A few years ago, there was an unspoken consensus that the real multiple transactions players in “big data” were limited: the retail sector, with their personal loyalty cards, telecommunications, based on call history and perhaps the banks, as long as there was sufficient confidentiality with regards to their customers details.
First of all, what do we actually mean with the term “big data”? In short, it refers to an approach whereby a data scientist uses specific techniques and analytical methods to find relationships and expectations, based on the data you provide. The saying “garbage in = garbage out” regarding the quality of the data still applies, but we’ll leave that aside for now.
Where it gets problematic for the average business these days is the size of its financial data, the foundations of the dataset. It is often small and only generated from internal sources.
Seriously, let us look at an ERP or accounting data. The number of records in identical format, the frequency and speed of storing the information can all be seamlessly managed using traditional Business Intelligence (BI) methods. Creating map-reduced datasets, these digestible chunks of information within a “big data” approach, makes little sense.
Such a “small data” BI solution usually does generate the appropriate answers, even more when supplemented by a well thought through data warehouse, or cube, built with a correct and frequent upload process (Extract-Transform-Load, ETL). Useful filters and zoom capabilities finish it off. Attempting to find hidden relationships within the data, another characteristic of “big data”, does not really any add value with such a limited statistical population, does it?
Possibly this is where the CFO can show its trump by guiding the organisation through mountains of information. By asking relevant “business” questions or outlining the right context, without only focussing on financial figures, she can lead her statistical and mathematical people in the right direction.
This is presumably the missing link to start having useful “big data” projects in companies and actually fits well within the picture of defining what the CFO’s duties are, within a much sought after role as strategic business partner of the CEO.
Martin van Wunnik - November 20th, 2016