JIS Senior Editors' Blog

Journal of Information Systems

This is a public blog  publicRSS

blog entry

    In a sea of Big Data, what becomes of accounting and...
    blog entry posted March 23, 2016 by Roger S Debreceny, last edited March 23, 2016, tagged research 
    In a sea of Big Data, what becomes of accounting and auditing?
    intro text:

    Dr A. Faye Borthick at Georgia State University is the co-chair of  JISC2016 - 2nd Journal of Information Systems Research Conference, to be held at the offices of Workday, Inc., Pleasanton, CA on October 13 &14. The theme of the conference is An Accounting Information Systems Perspective on Data Analytics and Big Data

    In this commentary, Dr Borthick sets out some of the issues that are relevant for the conference.


    Dr A. Faye Borthick

    Dr A. Faye Borthick


    In a sea of Big Data, what becomes of accounting and auditing?
    A. Faye Borthick

    Evolution of Big Data and Its Insinuation in Organizations

    Big Data and the software for doing interesting things with the data have developed far enough that some trends have emerged. People are clever. Leave them alone with resources, and they will do interesting things with them, giving both intended and unintended consequences. This commentary highlights the landscape of Big Data, not the technical aspects per se but how organizations are starting to use data in different ways. While it is true that some of what this commentary offers does not, strictly speaking, require Big Data with respect to volume, diversity, and structure, the connotations that Big Data bestowed have prompted new ways to stage and use data. For example, “70% of firms now say that big data is of critical importance to their firms” (Malone 2016 A17).

    We invite you to reflect on burgeoning data and its emerging uses highlighted below as you consider innovative practice on aspects related to data analytics and Big Data.


    Managers and investors are tantalized by the prospects of using more data to make companies more profitable and to make organizations more responsive to their constituents (Dwoskin 2014). The growing number of fintech company startups making loans based only on digital data illustrates the phenomenon (Rudegeair 2015). Furthermore, ideas about how to take advantage of data can come from anywhere. This is the phenomenon that propels startup companies into billion dollar IPOs (initial public offerings of stock) in just a few years.

    With new insights into consumer/buyer behavior and product performance, companies and organizations have been compelled to accelerate changes to their business processes to stay current with their product and service offerings (Zetlin 2015). Instead of weeks, General Motors’ profitability analysis of the Chevrolet brand in Europe took only days (Monga 2014). Food companies have noticed declines in sales of the worst offenders of packaged foods associated with packing on the pounds (Esterl 2016) or foods containing GMOs (Brat 2015). How fast can packaged food makers adjust to such a shift? When Ben & Jerry’s set out to source only organic products for its ice cream, it discovered that a key ingredient, organic milk, was not available in sufficient quantity for its volume (Gasparro 2014). The company is caught between consumers whose tastes change rapidly and a supply chain that requires years to evolve.

    Much though companies aspire to change their information systems to take advantage of the business insights that analysis of data can afford, they are hindered by existing systems that have accreted over decades. The bigger and older the company, typically the more systems a company has and the less they are integrated. For example, General Electric hired Amazon to help it “reduce internal applications to 5,000 from more than 9,000 and move them to Amazon and other cloud services,” in order to “allow GE to eliminate 30 of its 34 worldwide data centers and roll out new applications in as little as five minutes” (McMillan and Barr 2015 B5). Other companies such as Whole Foods Market Inc. and Wal-Mart Stores Inc. are “plowing into years-long efforts to merge disparate data sets, in the hope of extracting cost savings and insights about customers” (Norton 2015 B4).

    Where do the data come from that companies and organizations want to analyze? Some of it is the familiar transaction data in accounting systems, usually highly structured with careful editing before it is permitted to enter the accounting system. Some of the data is about the location or behavior of things, e.g., RFID (radio frequency ID) data, recorded in real time indicating the existence of products or pallets of products in a specific location at a specific time. These data are usually highly structured. Then come the unstructured data of interactions, often from social media, whose existence has spawned whole new categories of analysis.

