Page 129 -

MDM Maturity Levels 93

and post-condition during data updates. For example, an
interest rate is valid for the MDM system only if the value
corresponds to a maximum calculated in relation to certain
parameters related to a period and a sales network. Under
this rule, the period and the sales network are themselves
master data hosted in the MDM system.

Data structures can be complex, involving data
dependency links and associations between business objects.
Consequently, validation rules must ensure the data
integrity constraints even in complex data models.

In particular, data modifications must take into account
the state of the business objects involved in the updates. For
example the update of a product description can be
dependent on its state: “to negotiate”, “outdated”, “archived”,
“activated”, etc. Any modification may be disallowed if the
product is not active. We will return to this idea later in the
book. It is an important concept that concerns all the
business objects in the system: product, employee, customer,
partner, other third parties, statements of accounts,
organization etc. These data objects hold different states
which define their lifecycles. Depending on their lifecycles,
the updates allowed are not always the same.

These constraints, relying on the business objects’ states,
are significant validation rules that the MDM system must
take into account in a formal way. A static MDM system is
unable to do this. It is not just a technical limitation but one
of methodology; neither the data modeling nor the
implementation takes into account the different states. The
consequence of this is the implementation of hard-coded
rules in software surrounding the MDM system. This failure
to formally take into account business objects’ states in the
MDM system causes serious issues:

– this type of MDM system is not seamless for business
users. The MDM administration team is unaware of the