When Change Isn’t an Option but a Mandate: What Big Data Is Doing to Accounting
Brian Sommer | December 1, 2015 |
Change often creeps up on us slowly, and then it’s all over us. That’s happening in accounting circles today particularly as it pertains to big data. And, the changes it’s triggering will cause fundamental reassessments of what practitioners do and what accounting educators teach.
Big data has burrowed its way into virtually every aspect of accounting. Businesses are using it to do a better job of developing budgets, plans, and forecasts. It’s particularly effective in finding potential fraud. It also helps business operations reduce a number of costs and identify revenue generation opportunities.
Businesses are sitting on so much information now that approximately 43% of firms are selling their big data. But who is valuing such information? Who is making sure that the sale of such information doesn’t expose a company and its customers or suppliers to potential privacy issues, security risks and economic harm? Sadly, no one may be watching these matters.
What is big data? A quick answer might be something that won’t fit in a spreadsheet. But to be more precise, big data is found in a number of internal and external sources. It includes operational data from sensors, point of sale terminals, financial transactions and the Internet of Things (IoT) devices. It also includes external data such as weather, census, economic, and other information. There’s also social media, social sentiment and review data (for example, TripAdvisor or Yelp reviews) that can be particularly useful in assessing a product or service’s success or failure in the market. Vast amounts of digital breadcrumbs are generated by web and mobile users. That big data is instrumental in placing appropriate offers in front of prospective buyers. And, finally, there is ‘dark data’—the information companies already have in abundance (such as email data) but hardly use.
And, more data is coming. With homes getting ‘smarter’ every day, businesses are embedding Internet connectivity into televisions, watches, thermostats, and more. Industrial firms are adding smart sensors to equipment, vehicles, and engines. In just five years’ time, the number of IoT connected devices will mushroom from a few billion devices today to 25 billion devices. Each of these devices will have the ability to generate dozens to millions of pieces of data daily.
From the consumer side, people are generating vast amounts location, photo, and comment data every day. With the average smartphone user having over 100 apps, the amount of data being generated daily is astounding—and growing.
All of this data requires new analytical tools and new skills. Why? Well, not all of this data shares the same quality, size, and standards you might get with accounting transactions—yet businesses are using it to make key operational, pricing, and other decisions. You might be able to read a tweet and see it contains sarcasm, but could a software program determine that? Some product reviews may have been written by a non-existent person or someone hired to post positive content. Assessing the quality and usefulness of some big data will be a continuing challenge.
Adding to the challenge is that financial accounting technology is facilitating the use of these new data sources and types.
Newer financial systems have been designed quite differently from their constrained predecessors. The predecessor systems aren’t relevant in a world where disk storage, computer memory, and throughput are no longer expensive, time-consuming or constrained. Moore’s Law has converged with new technology capabilities to create financial accounting systems that can consume virtually unlimited amounts and kinds of data—not just basic accounting transaction data. The new software uses in-memory database technology and Hadoop, and billions of records can now be read, summed, and reported in nanoseconds.
What this means is that all of those prior systems just aren’t that relevant in a world where disk storage, computer memory and throughput are no longer expensive, time-consuming or constrained. Moore’s Law has converged with new technology capabilities to create a financial accounting environment that can incorporate and use more than just basic accounting transaction data. Big Data has been built-in to the new accounting software.
There are also new tools that supplement the new financial applications. There are data visualization tools, sophisticated fraud detection tools, continuous auditing software products, etc. There are tools to access massive amounts of social media data. Artificial Intelligence and Machine Learning applications have been deployed to handle the more mundane aspects of invoice presentation and transaction coding. These same smart technologies are using big data, algorithms, and pattern recognition to also spot potential fraud issues. The list goes on and on.
The issue for accountants and accounting educators is no longer “Is this big data thing real?” but “How do we get our heads around it?”
Implications for the Profession
The arrival of these new technologies and big data stores is not and will not be an incremental or easily digestible change for many. These new capabilities are disrupting the status quo everywhere—including education, business, and technology providers.
Within businesses, a number of firms are finding their current mix of old, loosely integrated technologies is wholly out-of-sync with the needs of the more modern, more digitally driven firm. Their current systems, which were designed for the Industrial Age, are a veritable dog’s breakfast of redundant, poorly interfaced software with too much data latency. Before these firms can utilize more big data, they’ll need to dramatically modernize their financial accounting infrastructure.
The skills required to make something from big data will be a challenge for many employers and educators alike. Businesses will need people with outstanding statistics knowledge as well as insights into algorithms, social sciences, and other disciplines. People will need to understand not just the dollars and cents behind a transaction, but also how all of the other data in and around the transaction came together.
How data is valued will become a hot topic, as will be the issues regarding the privacy, collection, retention, security, and sale of this information. Businesses will need executives capable of researching, creating, and enforcing guidelines where none previously existed.
And, every business will see big data impact virtually every process within the company. Big data feeds are already common inputs to sourcing, recruiting, marketing, and other processes. Does your firm understand what its future processes should like? Does it know where it will get the datasets for these modern processes?
Finally, big data will affect auditing in a couple of ways. For instance, the technologies that allow companies to parse vast amounts of data will also permit internal and external auditors to continuously review every economic transaction impacting a firm. Auditors may want to consider social sentiment data when assessing a company’s status as on-going entity. And, a number of fraud detection tools are using big data to spot actions such as insider trading, travel expense reimbursement fraud and more.
Big data is already driving change. Is your organization keeping pace or getting left behind?