Data Quality in Data Migration Projects
Data migration projects go wrong. Often. We can say this without equivocation. Bloor research puts the percentage of projects that either overrun or fail completely at over 80%. Our feeling, shared with Philip Howard, the author of the report, is this may be a little exaggerated and out of date. Philip will be re-running his survey later this year and we shall see. But we all still know that many, probably the majority, of legacy to new system journeys are fraught with anxiety and pain and often end badly.
But why so? After all, we have been creating computer systems, and therefore stocking them with new data items, for nearly fifty years now so surely we should be quite good at it?
Bloor identifies a number of issues, with which I concur (lack of a standard methodology, lack of trained staff, budget underestimation etc.), but in this article, I will concentrate on the Data Quality issues that bedevil most failing Data Migration projects (and I should know, I’ve parachuted into a lot of them).