The holidays are no holiday for first-year accountants.
While most of the working population is spending time with family and adding up presents, newly-minted auditors are in warehouses, storerooms, basements and random fields counting everything else: pens, computers, socks, lawnmowers, cattle. Physical inventory counts are crucial to a business’ bottom line, but are time-consuming and best left for days when transactions and deliveries don’t get in the way.
So make your list, and check it twice.
“New staff every year have to count the inventory at their clients and this usually happens during the last week of the year — that’s what you spend your time doing,” says Margaret Christ, an associate professor in the J.M. Tull School of Accounting. “They’ve gone from originally taking inventory on paper, then Excel spreadsheets and more recently firms use mobile tech and apps to record.”
But this past summer Christ worked on an analytics study at a Kansas feedlot finding a better way to count cows.
“We were able to get the internal audit papers of how long it took. Before they would audit only 10 percent of this feedlot of 32,000 cows, and it took them 40 hours including cowboy time (when cowboys have to corral the cows),” Christ says. “Using drones and counting software, which loaded your images and automatically counted, we counted the entire population in just six hours.”
They now have drones that fly though warehouses and do the same thing with office supplies, “and even make sure the boxes are full to assure they have what they say they have,” Christ adds.
You wouldn’t be alone in thinking “data analytics” was only about who buys striped ties or reads articles about Kim Kardashian (though it means that too). What analytics really means is scrutinizing data to help businesses augment productivity and become more efficient with their time and personnel. Analytics guides autonomous cars, detects bank fraud, and helps doctors track patient history to make better diagnoses.
It can even get first-year accountants home in time for grandma’s apple pie.
The purpose of analytics is solving problems, but the problem with analytics is people — people who believe in analytics, people who can gather data, and people who can then interpret what the data mean.
“We have oodles of data coming in,” says Rick Watson, the J. Rex Fuqua Distinguished Chair for Internet Strategy in Terry’s Management Information Systems Department. “But now the bottlenecks are with converting these data to information and the interpretation of the resulting information. The goal of business school is to give people the expertise to know what information they need to solve the problem and to then interpret it.”
And that’s where the Terry College of Business comes in. The college has responded to this dynamic trend by introducing new programs to complement its analytics programs already in place.
• The first class of Master of Science in Business Analytics students enrolled this fall. The one-year program teaches students technical expertise in collecting, analyzing, and interpreting big data while learning the field’s predominant programming languages.
• Tull introduced a data analytics initiative within the Master of Accountancy program in partnership with KPMG to prepare students for the digital marketplace. The coursework explores beyond traditional accounting principles and tax regulations to include technologies and methodologies used in today’s data-centric environment.
• The Management Department offers an area of emphasis in human resource analytics, where students study topics such as recruitment, selection, training and development, performance management, and compensation.
• Management Information Systems’ area of emphasis in data analytics for undergrads allows students to gain knowledge about how to leverage data — big and small — using unstructured data, query languages, statistical methods, visualization, predictive analytics, and data science techniques.
• The Master of Business and Technology program offers a degree addressing the gap between business strategy and technical know-how and preparing graduates to lead digital initiatives as business analysts.
• An MBA concentration in business analytics teaches core courses including emerging technologies, predictive analytics and data management.
Analytics, of course, is nothing new to the Terry College. It has long been part of Terry’s Master of Marketing Research Program, the first of its kind when it was created in 1979. Analytics courses in MMR focus on how to analyze and interpret data to guide marketing decisions at the strategic level as well as for tactical plans. And while business data may fluctuate from year to year, one thing is certain for MMR graduates — everyone gets a job.
“The MMR program was started to help train people to deal with marketing data, and you can see how analytics has changed marketing since then,” says Charlotte Mason, head of the Marketing Department and C. Herman and Mary Virginia Terry Chair of Business Administration. “For a long time there was a dearth of information, data were scarce and companies were really looking for it. But now so much is automatically captured with scanners, loyalty cards and online searches. If you look at marketing research classes, which are required for all marketing majors, it’s shifted from the design and collection of data to more the analysis of data, which is already out there.”
Marketing analytics — the art of predicting a consumer response — is a form of analytics most of us encounter on a daily basis. The ads we see on websites, the cookie displays at Kroger, the catalogs in our mailbox are all based on snippets of information companies know about us. What the companies don’t know (but want to find out) is the process of why we choose what we choose. It’s the question Terry professors ask their students to figure out.
