Whether it’s through machine learning, blockchain, or artificial intelligence, senior finance executives are increasingly employing brainy technologies to make their processes more efficient.
Among the finance functions getting their screws tightened are revenue forecasting and cross-border supply chain financing, executives say.
At Microsoft, for instance, assistant corporate controller Alice Jolla and her colleagues began using a machine-learning system hosted on the tech giant’s cloud-based Azure platform when it became clear that the process of forecasting revenue had become too cumbersome.
As a result of the inefficiencies, Microsoft embarked on automating the sales prediction process. That process is crucial because the forecasts affect items in the company’s quarterly earnings estimates and the company’s ability to figure out its effective tax rates, Jolla suggested at a press briefing at the Financial Executives International’s current financial reporting issues conference in New York.
“What we found prior to going down the automation path is that we have hundreds of individuals who are dispersed throughout the various business groups and geographical areas [whose estimates] were bubbling up through the forecast,” said Jolla, who’s responsible for Microsoft’s corporate accounting.
That made for an extremely labor-intensive and inflexible workflow. “If you had one change in [a] variable, it could take a day to cycle through that,” she said, adding that the new machine-learning technology can improve the quality of a forecast as well as curb cycle times.
Now, employees throughout Microsoft contribute to the process through centrally controlled business intelligence dashboards, according to Jolla.
A reporter asked Jolla if the automation would lead to layoffs in Microsoft’s finance and accounting functions. “We don’t think about it in terms of ‘will people be replaced? We think about it as ‘what skills will now be required in the future?’” the accounting executive answered.
Before the automation, finance had used many employees to perform such “low-value activities” as copying and pasting data “or moving information between one spreadsheet and another,” Jolla said.
Those are the kinds of tasks that are being replaced by machine learning. “It’s a different skill-set [that’s required],” she added. “But I don’t imagine that there will be a world where individuals are completely out of the process.”
Alluding to the words of another conference participant, Jolla said that it’s “where the analysts and the robots meet that becomes the important intersection in the finance [function].”
For its part, IBM is testing the use of blockchain technology to coordinate complex financing arrangements. The company is trying to boost participation in an automated ledger aimed at coordinating shipment information among suppliers and customers, with IBM acting as the financier, according to Keith Goodwin, a vice president of finance at the information technology giant.
In setting up a blockchain system within finance, the highest hurdle is coordinating the participants and extracting the needed data from them. “The challenge with blockchain is not the technology — the technology we can spin up in a few months. It’s [amassing] the data and the alignment with other organizations,” Goodwin said.
“When you’re dealing with a blockchain implementation, you’re dealing with multiple parties, because if you’re doing it right [that’s] the real value prop,” he added.
IBM’s blockchain pilot involves arrangements that span many countries. “We were having a large amount of disputes in that environment, and the number-one cause was just shipment information, not having transparency on the shipment status,” Goodwin said at the press session.
To deal with the crush of information, IBM constructed a “a distributed ledger of shipment information that is now visible and transparent to all the parties,” he explained.
As a result of the blockchain system, the participants were able to drop the number of dispute days from 44 to less than 10 and thereby generate extra cash, according to Goodwin. To be sure, IBM only has a handful of customers and suppliers involved in the effort.
The company has a goal of adding 4,000 participants to the ledger. “We’re going to accelerate that deployment through 2018,” the IBM finance executive said. “I don’t know if we’ll get to 4,000. But we’ll try.”
Photos courtesy of Twitter and LinkedIn