manage and enrich certain aspects of the data surrounding the employee to perform certain tasks. Currently we are re-looking an integration strategy and doing some investigation surrounding the flow of information. I find them talking about a certain piece of information, let’s call it X.
Q: Do we use X anywhere
A: No.
Q: Why do we have X?
A: We need to send it to a downstream system.
Q: Do we need to pass it on?
A: Yes: Cause we don't know who they send it to.

So why do the downstream system not get it from the source or the master? Why do we need to be the source of data that we are not responsible for?

The end result of not defining clear owners of data is that you end up with very brittle point to point integrations that can get out of hand and unmanageable very quickly. There should be only one master and downstream systems should be sourcing the required information from that system.
There are obviously many strategies for sharing/integrating information, but unless clear owners are defined, you have bigger problems that trying to get data from one system into another.