Technology Oracle

5 First Steps for Adopting Machine Learning in Finance Organizations


Sponsored by Oracle

Discover how machine learning is influencing finance operations to help drive strategy and identify business opportunities, beginning with these five first steps.

©Danor_a/ISTOCK/THINKSTOCK

For any executive who pursued a traditional path into the finance organization—business school, perhaps an accounting degree or an MBA in finance—the incessant media chatter about how machine learning will transform your business must be a little disconcerting. Most finance executives have little to no experience with the subject. And while many now have an idea of what machine learning is, few have a concrete understanding of how or where they can implement it.

And yet, machine learning’s influence on finance operations will inevitably grow, accelerating the long transition of the finance function from accounting steward and risk manager to valued business partner who helps drive strategy and identify and exploit business opportunities.

For finance leaders who wish to be part of this process, the way forward involves five key steps:

1. Developing a working understanding of what machine learning is

A branch of artificial intelligence, machine learning refers to the use of algorithms, or computing models, that allow computers to become better at performing a given task through experience rather than programming.

Machine learning is already at work behind the scenes in many everyday applications, from spam filters on your email inbox to the shopping recommendations that pop up while you’re surfing the Web. And now, it’s becoming more popular in finance organizations.

2. Understanding how machine learning can be applied to finance activities

Finance organizations typically find it easiest to take advantage of machine learning in areas that involve repetitive tasks and minimal inputs—classifying business expenses for accounting purposes, for example, or reconciling accounts. As they gain experience, finance organizations can begin to use machine learning to tackle higher-value activities such as fraud detection, forecasting, and, ultimately, strategic planning.

3. Training or acquiring employees with the knowledge and skills necessary to embed machine learning capabilities into their organization’s activities

CFOs don’t need to be experts in building machine learning algorithms, but they do need to build finance organizations with the skill sets needed to take advantage of machine learning. This means employing people who have some understanding of statistics and skills around data science and predictive analytics.

4. Identifying and planning for opportunities to solve problems utilizing machine learning, beginning with simple, quick-win projects 

To make the most of their machine learning initiatives, companies need to first think about the problems they’re trying to solve, why they need to be solved, what benefits will be realized by solving them, and what capabilities their solutions will enable. Companies should also take care to ensure that they’re using a good quality data set, since it’s what will be used to train the machine-learning algorithm.

5. Developing an internal culture that understands and embraces what machine learning can do

Finance organizations should take the time to educate employees about what’s being done and help them understand that machine learning is aimed not at replacing them but at freeing them from repetitive, lower-value tasks—empowering them to engage in higher-value activities and analysis.

Creating this sort of culture is important because companies don’t want employees to feel alienated by, and push back against, machine learning initiatives.

Machine learning algorithms are already being used by many organizations, and will only become more commonplace with time. Finance organizations need to be part of that process. To succeed, finance leaders need to be thoughtful and methodical about their approach to adoption.

This article has been adapted from a longer, more in-depth look at machine learning and finance organizations. Register here to read the full text.