Dave Frankel was recently boarding an American Airlines flight in London when he heard a song come over the intercom. Although it was a Muzak version of the tune, he recognized it as one by a group he liked.

When he reached his seat, he tweeted his appreciation of the airline’s interesting choice of music.  “And then I got a response on Twitter from American Airlines in a second saying, ‘I’m glad you liked it,’” Frankel, the president of EDGAR online, says. (He has since forgotten the name of the song.)

To be sure, he got a less positive answer when he tweeted back a request for a seat upgrade. Though the request was playful, the airline might well have considered compensating him, since it had received something valuable in return.

Frankel tells the story to demonstrate the second-by-second, almost exponential growth of potentially actionable business information under the rubric of Big Data — not only via Twitter, but also Facebook, Amazon.com, Yahoo and similar sites, as well as email. He marvels that “just by sitting on a tarmac and tweeting something,” he had provided American Airlines with data it could potentially use to improve its marketing and customer services. Other fliers, after all, might like such songs, and the company might subtly help boost its revenues by playing them more frequently.

Such opportunities are generated by the growing ability of companies to aggregate and draw correlations from huge amounts of unstructured data. Accompanying them, however, is the threat to corporations and accounting  professionals of falling behind. “If companies don’t wake up and understand that this isn’t going away, that it’s only going to get bigger, then they’re going to put themselves at risk,” Frankel warns, adding that “the accountants and [financial] reporting managers are still not there yet.”

Trax, a Singapore-based firm, provides an image-recognition ap that gathers data from photos taken of shelves at retail stores.

Trax, a Singapore-based firm, provides an image-recognition app that gathers data from photos taken of shelves at retail stores.

Indeed, outside of accountants and finance executives who actually work for companies in businesses that provide or deliver data products and services, few such professionals have reportedly been able to catch the wave. A big reason may be that those involved with internal and external financial reporting are limited, by training and inclination, to working with structured data – the kind that can fit readily into tables, Excel spreadsheets and, ultimately, financial statements.

The problem for accountants is that the future belongs to unstructured data — the kind that can be dug out of tweets like Frankel’s, as well as videos, photographs and the vast amounts of text floating free on the Internet. “Unstructured data represent the largest proportion of existing data and the greatest opportunity for exploiting Big Data,” according to a recent editorial in the Journal of Information Systems (published by the American Accounting Association).

Such unstructured data includes the plain text found in the Management’s Discussion and Analysis (MD&A) sections of company 10-Qs and 10-Ks, and in corporate press releases and interviews with corporate executives, the authors of the editorial note.

Valuable as such data may be, they’re outside the grasp of most accountants and finance executives. Thus, while traditional corporate accounting and auditing “are essential for economic production activities and will continue to be performed,” the authors write, “current accounting and auditing methods are in danger of becoming anachronistic
in the face of an economy increasingly being driven by Big Data.

In short, reality is swiftly outpacing the ability of accountants to gauge it. What needs to happen for them to start to catch up?

To some, the future evoked by the editorial’s writers might appear dark. Referring to employees and management as “nodes,” they suggest that accountants “must view the possibilities associated with Big Data, of knowing much about a corporation, including knowing a substantive amount about who works in a corporation.”

Granting that many questions still need to be answered about the propriety of surveillance of employees, they wonder if such employee metrics as time on the Internet, sites visited, telephone calls made and geographical location should be created and reported.

Yet although “it seems objectionable and invasive that a stranger could know virtually everything about another person, knowing as much as possible about a corporation is much more palatable,” they write, noting that employees have less privacy protection than exists for individuals outside a corporate setting. Speech, documents and email generated during work and via corporate resources may be fair game for the Big Data gathering efforts of corporate accountants.

Swimming in Big Data
Yet before diving into such difficult waters, finance and accounting professionals might do better to learn from peers who actually work in the world of data analytics. While Nina Tan, CFO of Trax Technology Solutions, for example, is involved in the valuation of her company’s data assets, she also uses unstructured data in more traditional accounting and finance functions.

At Trax, a Singapore-based firm that provides an image-recognition mobile application that gathers data from photos taken of shelves at retail stores, Tan is “moving finance from a reporting-centric to analytic-centric financial function,” she wrote in a recent email in response to questions from CFO.

Thus, Tan pores over data to unearth the causes behind the company’s sales, costs and profits and to create budgets and provide input to rolling forecasts. Via analytics, Tan, a CPA, gets an overview of the data that drive revenues and operating and sales expenses, including when, where and by which salespeople different products are sold.  When she needs to, she can also drill down into more granular information underlying financial statement line items.

Beyond such fairly traditional “dashboard” type information, Tan is working with other senior managers to set key performance indicators (KPIs) for use in aligning employee motivations with company strategy. To confirm that the KPIs are leading the company where it wants to go, the executives will analyze the correlations between the success of certain salespeople and the performance indicators.

In one recent example, to find out whether the firm was using its people wisely, Tan used data analytics to measure why Trax’s website experienced certain peak and trough periods. She first found correlations among traffic, time of day and location. With Australia about three hours later than Asia, the firm’s site traffic builds to a peak in Asia’s morning and Australia’s afternoon. The trough occurs during Asia’s afternoon, as Australia knocks off from work.

The question she was attempting to answer was: “Does it makes sense to hire additional headcount to handle the peak, but leave them idle during the trough?” Employing what she calls her “cause and effect analysis” of time zones and traffic data, respectively, she “suggested a practical solution that requires only use of our existing headcount. Basically, it only requires overlapping of two different shifts during the peak hours.” The analysis enabled Trax to trim its headcount costs and capital spending, she said.