“A huge issue in marketing is called attribution modeling — somebody goes in and they buy something. But what marketers want to know is what put them there,” says Mason. “Was it a banner ad, a billboard ad, a catalog, word of mouth, TV ads? Companies are trying to figure out these marketing communications and they want to create a model that says ‘OK, this is worth this much.’ Putting the data together is a big issue.”
It’s not enough for students to be prepared; the faculty teaching this data revolution need to be constantly learning too.
Analytics in accounting isn’t as well-known as it is in marketing or social media, but there’s no arguing its importance. Through audits, analytics can ferret out fraud on the company books, but also offer insights on what the numbers reveal, which can lead to future innovations.
For more than a year Christ and fellow Tull associate professor Tina Carpenter have investigated the use of analytics throughout the financial reporting process, conducting nearly 60 interviews with financial executives, public accountants, and standard-setters. They learned using analytics in audits is more widespread as well as a better way to validate financial reports. But challenges remain on several fronts.
“One of the first challenges for auditors is they’re still having to talk to their clients about getting all the data — some are all in but many are not,” Christ says. “We heard concerns that didn’t have anything to do with the analytics themselves but with cybersecurity. There is a challenge of changing that mindset, internally for companies, internally at audit firms, and between audit firms and companies. Someone described it as trying to turn the Titanic.”
“And,” Carpenter adds, “they know the iceberg is there.”
Another challenge for auditors is finding people with the skill-set to analyze the data. Vetting the numbers on audits isn’t the sticking point. It’s the critical thinking needed to determine why the numbers do what they do.
“What we hear from firms all the time is they need people who can look at a set of data, think about the question they’re trying to answer and figure out how to get that data to answer that question, maybe by even bringing in other data,” Christ says. “They want people with the ability to think like that and not solely focus on what they did in last year’s audit.”
Auditors with access to all the data and the personnel to decipher it found it improved the client-auditor relationship because it validated reports and sniffed out elements of fraud.
One person said data analytics makes the needles in the haystack shine brighter, referring to fraud,” Carpenter says. “The other thing is connecting the dots. They used to be able to connect one or two dots but were limited by their own imagination. But with more sophisticated data analytics, they can connect four and five dots and start to make significant progress on new ideas. There’s a story the data are trying to tell, but they can’t understand the story unless they understand the accounting.”
Through the insights they learned in interviews, Christ and Carpenter have transferred their findings from “The Data Analytics Transformation: Evidence from Auditors, CFOs, and Standard-Setters” to the classroom. The two have won awards and grants for their creative and innovative teaching, employing the patented technique of “show, don’t tell” to immerse students into situations they’ll face as professionals.
“In fraud class I do a simulation which enhances these problem-solving activities that definitely embrace data analytics,” Carpenter says. “They get evidence from one place and have to match it up with evidence from another place, then ask additional questions to keep probing the issues and looking for more people to help them solve the fraud. It’s a simulation that runs seven weeks out of the course.”
“What we are hearing from firms is a challenge,” Christ adds. “Make (students) more technical, make sure they know accounting, but let them be free-thinking creative artists. There does seem to be some shifting in public accounting firms where they are talking about innovation and trying to figure out how to reward and encourage it. It’s quite a big shift from when I was there years before when nobody was talking about that.”
The introduction of analytics has also changed the field of accounting.
“Accounting as it has been historically understood — the stereotypes about accounting — don’t hold true anymore,” Christ says. “It’s a different skill-set, and the work is exciting.”
Mason brings her unsolicited mail to class to show how a slight change in response rate percentage can affect the bottom line.
“I can’t tell you how many credit card offers I get, and I have never responded to a single one but they keep sending them,” she says. “The cost to send them is trivial, but if I bite the payout is actually pretty big.”
Direct mail and email, it turns out, is marketing analytics in its simplest form. Your information came to companies in a variety of ways, and once their letter is sent out, all they have to do is wait.
“There’s a famous expression — direct marketing is the only business where you can be wrong 99 percent of the time and still be a winner,” she says. “For credit card solicitations, 99 percent of the people throw out the offers without even opening them and yet the direct mail offers are still profitable for the firms. If firms can use analytics to increase response so that they’re wrong 98 percent of the time, they just doubled their money.”
The role of artificial intelligence in analytics cannot be understated, but its influence wavers from business to business. In accounting, drone analytics uses AI to take away guesswork, conducting activities more efficiently and accurately than people can. In marketing, AI focuses more on making people’s decisions easier, as opposed to lightening their workload.