    Lest one think that these data capabilities are only available to large companies, consider the experience of some startups in which middle management has been subsumed in data. Instead of hiring people to get the data needed for decisions (the traditional middle management role), the startups make extensive data dashboards available to everyone in the company. Middle managers are not needed to gather information and make decisions because “every employee can have the tools to monitor progress toward any goal” (Mims 2015). The transparency and accountability afforded by all-employee access to dashboards means that leaders can find out how the business is performing directly without relying on middle management.

    Where do the data live? More and more in the cloud, of course, with a dashboard interface most likely designed by the cloud-services provider or a consultant. This approach arose first in startups because it allowed them to run lean, minimizing headcount. Big companies can move to dashboards, but the costs of taming non-integrated systems loom large, and corporate cultures will have to be transformed. Big companies with their non-integrated legacy systems realize that time is of the essence because startup companies commence operations with fully integrated systems run from the cloud (Ismail et al. 2014; Loten 2015).

    What are the implications for employment of the people that now summarize, categorize, and report data? Merchandising staff are discovering that management wants to rely more on data analysis than instinct for product selection, much to the dismay of chief merchants, “once lionized for their knack for spotting trends, are finding their intuitions being displaced by algorithms” (Kapner 2015 B7). Floor stock traders have been replaced with software, and now financial analysts are being replaced with software (Popper 2016).

    Where is the accountant in this uncharted sea? Because accountants understand traditional accounting data, they are uniquely poised to analyze it, including the related location and interaction data. As more accounting functions are automated in software, managers and investors are expecting higher levels of analysis of all the data. Investors are eager to spot business trends that portend changes in revenue. Thus, they are interested in feeds of transaction data, e.g., summaries by day of consumer purchases at publicly traded companies . The transactions are captured in the normal course of business but become useful for non-transaction purposes, a strategy known as cashing in on “exhaust, ” i.e., “data collected while doing other business” (Hope 2015).


    What happens to auditing when data volumes grow, when data about interactions and observations become available in addition to the traditional transaction data? Transaction data are typically well structured, which makes them amenable to analysis in relational database systems. Interaction data, e.g., from social media or other sources, are typically unstructured. Data about observations may be structured, e.g., RFID data or logs containing process events, or unstructured depending on the context, e.g., comments appearing in logs of process events.

    Data proliferation challenges auditing because auditing has been formulated and conducted in a world in which data were limited and there were no good software tools for analyzing large volumes of data. But conditions are changing.

    Data limitations in auditing gave rise to sampling as a way to obtain evidence about account balances and flows. In the absence of data and computational software, auditors developed manual procedures based on sampling a small number of items (transactions usually) and checklists to ensure that lower level staff could execute audit procedures. But data limitations are falling away. When they know the whole population of data could be analyzed, people just laugh at the sampling mentality. Instead, they want data analytics applied to the whole population to make the data give up their secrets (Murphy and Tysiac 2015). This presents a problem to auditors in that they have decades invested in a sampling/checklist/procedural approach to auditing in a time when their constituents want data analytics applied to whole data populations. Eventually auditing standards premised on sampling will be revised to embrace better evidence (Titera 2013).

    As auditors have dived deeper into company data, the Public Company Accounting Oversight Board (PCAOB) has stepped up its scrutiny of auditors’ testing of system-generated data and reports as a means of prompting auditors to detect more of the latent internal control deficiencies. In essence, the PCAOB is demanding that auditors vouch for controls over “the accuracy and completeness of the system-generated data or reports” (Munter 2015). Thus, even as auditors are facing more data, they are being pressed to detect deficiencies in internal control over the data on which they rely for evidence.

    Accounting firms are taking advantage of growing data availability and increasing software capabilities to create dashboards populated with operating and other related data streams (PwC 2015). The purpose of the dashboards is twofold: to facilitate auditors thinking analytically about risks and their instantiation in data patterns and to enable drilling down through the data to look for underlying causes for anomalies or for business opportunities. This approach can be called an analytics mindset.

    Auditors are shifting from a sampling mentality to a data analytics approach as a competitive necessity. Their investments to reorient audit methodologies are large, in developing auditing based in analytics, retraining staff, and seeking and cultivating an analytic mindset in new staff.


