In the last few years, big data has become simple to acquire, store, and process. Amazon, Rackspace, and Google provide amazing services that can store and process a lot of data easily and quickly. IBM has developed a machine-learning tool (Watson) that you can easily integrate over your data. Automation is no longer a dream of the far-off future but a foreseeable present reality.
Companies can track millions of events every day on their website, record the entire purchasing history of customers, and collect endless information about stock capacities supply and demand. Yet many are still calling the resulting information “big data” and failing to make full use of it.
Yes, the data is not always stable and correct. This is why it can be helpful to rethink the way your company works with data. Adopting more lean and agile approaches to dealing with your data will help you build the right infrastructure to meet your needs. You’ll need a well-chosen set of raw data to serve as a base for creating aggregate tables and good actionable reports. If you plan to have an automated performance-optimization tool in the future, that also will require well-chosen data.
Hiring a team of data scientists won’t save you if you don’t have stable and correct data. In the same way, having a large marketing team working with manually processed data will do you no good. Adopt a data strategy today. Start by collecting everything, then aggregate only a reduced subset, and report on a further reduced subset. Then slowly increase it again.
Takeaway: Big data is no longer BIG– it’s just data, and can enable many things, but only if the right infrastructure is built to begin with, in a lean and agile way.