“Think about the recommendation systems in Amazon or Netflix, that’s something a person can’t really do because the universe is just too big, and that’s AI,” Mason says. “And it gets smarter as Amazon or Netflix collects more information about you.”
But as companies seek to use AI to gauge more on what you want, a “creepy factor” has slid in. For a while putting radio-frequency identification tags in clothing was a hot topic — you walk into a store and the sensor reads the tag and knows your name and shopping history — “but it freaked people out,” Mason says.
Facial recognition, a staple in Steven Spielberg’s futuristic “Minority Report,” is used to open iPhones, check in bags at airports, and in Japan, buy sodas.
“There are vending machines that capture your image as you’re standing in front of the machine, and then make recommendations based on inferred age and gender,” Mason says. “So if you’re a middle-aged male one kind of drink might be recommended, but it would be a different one if the image is identified as a younger female. The algorithms were pretty accurate.”
There’s more to be done in sharpening the message without repelling the audience. Mason sees the influx of more and more qualified students entering analytics, using their background in statistics or computer science to improve the growing field of study.
“There’s a signal and a noise problem, and it’s far from perfect,” Mason says. “Things get lost in translation. It’s a little like telephone — every time the message gets transmitted it gets a little fuzzier. ... I have received ads for online advanced degrees that read ‘Dear Professor Mason, have you ever thought about getting your advanced degree in business?’ Not only did they send it to Professor Mason, they sent it to my school address. So sometimes it’s not so smart.”
About the autonomous car.
Rick Watson has been at the Terry College since before it was called the Terry College. He has written 10 books, including “Data Management: Foundations of Data Analytics” (now in its sixth edition), and has been part of numerous studies in the field. For him, cars, and forms of transportation in general, are the real future movers in AI analytics.
“The autonomous cars have huge implications for society — here’s a simple one, there are no longer any organ donations because you don’t have accidents,” Watson says. “There are no speeding fines, which has implications for municipal governments. The car is a tipping point in many ways and that comes about because you can collect so much data through these sensor networks and then analyze it.”
Cars already have automatic braking, and sensing devices that tell if you are too close to the car in front or if you’re going to hit something behind you. But “with an autonomous car you have huge amount of data coming in all of the time because what the car has to do is build a digital representation of the world around it and then drive through it, and it’s doing that yard by yard,” Watson says.
In his role as an MIS professor, Watson teaches his students to clean, clarify and produce knowledge from the massive amounts of data coming in to give companies the opportunities to ask “how am I going to change the way we think about the business?”
He pointed to two trends — digital data streams (DDS) and digital twins — that mark the future of analytics. In DDS, data are generated in real time based on the environment it sits in. Watson references a bus stop in London, where a flat panel display can react to rain in the area by telling people where to buy umbrellas, or acknowledge the local soccer match by pointing out pubs showing the game. “We’re moving to where every asset is online, streaming data about its status and when its status changes,” Watson says.
A digital twin is just that, a replica of the physical item that uses sensors to fix and address issues. “One example is if the owner of a Tesla had a steep driveway and he would scrape the bottom of the Tesla each time he would drive out,” Watson says. “He’d contact Tesla and say, ‘hey I’ve got this problem,’ and it would transmit a software patch that when the car is at a specified GPS location, the point where the driveway meets the curb, the software raises the suspension. The digital equivalent talks to the physical, and they interact the whole time.”
Digital twins are used in trains and jet engines, but could include people as well. A digital twin could monitor your heartbeat, measure glucose, track weight, and offer alerts to help promote a healthier lifestyle. The shared information could shrink the cost of health care.
“But there’s a fundamental problem with AI at present, it generally can’t tell you how it makes its decisions — the reasoning is not there,” Watson says. “That’s the next generation they’re working on.”
The present generation continues to show and prove to businesses how important analytics is. There is no doubt every industry is gathering data, but figuring a way to positively drive the data to enhance a company’s direction is the charge of today’s students.
Watson recalls a story about a mining company with a treasure trove of data but no internal capability to tell them where X marked the spot.
So they put out a competitive call.
“There’s a really great example in the gold industry where a company had all the data but didn’t know where the gold was, so they had a competition and offered something like $600,000 and gave the data out,” Watson says. “It was a company in West Australia near where I used to live who found where the gold was. The value of the company went up astronomically, and it was all from data analytics and visualization. Of course they got $600,000, and the company that started the competition got billions out of it.”
Everyone wins, but as it is with analytics, the numbers aren’t always the same.