3 Reasons Excel Doesn’t Deliver on Data Analytics

Hint: It’s not the number of rows; it’s the relationship with data.

The power of Microsoft Excel for the basic audit is undeniable. The sheer number of businesses that built the foundation of their internal audit program with the world’s most ubiquitous spreadsheet tool is doubtlessly staggering.

But what is confusing is the status quo of using Excel for advanced auditing and data analytics when the tool is fundamentally ill-equipped to meet the complex requirements of such tasks.

To be clear, there is and will always be a place for Excel and the few alternative electronic spreadsheet programs on the market. They’re nearly universally accessible, highly affordable, easy to learn, and just about everywhere.

But there’s no need to further celebrate the well-known strengths of spreadsheet software for basic business functions and the limited internal audit. Instead, it is important to consider where it falls short, and the cracks in its armor become apparent when the advanced audit and data analytics enter the equation.

Auditing in a data-obsessed world

Firstly, let’s establish what we mean by that: the advanced internal audit today is one that leverages data analytics capabilities to assess massive amounts of data from multiple sources. And while it was once considered a nice-to-have, data analytics is widely viewed as an essential part of the mature, modern audit. And frankly, it’s critical these days. Somewhere between Big Data, cybersecurity risks, and AI, the complex needs of today’s audit arise and the limitations of conventional software start to show. These limitations go beyond Excel’s cap on rows and columns, at about a million and 16,000 respectively. Indeed, when it comes to the modern audit, the extents of Excel are found more in its relationship with data than with the amount of data it can retain.

With that, let’s look at the top three limitations faced when we try to use Excel or a program like it to handle the requirements of an internal audit fueled by data analytics.

1. It won’t protect the integrity of your data

Internal auditors will probably agree that an audit is only as accurate as its data. Most people would agree that humans are, well, error-prone. An auditor can bring in as many external records from as many external sources as they like. They can call them accurate, but in the hands of a fallible mortal, the information contained in spreadsheets is subject to sloppy keystrokes, a bad copy-and-paste, a flawed formula, and countless other errors. Spreadsheets emailed between colleagues risk being further compromised with every set of hands they pass through, compounding the risk of error. In a field so synonymous with risk aversion, it’s remarkable any auditor would feel comfortable managing massive datasets with such fickle controls – especially when there’s an alternative.

Dedicated audit data analytics software circumvents the problem by minimizing the element of human error and protecting the data — generally imported from Excel spreadsheets, no less — into a centralized and secure system where the possibility of keystroke mistakes or emailing the wrong file version are entirely eliminated. In other words, the data analytics solution has a very intimate relationship with the data and protects it accordingly.

“In a field so synonymous with risk aversion, it’s remarkable any auditor would feel comfortable
managing massive datasets with such fickle controls – especially when there’s an alternative.”

2. There’s no real audit trail

When human or other error does occur, or when the wrong data enters an audit process, it’s important to be able to look back and determine what went wrong and when it happened. It’s even more critical when dealing with multiple data sources or in continuous auditing situations.

Which points us to another limitation of conventional tools: The run-of-the-mill spreadsheet solution has no intrinsic record-keeping capacity that meets the demands set by even basic audit trail requirements.

Alternatively, data analytics tools naturally create an audit trail recording all changes and operations executed on a database. File and format imports, types of analysis performed, and analysis results are all contained within inalterable file properties and that’s the kind of reliability that lets an auditor sleep at night.

3. It doesn’t have data analytics libraries

The key deficiency of traditional auditing approaches is that they don’t take advantage of the incredible possibilities afforded by audit data analytics. When audit data analytics tools start to talk to data analytics libraries, magic happens.

Theoretically, some of the basic tests data analytics allow can be accomplished in standard spreadsheet programs, but these are time-consuming and complicated pursuits since users must program intricate macros or multiple pivot tables.

Contrast that approach with tools that let users duplicate, join, or stratify data or else run or gap detection or Benford’s Law test effortlessly – no coding experience required.

Also, part of our problem right now is that we are all awash in data. There’s too much of it, and that’s a double-edged sword insofar as it lets us discover incredible insights if we can actually comprehend it and the vastness of it. Data analytics tools help users navigate a data analysis process from start to finish with predefined routine tests that can help a relatively inexperienced user execute, say, a set of routines to detect security issues in an SAP implementation, for example.

Progressing to data-driven solutions

As has been well-documented, internal audit is a little slow to adopt new technology. And unsurprisingly, most auditors’ familiarity with technology extends to electronic spreadsheets only. Which is odd, because between data mining, predictive analytics, fraud detection, and cybersecurity, data analytics and internal audit are natural bedfellows.

In this age of digital transformation, the data-driven audit is becoming the standard and it is interesting that the argument for advanced data analytics still needs to be made in 2019. But with an industry too reliant on aging solutions and with data analytics and data mining deemed the skills most in need of additional training, it’s a point worth driving home.