Study says: More than half of German banks don't even know what data management costs them.
Big data is becoming the "next big thing" in IT. This is all too understandable given the rapid growth in data volumes. "The hay is there, but you want the needles in it," writes the Süddeutsche Zeitung. What many companies forget about the new hype: There is still a lot to be done in traditional data management too. More than half of German banks do not even know what it costs them.
"Professional data management can save operational costs of 15 to 20 percent," says Bernd Richter from Capco. The consultancy firm has also determined that only 53 percent of institutions have guidelines for data governance. In order to make data more available, better controllable and usable, Richter also believes that "in some cases extensive restructuring" is unavoidable.
All study participants stated that their data volume had increased significantly within a year. They are particularly concerned about the quality of the data and its meaningful preparation for management. Unstructured data in particular is growing at an enormous rate, but there is still no usable recipe for it. "The buzz around big data is initially sheer despair at the large numbers, sold as joy" - as Bernd Graff aptly writes on SZ Online. Traditional database applications are overwhelmed, there is a lack of computing power. We look helplessly at mountains of information whose value we do not yet know. However, until the market provides appropriate analysis techniques, there are still plenty of tasks in classic data management. For example, when it comes to overcoming data silos. Capco therefore advises
consolidate the various IT architectures,
automate processes (straight through processing),
clean up existing data,
make processes more transparent and therefore more manageable,
procure and record data in a consolidated manner,
store data centrally and
improve data governance throughout the company.
At the end of this classic path, you will have a better view of the customer, lower data management costs - and a usable basis on which big data can later be implemented.
