Reliable data mining and reporting depends on being able to trust the assumptions that are made about content and quality of the data that is being analysed and aggregated. But after years of use there may be issues with the quality of the content; sometime's it's not even close to what was expected, and often for good historic reasons.
Our Data Content and Quality Review service can investigate and report on existing database content and structure. This could be used before initiating projects to ensure that:
We can also help organisations that have commissioned analysis and mining / reporting projects, but been surprised and disappointed by the results in some areas.
Our reviews cover structure and design of the database, as well as content. If denormalisation was used (often done for performance reasons - servers weren't so fast in those days) this will have been known to application developers but may not have been more widely exposed. If new queries and access strategies are used for new reports or to mine the data, unexpected denormalisation and null values can lead to misleading outcomes. We can help you uncover these situations and develop plans for working around them or for fixing the design.
Data Cleansing work may be required before data that's been built up over years of business transaction processing can be confidently used for new reporting, analytic and marketing work. SMH Systems can:
We may also recommend improvements that could be retrofitted to database or application programs - such as adding triggers or additional referential integrity, to prevent poor data from getting entered in future.
Another situation where content quality review should be performed, is when existing data is to be exported and taken up into a replacement system. Application replacement and business mergers are common drivers for this.