5 Reasons Why Finance Leaders Are Considering Machine Learning


The role of a Finance leader is changing rapidly with the introduction and democratization of machine learning. The reason for this is that Finance departments are sitting on incredible data assets – vast amounts of high value data that when combined with machine learning lead to incredible insights, critical strategic thinking and decision automation for the company. Bringing that data together to create high impact predictive models, what-if decision support models, and more accurate forecasting models is no longer a nice to have, it's increasingly become a necessity made possible with machine learning.

Organizations that look to drive decisions off of data and embrace a data driven culture are turning more and more to their Finance counterparts for help and answers. Finance plays a pivotal role in connecting the broader business to value in the data.

Product groups, online commerce providers, search engine providers and even the United States Postal Service have used this technology for many years. It's not a new technology. The idea of using it in the back-office however has recently crept into mainstream thinking. Why? Because the cost and complexity of using this technology have come way down. So as Finance leaders are considering this technology, many are asking, what's in it for them?

1. Strategic instead of transactional

Gain trusted advisor status with your CFO, Sales & Marketing, Manufacturing and Supply Chain, Finance, HR, IT and Operations partners.

2. Objective and transparent budgets & forecasts

Remove many natural biases in the budgeting and forecasting process with transparent and data driven machine learning models.

3. Robust and real-time decision support

Quickly identify opportunities or flaws in critical initiatives and accelerate decisions with predictive modelling.

4. Optimized use of cash and capital

Add to the bottom line with improved use of cash and capital and make trade-off decisions faster with predictive modelling.

5. Early detection of compliance risks

Automate auditing activities and leverage machine learning to identify anomalies reducing risk exposure and business disruptions before they occur.

Breaking out of the "transactional" mode takes a conscientious shift in thinking. In speaking with Finance leaders across many leading Seattle area companies, some common themes surfaced when asked about their plans for machine learning. Most wanted to start exploring the use of machine learning but were struggling keeping up with the transactional demands of their day-to-day processes. At the same time, providing decision support to fast paced changes in the business has also been grabbing a lot of their attention. Others expressed concern about their data quality.

Machine learning creates more capacity for strategic thinking within a Finance org by helping streamline and accelerate decision support, forecasting, budgeting, auditing, compliance tasks and automating certain decisions. Benefits are realized quickly using data where it exists and deploying machine learning models within existing BI platforms or line of business Finance applications. By embedding machine learning within existing Finance applications not only do you accelerate the usage, you also make sure benefits are delivered to the right people, within the right process, at the right time - this is critical in getting the most out of a machine learning model.

data:image/gif;base64,R0lGODlhAQABAPABAP///wAAACH5BAEKAAAALAAAAAABAAEAAAICRAEAOw==

There are many practical and high value Finance scenarios where this technology can help. Models offering the highest value and impact depend on the underlying business and characteristics of the industry you're in. Here's a summary of some of the more commonly considered machine learning use cases across Finance:

1. Financial analysis & reporting

  • Intelligent pricing and improved profitability

  • Reduced reporting cycle times

  • Audit automation

  • Reduced customer churn

  • Intelligent expense management

  • Optimal use of labor

2. Forecasts and budgeting

  • Cash optimized to liquidity needs

  • Improved sales & revenue forecasts

  • Single demand planning models shared across Finance, Sales, Marketing, Ops

3. Decision support

  • Intelligent pricing & promotions

  • What-if impact modelling (transfer pricing, entity changes, IP transfers, legislation impacts, tax implications)

  • Accelerated investments decisions

  • Optimal budget prioritization

4. Management of cash and capital

  • Cash management

  • Predict bad debt

  • Forecast capital demands

  • Manage capital assets, predict when they need replacing

  • Automate cash application to invoices

  • Accelerate credit extensions

5. Risk and compliance

  • Predict internal disruptions

  • Detect fraud

  • Predict credit risks

  • Prevent money laundering

  • Detect potential company policy, expense and procurement violations

  • Prevent channel stuffing

While the technology is widely available and affordable, knowing how to apply it to create high value business outcomes requires experience. We have found the best way to deliver high value business outcomes with machine learning is to keep the team lean, involve the right people at the right time and let them collaborate with very little project "overhead". The process is best performed with a high quality data scientist, industry expertise in Finance systems, a data analyst and a passionate expert from the business. Assembling the right team and then getting out of way allows the "magic" to happen.

Whether you are in the beginning stages of a digital transformation or consider yourself well underway, putting off the use of machine learning technology in Finance only delays the benefits and some would argue delays the inevitable. The sooner you experience it for yourself the sooner you will become a believer.

Please let us know if you are interested in hearing more about how Machine Learning can benefit your Finance Department.

Matt Wilson

matt@winigent.com

http://winigent.com