Let’s just call big data what it is: the pile. It doesn’t have quite the same ring to it as the cloud, but it’s just as nebulous and ill defined. Besides, big data is a pejorative. It sounds sinister and scary, especially with all the news about privacy, the National Security Administration, metadata, and so on. We already have big business, big oil, big tobacco, big pharma, big government, big brother, and the big bad wolf: Do we really need another bogeyman?
The pile is simply the vast quantity of unstructured data that businesses and organizations collect and store. The data sets collected could be anything from consumer demographics and buying behavior to manufacturing and business processes. This information accumulates and grows because of the reduced size and expense of memory storage and the proliferation of ubiquitous data-sensing devices.
For perspective, think about all the data passively collected by everyone with a smartphone. The smartphone has as microphone, camera, global positioning system, Internet capability, and other sensors. Just the pictures on a smartphone alone accumulate quickly and become unmanageable. The photo collection keeps growing, because it’s easier to buy another microSD card than wipe and reuse memory after selecting and editing photos.
The pile has huge potential, but the issue is that not enough benefit has been realized from the effort required to manage and process the complexity. The gains have been incremental, and progress has been slow. The ability to store data has outgrown our ability to manage it, so it piles up faster that we can process it.
The processing of this complexity is often compared to mining for gold. Likewise, it’s worth noting that it’s seldom a gold vein is ever struck. Most gold comes from the tedious drudgery of extracting disparate gold dust, flakes, and nuggets from enormous amounts of earth. In much the same way, data have to be dredged, sluiced, and panned; because not all data are pay dirt, larger companies are better equipped to handle the volume than small and medium-sized businesses. Therefore, enterprises are more likely to reap the benefits.
However, the difference between gold mining and data mining is that we can control the content of the material being mined. For example, if we like a photo we take, we usually share it on social media in a specific album with information about the occasion, location, and identity of the people in the image. Likewise, we should diligently identify, classify, and organize our business data at its initial compilation so that we can more easily normalize, compare, perform analysis, and derive correlations.
Invariably, enterprise resource planning (ERP) is data centric. The whole point of ERP is optimization—maximizing efficiency by enhancing processes and improving visibility. The only way this is realized is through the collection and dissemination of accurate, reliable, and timely data. So, the focus on big data as with all of ERP should be, “How can we do this better and smarter?” For the pile, the focus should be on curating and quality rather than quantity.
No comments:
Post a Comment