How good the data is in the company's business
software systems is critical to enabling usable
reports to be generated. But how "good"
the data is involves looking at the data in
several ways. Further, how the data is setup
in the software can literally enhance or prohibit
the functionality of the software. User frustration
with a computer system may be due to the poor
setup of the data.
Before you consider scrapping a computer system,
make sure the system is really the problem,
and not how the data is setup!
There are numerous aspects to look at when
speaking about the "data":
Data Setup looks at the entry of the data in
the software based on the data fields available
for use. Many software applications come with
extra data fields that can be used to setup
characteristics about customers, finished goods,
raw materials, and vendors. Further, there is
a distinct difference in setting up data "horizontally"
versus "vertically". Many software
applications allow for flexibility to do both,
though one method may be functionally better
for one company than another.
Data Standards are required because they define
how data is setup in a computer system. Without
data standards, the data setup can be somewhat
hodge-podge and thus negatively impact functionality,
such as searching. For example, some companies
have intelligent ID's for customers - this intelligent
identification pattern is thus a standard. A
data standard may have been defined to help
enable system functionality.
Data Accuracy is concerned with the "correctness"
of the data, much to do with the mathematical
calculations performed by the software. For
example, does the quantity ordered multiplied
by the unit price arrive at a correct extended
amount on a customer order? Do the pricing adjustments
correctly impact the extended amount? Are sales
commissions calculated correctly?
Data Integrity looks at the "completeness"
of the data across the system. When finished
goods are picked or shipped, is inventory reduced
correctly? Is the general ledger updated accurately?
Is raw material inventory back-flushed properly
after finished goods are moved from manufacturing
to inventory? Do the history log files correctly
reflect the transaction that took place?