Data Quality Elements
Topics:
data quality company services | roi
data scrubbing info | maximize
data quality | service material
data cleaning info | complete
data quality service information |
data quality company | repairs
data cleaning info; | more
database quality information | better
data quality company | about
data cleaning info |
data quality company information|
In order to determine how much data scrubbing is necessary to tune your data, we
need to establish some basic data standards to which we can compare your data. So,
what is good data or "clean" data? Good data usually have the following traits:
Topics:
data quality company | service info
data quality information | how to fix
data accuracy | what is important
data normalizing | why significant
database quality information improves
data quality |
-
Relevance: Do the data meet the basic needs for which they were collected, placed in a
database, and used?
Also, can the data be used for additional purposes (e.g., market analysis)? If not, how much
time and expense would be needed to add the additional features required?
Is it possible to use your database for several different purposes? For example,
use your database for determining what subsets of customers are more likely to purchase
certain products; or which advertisements or e-mails may be more successful with select
groups of customers than others.
-
Accuracy: If your data are inaccurate, then they can't be trusted; and if they can't
be trusted, you must fix or discard them. The best you can hope for is that they don't
contaminate your key decision-making queries; providing you with erroneous information
upon which to base your business decisions. How accurate do your data need to be? How
accurate do the decisions based on them need to be?
-
Normalized: Your data may be accurate, but accuracy doesn't make much difference if your
data aren't normalized. In short, data normalization means that every bit of information
is stored in its proper place. For example, the First Name field in your data base only
contains first names (or initials) or nothing at all. There are no last names in there;
no prefixes (i.e. "Mr & Mrs") in there - and under no circumstances would there be any
"pseudo data", such as "?????", "Unknown" or "N/A", etc.
-
Timeliness: The USPS says that 35% of all bulk mail mailed every year ends up in a postal
service dumpster due in great measure to out of date addresses. National figures indicate
that about 15% of consumer addresses change throughout the year; while as many as 20% of
business addresses – even higher with more volatile businesses (such as restaurants) -
change over the same time. In addition to wasting your mailing budget, dirty data
provides a deceptively misleading map to guide your all important business decisions.
-
Completeness: in database terms completeness means that there are no missing records and
that no records have missing data elements. If there are fields that are more than 50% empty,
you might consider dropping them from your file; unless they represent exceptions (i.e. a
"Don't Mail To" field in a mailing list where there is a reason to mark records that are no
longer mailable rather than deleting them).
Find out how data scrubbing improves data quality.
Contact us for more Data Scrubbing information
Topics:
data quality company |
data scrubbing info |
data quality service company |
database quality information |
data accuracy |
data quality |
data normalizing |
database quality information;
|