From her perch in a firm that uses data analytics on day-to-day basis, Tan waxes bullish on the ability of the accounting and finance functions to expand their roles to encompass huge amounts of unstructured data. “In three to five years’ time, I see the finance function as a strong, formidable strategic partner to operations and sales. Finance is the natural gatekeeper of data, as information normally flows through the function,” she says.

Accountants still need to become more familiar with statistics and decision science in order to grasp how Big Data can enhance their skills and vice versa, she thinks. Ultimately, however, they can play a part in enabling companies “to separate the wheat from the chaff and focus on the information that counts — not on the information overload,” Tan adds.

Photo: Trax Technology Solutions

, , , , , , , , ,

17 responses to “Accounting’s Big Data Problem”

  1. The only big data problem in accounting is accuracy, verification, and reliability. The other six general accounting principles are damn important, too, but if you’re pondering another best method to acquire or receive data, you should also consider more efficient and optimal methods of verifying data from every source that reports financial, economic, and accounting data for collation or compilation. Social media and other forms of customer feedback are like a temporal suggestion box of ideas which may or may not solve any real world problems, especially in accounting like estimation, probability, disclosure, representation, theoretical soundness, logicalness, correctness, and timeliness, e.g. THESE ARE REALLY BIG DATA PROBLEMS AND THE GREATER THE NUMBERS …. BILLIONS AND BILLIONS …. THE MORE COMPLEX AND CHALLENGING In the near future, I also predict that accounting and financial data will be made available through multi-sources, let’s just hope that the data and information will be useful and reliable

  2. Thanks Michael Hemingway for your reality check. There is substance in this article, but its relevance to accounting is questionable. Nothing on twitter or facebook is part of the accounting records now, nor is it likely to become part of the accounting records. Social trends matter in business. But they don’t show up in the financial statements.

    • Considering how much of what’s on Facebook and Twitter is rumor, innuendo, and opinion. I’d be really skeptical about using ANY social media source as a point of reference of information in any financial statement package I’d publish to any internal or external user.

    • Questionable? I respectfully disagree. Perhaps the connection is not clar from where you sit. The topic of Big Data and Accounting Information Systems could not be any more intimately related.

  3. Interesting article and comments. Would like to add one a thought with regard to “big data” in the accounting space. The examples used in the article focused more on operational aspects of running the business not the accounting and or external reporting aspect of the business.

    Having worked at building more analytical capabilities into corporate reporting systems, I find this data far more useful in operational, sales & marketing and risk assessment functions not Accounting. Accounting has been and will continue to be backward looking in that accounting records what has happened. These other functions are interested in not just what happened by what is going to happen.

    As Jeff stated, Big Data’s relevance to accounting is in fact questionable.

  4. In the 1960s (or thereabouts), accountants disregarded their role as information managers and ceded control of computer information systems to a new breed of professionals. Their argument was that computer programming is not accounting; therefore, accountants don’t need to know anything about computer programming and related data management functions. It took the profession 20-30 years to catch up. Now, those myopically focused accountants are at it again, arguing that if it doesn’t show up in the financial statements, it’s not accounting. Alternatively, if one views accounting through the lens of a modern definition of accounting as “a profession that involves partnering in management decision making, devising planning and performance management systems, and providing expertise in financial reporting and control to assist management in the formulation and implementation of an organization’s strategy,” then the insights that can be provided by the crafty application of data analytics are most assuredly within the realm of accounting-relevance. The relationships “hidden” in unstructured data can certainly provide insights related to the development, interpretation, and even content of structured accounting data. Don’t let the bean-counters among us talk the rest of us accountants into ignoring the future of our profession.

    • I think that these analysis techniques will be much more fruitful in the internal management accounting function, rather than in the financial accounting & reporting function. I can certainly see the boon to a company’s operations that Big Data techniques can provide, and I think that utilizing these techniques for internal reporting is a great idea. However, I would be very, very hesitant to attempt to integrate unstructured data and non-standard sources of financial insight into external reports.

      Ultimately, I think that Finance, Operations, Sales, and Risk functions could strongly benefit from effectively utilizing unstructured data, as well as the internal management accounting systems of corporations. External reporting, however, would probably be best served to continue with a “business as usual” attitude, using the insights gained from big data as a source of (possible) sustainable competitive advantage.

    • Bravo! Reading the response comments here from accountants makes me gasp! I am amazed that my accounting colleagues, highly educated professionals, go through their professional lives with their heads in the sand. Amazing and puzzling.

  5. It is difficult to imagine how Big Data may be employed in the operating aspects of business and frequently tax laws govern depreciation expenses. Asset valuations are a far different matter, however. Assessment of the value of brands, intellectual capital and unmined mineral reserves,for example, may well benefit from the use of Big Data.

  6. I agree that it’s hard to see the value of Big Data in traditional accounting.
    On the other hand, Big Data could be very useful to assess the value of “softer” assets: for instance the value of the company’s customer portfolio could be measured as a combination of number of customers + their likelyhood to buy again + their likelyhood to convince somebody else to buy etc. etc., all info that, at least in some industries, now can be measured quite well. This may not impact the traditional P&L, but it would certainly impact the market value of a Company, and accountants would be very short-sighted in excluding this kind of analysis from their profession.

  7. There are already aspect of the traditional accounting practice that require indeepth analytics. Accounting estimates could
    be strengthen by tapping into the power of bigdata. An example is depreciation rate which are often determined without no real-life relation
    to the operation of a company, the machine work hours, etc. Analytics using real life data collected from a machine would enable accountants to review accounting policies line with company’s reality and will strengthen external reporting in this area which focus in fair presentation that is really fair to the circumstance of a company’s operations..

Leave a Reply

Your email address will not be published. Required fields are marked *