    The ripple effect of Big Data on university level education for accountants comes through calls from employers for entry-level accountants and auditors with analytical skill sets. The business press has documented the shift from armies of people tracking and paying for orders to automation of the task (Monga 2015). “Since 2004, the median number of full-time employees in the finance department of big companies has declined 40% to about 71 people for every $1 billion of revenue” (Monga 2015).

    If manual entries, which used to require armies of people, have been automated, what skills do entry-level accountants need? A typical response usually includes a variant of “analyze data and present findings coherently to colleagues “ (Johnson 2015). The data analytics response has been written into AACSB Standards for Accounting in the form of Standard A7 on data analytics (AACSB 2013).


    AACSB. 2013. Eligibility Procedures and Accreditation Standards for Accounting Accreditation Standard A7. Tampa, FL: Association to Advance Collegiate School of Business. Accessed March 30, 2014: http://www.aacsb.edu/accreditation/standards/2013-accounting/Learning%20and%20Teaching%20Standards/standard7.aspx.

    Brat, I. 2015. Food goes 'GMO free' with same ingredients. The Wall Street Journal, B1 (August 21).

    Dwoskin, E. 2014. Tons of data. Now put it to use. The Wall Street Journal, R6 (October 20).

    Esterl, M. 2016. As sales fizzle, pop makers bill more for a sip. The Wall Street Journal, B1 (January 28).

    Gasparro, A. 2014. How we eat: GMO fight ripples down food chain. The Wall Street Journal, A1 (August 8).

    Hope, B. 2015. Firm tracks cards, sells data. The Wall Street Journal, A1 (August 7).

    Ismail, S., M. S. Malone, and Y. Van Geest. 2014. Exponential Organizations: Why New Organizations Are Ten times Better, Faster, and Cheaper Than Yours (and What To Do About It). New York, NY: Diversionbooks.

    Johnson, K. S. 2015. Outdated: The plain-vanilla accountant. B7 (May 19).

    Kapner, S. 2015. Data pushes aside chief merchants. The Wall Street Journal, September 23, B7  (September 23).

    Loten, A. 2015. Cloud tools tackle new tasks. The Wall Street Journal, B6 (June 4).

    Malone, M. S. 2016. The Big-Data future has arrived. The Wall Street Journal, A17 (February 23).

    McMillan, R., and A. Barr. 2015. Google taps director for cloud push. The Wall Street Journal, B5 (December 24).

    Mims, C. 2015. Data is now the new middle manager. The Wall Street Journal, April 20, B1, B2 (April 20).

    Monga, V. 2014. Big data chips away at cost. The Wall Street Journal, B6 (July 1).

    ———. 2015. The new bookkeeper is a robot. The Wall Street Journal, May 5, B1, B7 (May 5).

    Munter, H. A. 2015. Importance of audits of internal controls: Public Company Accounting Oversight Board (PCAOB). September 9. Available at http://pcaobus.org/News/Speech/Pages/Munter-Audits-Internal-Control-IAG-09092015.aspx. Accessed March 11, 2016.

    Murphy, M. L., and K. Tysiac. 2015. Data analytics helps auditors gain deep insight. Journal of Accountancy (April/May): 52-58.

    Norton, S. 2015. Big companies rein in data sprawl. The Wall Street Journal, B4 (October 22).

    Popper, N. 2016. The robots are coming for Wall Street. The New York Times, February 25. Available at http://www.nytimes.com/2016/02/28/magazine/the-robots-are-coming-for-wall-street.html?rref=collection%2Fsectioncollection%2Fmagazine&action=click&contentCollection=magazine&region=stream&module=stream_unit&version=latest&contentPlacement=9&pgtype=sectionfront. Accessed March 5, 2016.

    PwC. 2015. Data driven: What students need to succeed in a rapidly changing business world: PricewaterhouseCoopers LLP. Accessed May 31, 2015. http://www.pwc.com/us/en/faculty-resource/assets/PwC-Data-driven-paper-Feb2015.pdf.

    Rudegeair, P. 2015. Online firms seek to eat banks' lunch. The Wall Street Journal, C1 (June 29).

    Titera, W. R. 2013. Updating audit standard--enabling audit data analysis. Journal of Information Systems 27 (1): 325-331.

    Zetlin, M. 2015. Breaking free. CIO (October 1): 22-29.