By Nathan E. Myers, 3/2 MBA ‘99, CPA
Digital Transformation Consultant
Author of Self-Service Data Analytics and Governance for Managers, by Myers and Kogan (Wiley, 2021)
There is so much emphasis on data and analytics in the graduate accounting programs. Isn’t learning Excel enough for most accounting applications?
Yes and no. Clearly, Excel offers excellent functionality to explore a dataset. When I graduated from the 3-2 Accounting MBA Program in 1999, the fact that we were using pivot tables to analyze field values and ranges, were performing time-series analysis of target fields to identify trends and anomalies, and even performing regression analysis to understand relationships and drivers of outcomes put us ahead of the curve. Excel offers a full suite of data analytics features and functionalities that have allowed a full generation of analysts to add significant value to organizations and it remains an important tool in an analyst’s kit. However, with advances in self-service analytics tooling, Excel is increasingly relegated to the quick and dirty exploration of datasets. While much can be gleaned about an array in the Excel sandbox, more robust tools are needed to assist with processing, as we move from understanding and mastering source data to answering the pressing business questions we face.
Why are more robust tools needed, and what are the problems these tools are addressing?
It may be that your role calls for you to be the “tip of the spear”, always answering new questions – and answering them only once. More often than not, however, you will find that when an analyst develops a way of interrogating a dataset in pursuit of value-added insights, the answers will be demanded with frequency thereafter. Remember that systems are the strategic home of processing, whereas analysis performed outside of them is tactical. Where possible, the newly developed analysis scripts should be added to core platform functionality. Many large companies deem Excel-based processing as risk prohibitive, as it introduces the opportunity for – nay, the probability of – manual error. Organizations are increasingly looking for process transparency, process stability, efficiency, and ready repeatability, once the new interrogation runbook or script is bedded down into a proper process. These requirements detract from Excel’s desirability as an analytics platform to live in, on an ongoing basis.
Ok. Systems are strategic. Other tools, including Excel are tactical. So, are you recommending other tactical tools as replacements?
Just remember that while analytics is rightly characterized in its purest form as beginning with a question and using advanced diagnostic, predictive and prescriptive tools and techniques to uncover hidden answers and react to them, it is a term with even broader meaning. I have been driving the replacement of routinized process steps performed manually in Excel with other tactical tools such as Robotics or RPA (for high frequency, repetitive processes), Tableau (for interactive dashboarding and visualization use cases), and Alteryx (for Extract, Transform, and Load use cases). This, too, is data analytics. Remember that if systems featured all required functionality, no such accommodation work would be required in spreadsheets. Until core systems catch up to what is needed on the ground, using self-service data analytics tools as a stopgap to emulate system-based processing can achieve those very same control and efficiency goals that systems support. Even though prominent software names may change over time, use cases persist. Ultimately, operators must capture inputs, master and enrich data, apply formulae, and format outputs for load, further processing, or reporting. This work can be structured and accelerated outside of Excel in a way that better protects the value delivered by the process.