Another way to extract value from curated, expensively stored, legacy data.
The previous Ugly Duckling piece highlighted monetizing data by distributing it. Today’s focuses on value from reanalysis.
Reap analytical results using cheap compute power that was only dreamed of when lurking data assets were generated. (A 1991 x86 chip was about 12 MIPS, a current i7 comes in at around 220,000 MIPS.)
Applications designers dream and scheme to collect, process, and store data in hopes of future exploitation. Lots of these features and projects get cut for time and budget getting to working applications. After the efforts to get full operational features working many of these dreams are buried away deep in the wistful thoughts of applications designers. Processing time in a business cycle, compute power, and computing tools are all reasons that “never got to this”.
Some examples:
1. Cost history and allocations for costs that obscure and obfuscate the ability to compare older baseline costs to current costs
2. Context information linking customer order management patterns and consumption to external economic conditions and supply chains
3. Improvements delivered thru business relationships that demonstrate the bonding between enterprises
The compute and storage power can be used not just to “heap things in the warehouse” – the new cheap compute and storage power can be used to:
a) Link highly curated data to provide context for newer data
b) Flatten data structures to make processing easier and faster, and
c) Prepare data (by cleaning up names, addresses, and associated history, etc.)
Our next post will discuss how to realize the value of transforming and better analyzing data.
Comments