With the push continuing for firms like yours to digitally transform and fully automate Finance processes, the term “AI” (artificial intelligence) gets mentioned a lot. But what is it exactly?
In simplest terms, it’s the technology that allows computers to learn how to analyze data, identify patterns and predict outcomes.
The goal is to make machines smart enough to interpret information and make decisions like humans do to take over repetitive and labor-intensive tasks. Examples include self-driving cars and natural language processing to review contracts.
AI and machine learning
A specific subset of AI used by Finance teams to take on reporting and data reconciliation is machine learning (ML). Using algorithms, ML allows a machine to teach itself based on historic financial transaction or operational data it receives. As the algorithms come up with predictions or insights from that data, and receive feedback on the outcomes, ML’s capabilities change and adapt.
In a Zoom interview, Mark Sheldon, the Chief Technology Officer of Order-to-Cash (OTC) SaaS platform Sidetrade, said that it calls their AI engine that analyzes B2B payment transactions to predict customer payment behavior and risk “Aimie.”
Aimie draws from what Sheldon described as a cloud-based, aggregated “data lake” based on the transaction activities of millions of businesses worldwide to recommend cash collection strategies and automate actions in the OTC process.
“So when an invoice gets loaded into our platform, we match the buyer into our network so we’re able to see (whether) this buyer already exists in the network. We know what their normal payment terms are, when they normally pay, (and on) what day of the week,” he said.
By automating tasks that used to require manual attention, some of the benefits of ML’s statistical approach include:
- improving the accuracy of the accrual process
- analyzing and interpreting expense data, and detecting suspicious expense claims
- segmenting key customer accounts with outstanding payments that need collection reminders
- predicting which customers are likely to make repeat purchases
- determining whether a potential customer is creditworthy
- predicting which suppliers are likely to default, and
- predicting when outstanding invoices are likely to be paid.
Having more predictive capability is highly valuable, especially with bad debt risk and bills coming due at higher prices due to inflation and rising interest rates.
ML’s self-learning means it gains value over time. But to get a short-term feel for if the return on investment (ROI) for the technology would be beneficial for your business, it’s best to focus on the desired outcome for a single problem you’re trying to solve, such as streamlining transaction processing.
Laying the foundation
To set AI tech up for success, it’s good to first fully evaluate your current operations to identify what needs to be updated and look at areas of your business where AI may already be in use.
You may need to assess and address any data flow issues that would affect ML integrity and accuracy. For instance, not having enough relevant buyer and seller data means you won’t get the full potential ROI from ML tech. Your Finance team will need a strategy for setting siloed data free.
But if completely breaking down silos is too disruptive to your company’s data governance, an alternative suggested by Deloitte is a pilot project that would maximize the data within a silo, such as classifying sales data for use by Finance and commercial teams.
In addition, keep in mind employees will probably need some training to manage ML systems to best leverage the tech’s capabilities. For example, managing issues from inconsistent data quality or human bias requires ongoing maintenance and quality control testing.
However, Sheldon said that even if your Finance team heavily relies on spreadsheets, the major SaaS providers in the marketplace are making entry into the ML world so easy that you might not have to depend on your IT team as much to consolidate data and deploy a solution.
Controlling the cost
The experts at Deloitte say that it’s OK to start with a point solution that’ll have a strong impact on the business, instead of investing in a more costly enterprise-wide solution. Consider if the specific problem is shared across other areas of your organization. Is the process you’re looking to streamline specific to Finance or is it a solution that could benefit other areas (e.g., invoice matching)?
In addition, it may be worth exploring creative funding sources that are out there for businesses, such as vendor subsidy programs, co-investment strategies or even a venture capital model.
You can also try huddling up with your CIO to consider partnering with a third-party AI provider, developing a solution in-house or a hybrid combination of both approaches.
Finance leaders will “get a competitive advantage from AI for sure,” said Sheldon. “The less you invest in this kind of technology, the more you will be left behind from your competitors.”