Link: http://www.etc-architect.com/?p=51
From ETC-Architect » Architect Global | Data Architect, Global | Enterprise Architect, Global
Analytical Architecture is usually used to explain things long after they happened on the assumption that it helps us to do things more efficient in the future. Now the main failure of many analytical projects was not really down to the architecture, but more down to the resistance of business owners to try to learn of mistakes and to action change. One of the ares where is became quite clear is marketing as the hit percentage to transform a campaign to a sale (C2S is actually the name of the standard business architecture in marketing) has not grown with all the analytical work done. Analytical processing has success stories but they are mainly on the production floor or stock control.
The going joke of marketing is actually that it should be forbidden as unmoral, but since it is so inefficient no one cares. And this is just what changes with big data architecture as big data architecture closes the link for influencing customer behaviour or recalculation on personal risk profile through data mining emails, hoping behaviour and similar. This is why now a lot of people suddenly scream that this marketing should be forbidden, as it suddenly becomes much to much efficient.
So the real value of Big Data is not in the amounts processed and also not in the speed, but in the fact to close the loop to help or manipulate the clients through integrating of analytical data at a fine personal level. So a Big data Architect exploring the architecture beyond the technology will need to understand this and create the right solution that will not focus on the reporting, but on the mining of useful information that will (not could) be integrated in customer decision processes in real time. As such the classical divide between OLTP and OLAP will no longer be there. Additional he needs to remember to also focus on the privacy issues as they are likely to backfire if one only concentrates on the doable instead of the sensitive.
