“Big data” is a phrase we’re increasingly used to hearing. In business terms, it’s often associated with online giants such as Amazon, Facebook, and eBay. At their core, these businesses each have a platform on which huge volumes of data are stored and processed.
This data is analysed to highlight trends, identify correlations, and ultimately increase competitive advantage through targeting and segmenting. This big data trend has emerged due to a number of factors: two key drivers are the dramatically reduced cost of data storage, and increasing web transaction volumes driven in part by the number of connected mobile devices.
Any FD can increase the value they bring to their business by applying some big data thinking to the way that they work and think about their role. All businesses have multiple data sets – in the general ledger, in order processing and billing systems, and in operational systems for example.
Powerful insights can be generated by bringing together data from different sources and analysing it in new ways. This is true in any business – whether or not it happens to have a high-volume data platform at the core of its operating model.
FDs are in a unique position to do this work – they have the established functional strengths of numeracy, accuracy, discipline and independence; they have a solid understanding of the drivers of the business and how its internal functions interact with each other; and they have a key role in measuring how the business is doing. From this starting point, FDs can then use big data thinking to bring unique insights to the management team, and ultimately improve the performance of the business.
A great place to start is to think about four dimensions of analysis and how they can be developed using this approach:
How can we combine different data sources to create reporting that gives additional insight to how the business is performing? For example, rather than reporting the business performance at total level, can we break it down into meaningful segments that behave differently from each other? Are trends getting enough attention, as opposed to period-by-period comparisons to plan or previous year? Can we use data from different sources to illustrate relationships that aren’t visible from our normal financial reporting?
There will always be ad hoc questions in a business arising from particular issues or changes in the market. Why are our gross margins lower in this particular period? How have we increased unit volumes compared with the previous year? How has our productivity changed over time? Combining data from different sources can be useful in addressing these questions. Once the analysis has been prepared, think about whether it really is a “one-off” analysis, or whether it should be incorporated into ongoing management reporting.
A big data approach can be particularly useful in spotting exceptions or outliers. Are these errors that need correcting, and if so how can we prevent them re-occurring? Are they genuine one-offs because of unique circumstances? Do they highlight opportunities or risks to the business that have so far gone unnoticed? Certain exceptions could for example point towards a potential new line of business, or a loop-hole that can result in a loss being made on certain products.
Should we enter or invest in this new territory? Where should we be establishing a new location? Should we buy this piece of equipment, or find a service provider and outsource the work instead? Once again, combining data from multiple sources is key to getting a complete picture of all relevant factors.
Read more about big data:
- Big data and the SME
- Why big data isn’t living up to the hype…yet
- 4 ways the European Commission is driving big data
The tools and skills needed to carry out this analysis will vary depending on the size and nature of the business. In any scenario, I find it helpful to think about three levels of data.
1 The original data sources. In some cases, all the data you need will be in a single integrated ERP system – great. In many businesses however, there will be separate files containing the required data. Even if there is an ERP, the data required is likely to be stored in multiple data tables.
2 The data extract. In order to keep things manageable, define the data points you need for the particular analysis – the time periods (e.g. by month, by week), data ranges (e.g. all operating units in a particular geography), and data fields (e.g. unit sales by product category). In many businesses, the data will be extracted in the form of a csv file (comma-separated values) or other standard data format. If multiple data sources are being used, these various extracts can often be combined into one master data file. As finance people, this may commonly be a spreadsheet – but could also be a database or a web-based analysis tool.
3 The data analysis. This is where the calculations and comparison are done, based on the data extract. These may be in the form of tables, charts, or exception reports depending on the application.
Thinking about these three different levels is helpful in capturing the appropriate sub-set of data for the required analysis, and ultimately producing an output analysis file that’s both useful and manageable.
From a standing start with this big data approach, I’d advise to start small and then build from there – identify two or three areas to work on where the impact will be greatest to the business. On an ongoing basis, continually think about how multiple data sources can be used to generate meaningful analysis and provide a richer understanding of how the business behaves. Adopting this approach will broaden the definition of the FD’s role, increase their value to the business – and of course make the job more interesting and rewarding.
Nick Williams has been a finance leader with private-equity backed, high-growth technology businesses including telematics group Masternaut.
Share this story