Very coming year, CIO's will be very much busy about exploring or deploying strategies about bigdata, cloud or IoT... I will try to explore bigdata in CIO's shoes.
I do not want to explain again "what is bigdata?" This question is answered many times and this answer can be easily addressed by a simple google search.
However, these answers are not mature enough to shape up a strategy. These macro findings will only increase the mess up and complexity. CIO are requested to put a strategy and roadmap on the table. First things first, listing the challenges is best the step to take first: I have used Thierry Winckelmans presentation called Challenges of Big Data, thanks to him.
CIO's challenges: Marketing Analytics, Sales Analytics, Operational Analytics, Financial Analytics: All these four topics are required deep preparations, exploration and investment. It is obvious that not possible to mention details in this posting, however all comes up with one single question or challenge: which platform(s) or vendor(s). This single question will bring answers for skill set and technology.
Let's take Operational Analytics as the starting point for investment. Collecting machine data is the easiest part.
Thanks to MII, this can be derived smoothly, but preparing reports and analytics are last decades topics.
This era requires more, requires prediction and live action. All collected data should be segmented and bonded to a strong entity. Defects per 1000 production orders by BOM variances should be well analyzed by statistical methods in order to predict an machine failure. This means solid, accurate and really fast platform in terms of developing and applying the solution.
Sales and Financial Analytics are similar to Operational Analytics, but if we look at Marketing Analytics, things are dramatically changing. All others than Marketing we are working mostly controlled and structured data sources and environment. However marketing analytics are bringing 3 major dimensions of data: Volume, Variety and Velocity.
Mostly unstructured data should be found and collected from anywhere, like social media, customer data, RFIDs...
This means you have to listen, observe and build data traps to anywhere to capture what you need. Your platform also should support you in all aspects as well.
Finally platform and vendor is the key for the Bigdata Strategy. Let us look what we really need from the platform:
The platform or vendor should provide:
Accessible from any device form anywhere and simple User Experience: SAP UI5 and Fiori.
Simple in-terms of data storing and retrieve: SAP HANAInstant to stream and real-time : SAP SLT and DS
Support massive scale, type and volume: SAP BW on HANA, SAP IQ
Answer complex questions and predict next: SAP Event Stream Processor and InfiniteInsight KXEN.