What about the status of my brand compared to competitors? Why do discount
stores with there no-name products make so much profit compared to
traditional food retailing? How much did we earn extra from last weeks
sales campaign?

For product managers of brand manufacturers and analysts from traditional food
retailing,the answer to those and similar questions is essential.

Analysing systems basing on latest OLAP-technology shall now support them. But
before the analysis can be realised, there needs to be the performance of
duty - the data supply.This process contains data cleaning, data finishing
and data load.

Data Cleansing is the basis for the quality of every single analysis.
Depending on the data source, the extracted data is available as a more or less
significant basic form.The data cleansing process starts here.
At this time, one can find technical as well as logical data alerts. Data may be
missing or incorrect as regards measurement or dimension.
A validity Check provides an indication of unexpected discrepancies.In case of
doubt, the whole data record must be rejected.In many cases, data alerts may be
recognized only in connection with other parameters.

To quote a simple example: 100 sold products at an unit price of 3,99€ and a
turnover of 499€ at the same time show that one parameter is not correct. But
the adequate process allows an analysis and, potentially, a revision of
the incorrect date.

Another part within the process of data cleansing is the adjustment of already
existing data and also the transformation, i.e. the unification of scale units
and decimal places.

Data cleansing is an important precondition for the later analysis, as every
analysis is depending on the appropriate data quality.By an effective data
cleansing it is thus possible to create an initial point for high-quality
analyses, which do support firm management decisions.

topcom possesses a long lasting know-how in the field of data cleansing.
Within the scope of numerous projects, our experienced staff already analyzed,
cleaned and integrated miscellaneous data sources from industry, commerce and
market research institutions. With the help of well-defined processes and
system modules we are able to create an efficient and future-proof
analysing basis.