Going Back to the Raw Data

In finance, we tend to report based on cost centers and categories. We combine cost centers or cost categories and report on performance based on this. We forecast and load based on those cost centers and cost categories. We are tied to the way we have organized our data, and we tend to invest in projects in order to organize our data in a different way so we can better report if the current system of organization is not working well. 

But in the end, are we looking at and analyzing the raw data? For us in finance, the raw data consists of our journal entries, sales, invoices, etc. How much time are we spending looking at that raw data and making sense out of it? When we look at reorganizing our data, do we spend time putting together rules of how to create journal entries that make more sense so we can do data mining and machine learning? As FP&A professionals on the cusp of harvesting the potential of "Big Data," we must escape the view of cost centers and cost categories and start doing deep dives into the raw data. 

We have a treasure trove of data in the current financial systems and we need to be able to use it. As

an example, lets dive into IT expense analysis. An IT department today in any company may be organized based on reporting structures. The function may have decided to put all the licenses in one cost center to manage the procurement process much better rather than by putting it in different cost centers. So, if the company is spending money on data center, help desk or SAP licenses—the expense for those licenses might come directly into that one cost center. The function could have employees in one cost center working on two different services such as data center and help desk.  To help understand the actual cost of providing data center services or help desk services, one must pull the expenses from various cost centers and categories. One solution is to reorganize or have more cost centers so we can track these separately, but that can cause other issues such as inefficiencies in license management.  No one way of organizing the cost centers is perfect and needs will be changing constantly, so to obtain any insights, we need to go back to the raw data.

Doing the analysis on the raw data we currently have as well as connecting that data to business drivers or operational metrics today does not require huge investments in ERP systems. It requires more manual work of extracting the raw data and doing the detective work.  The majority of that work is done in Excel today. This kind of manual analysis is not sustainable in the long term as it takes time and resources and is prone to errors, but it might be the first step to take to prove the value of such analyses.  It is very important to prove the value of the analytics needed in small pilot projects or proof of concepts so the company can invest in the right software and tools to enable to the analytics. 

So, go out and explore your raw data and breakout of the cost centers and cost categories.