Imagine that you had to search for a cooking pot within your house. Where would you start and search first? Most people would focus on the kitchen. Where would you look for some toothpaste? Most probably in the bathroom and maybe in the room where you just put the bag from your shopping trip.
Using the context information speeds up search, you are only considering some places and avoid searching the entire house. This is data skipping in normal life. DB2 with BLU Acceleration uses a synopsis table to provide the context information. By avoiding work less resources are needed, less data needs to be processed and you have the result much faster.
Now imagine that the cabinets are labeled, the kids would have cleaned up their room with clothes nicely folded and small junk sorted into plastic containers. In DB2 BLU this would be called "scan-friendly". Some people use "space bags", plastic wraps that can be vacuumed to reduce the storage size of clothes, pillows, etc. Because you can still see what is inside and handle it like everything else, it is "actionable compression" - same as in DB2 BLU which can operate on compressed data.
Now if I could create an analogy how DB2 BLU does the dishes - something I have to do now. Household chores. Enjoy the weekend!