Philip Hung Cao

Stay Hungry. Stay Foolish.

How to Overcome Common Pitfalls in Data Analytics

4 min read


The proliferation of data analytics in the world of internal audit has been a boon to some companies and a disappointment to others. Companies using analytics effectively are auditing with fewer resources and getting better coverage; reducing fraud, waste and abuse by the millions; and providing executives the ability to make smarter, data-driven business decisions.

But far too often, audit functions are launching data analytics that sputter, flame out and waste precious resources, and most immature data analytics programs are stumbling over the same few hurdles. Here are the five main hurdles and how to clear them.

Get Executive Buy-In From the Start
Data analytics is a disruptive technology, particularly for internal audit functions, and switching to data-driven decision making can be jarring for executives accustomed to basing decisions on intuition and business acumen alone. Preventing new data analytics initiatives from being stymied early on requires investment, agreement and evangelists at the top of the organization.

To win the hearts and minds of your executives, cherry pick a few analytics projects that promise a high return on investment through cost recovery or an increase in internal audit efficiencies. Audits of financial processes like T&E reporting or accounts payable often yield big returns from easy-to-access data. Demonstrate the value of a modest investment in analytics, and get your executives on board.

Plan for the Data Overload
With the quantities of data collected in most corporate environments today, sifting through all of them can be a headache. But byfront-loading your time—spending more time planning, identifying requirements and cleansing data—you can drastically improve the likelihood of a successful project and reduce the time spent on downstream analysis and review.

A key objective during the planning phase should be to identify all possible data streams that could impact the result of your planned analytic. Document those data streams, where they are stored, who the owners are, what related data might be relevant and what permissions will be needed to access them. By investing in an accurate inventory of the relevant data upfront, you can launch the project with confidence and avoid unexpected roadblocks or delays during the project.

Build a Data Analytics Team that Works
A recent Sunera survey of Fortune 50 to Fortune 2000 companies highlighted two key success factors for companies building internal analytics teams. First, companies with mature internal audit data analytics programs typically have resources dedicated specifically to data analytics. Second, program maturity correlates with the size of the data analytics team in relation to the overall internal audit function. A mature program often requires an investment of 10-15 percent of the overall internal audit budget. But with salaries skyrocketing and a high demand, it can be hard to snag the right professionals for your data analytics teams.

Using a strategic approach, you can build a skills-based team that fits the needs of your organization. Start by targeting individuals who have skills across a few key areas like business analysis, project management and data analysis. From there, look to add roles like IT specialist and data hygienist. These resources ensure you have the infrastructure support to scale your analytics programs. And finally, when you are ready for your data analytics program to really take off, consider adding individuals with graphic design skills for visual analytics and predictive analytics capabilities.

Fight Back Against Dirty Data
More than 25 percent of critical data in Fortune 1000 companies are “dirty,” meaning they are inaccurate, incomplete or duplicated. Dirty data can plague your data analytics initiative at any stage, derailing your budget, generating misleading results and potentially causing project failure.

The best (and only) way to fix the dirty data problem is through better institutional data governance. Proper data governance means clearly communicated objectives, processes and metrics; data quality controls and issue resolution processes; clear and documented data conventions; well-understood roles and responsibilities for analytics resources; and data governance training. Developing a robust system of data governance is worth the investment to ensure that quality data are available to your team.

Make Data Accessible With Visual Analytics
What good is data analytics if executives do not understand the results? That is the central question driving the spread of visual analytics platforms through audit functions nationwide. Visual analytics allows for meaningful exploration through analytic results, giving executives the ability to discover cost-saving or efficiency-improving trends, opportunities and relationships in the data.

Visual analytics gives executives and auditors a powerful tool to ensure that data deliver strategic and significant value to the organization. As you undertake each data analytics project, consider how the output can be visualized to communicate data in the language of your executives, speaking directly to the business challenge at the heart of the analytics initiative. With proper planning and a modest investment in tools and expertise, your organization can provide a far greater level of insight to executives making critical business decisions.

Cliff Stephens is a director in Sunera’s Data Analytics practice where he develops analytics to improve clients’ internal audit functions and provide value to executives. Before joining Sunera, Cliff created and led the Internal Audit Data Analytics team at Home Depot. To read Cliff’s full white paper on this topic, please visit

Cliff Stephens
Sunera Data Analytics

[ISACA Now Blog]

Leave a Reply

Copyright © 2006-2022 Philip Hung Cao. All rights reserved