Artificial intelligence (AI) is in a phase of accelerated growth and the full gamut of its business applications has not yet been fully explored; however, AI may hold the potential to fundamentally change finance departments.
AI is the use of "thinking" machines to perform cognitive functions previously only able to be undertaken by humans. Its close cousin, machine learning (ML) is a form of AI that uses machines to crunch data to produce insights from it.
Human decision making is defined and limited by the amount of information a human is capable of understanding. Because AI can analyse substantially more data than humans, businesses utilising AI can make more informed business decisions.
The algorithm that is at the heart of an AI program can incorporate a wide range of internal and external data sources, including news, market and pricing data. With increased data analysis, finance departments can use more accurate, data-driven insights to inform business decisions.
Mark van Rijmenam is currently completing a PhD in AI at the UTS Business School in Sydney. He says many CFOs have already started to work with big data and AI is an extension of this. Big data is the analysis of extremely large amounts of data, usually millions if not billions of lines, to identify patterns or trends.
He says CFOs who wish to grasp the potential of AI for their organisations need to start experimenting now.
“Gain the knowledge to understand what it means, which will differ from business to business and sector to sector. Now is the time to build connections with big data scientists and machine learning specialists; CFOs can choose to either employ them or draw on their knowledge as consultants," he advises.
Van Rijmenam says there could be multiple benefits for businesses that embrace AI. These include a more efficient business, lower costs, fewer employees, better products, better customer service, lower risk and reduced fraud.
According to van Rijmenam, thanks to AI, finance departments are likely to be unrecognisable in five or ten years' time from how they look today.
“It will more than likely result in a lot more automation in finance departments and across businesses. Many more manual tasks will be automated. Also it will be possible to produce up-to-date balance sheets in an instant," he says.
Businesses can use this analysis to create products that better fit customers' needs at a lower production cost, potentially improving margins for businesses.
Van Rijmenam anticipates it will be easier for finance departments to understand what is going on in a business in real time, as well as being cheaper to monitor cash flow, and take necessary remedial action should the business experience cash flow problems.
The first step for CFOs when it comes to applying AI to their organisation may be to identify the organisation's most pressing challenges, and pick off the low-hanging fruit.
An example might be adding chat bot functionality to the organisation's website. Chat bots use a form of AI known as natural language processing, which allows a machine to interpret simple questions posed by humans.
“Start with problems that do not require a high level of skill to resolve and gradually build up from there. If you start with something ambitious related to a major part of your intellectual property and the AI fails you, you're going to potentially be facing very serious problems," says Tim Lea, CEO of technology business Veredictum.io.
For example, using insights from data to completely change a product that is already selling well; the risk is that change could destroy the existing market for the product.
Lea advises using the minimal viable product approach to introducing new tools that use AI in business. This is a technique through which a business develops a working prototype to test with consumers, then improves the product over time based on customer feedback. The advantage is the ability to find out if there is an addressable market for the product.
So using the example above, a business might develop a workable chat bot and then test it with a small group of users or even staff, known as beta testing, to iron out any bugs before a wider release. Says Lea: “Recognise it takes time to put new technologies in place. You can only really have effective AI if you have access to huge amounts of data."
Lea says it's also essential to have contact with the right talent; there's a massive shortage of big data specialists and AI professionals are incredibly rare as well. So businesses should be building relationships with consultants and universities now, to be able to access the right talent pool when AI becomes a commercial reality for their operations.
He says businesses that are able to do this should be able to identify problems earlier, develop better business models and generate better results. “Finance functions will be far more efficient and effective – it's not just about replacing roles in the finance team. AI is more about getting access to a lot more information and giving the customer a better end result."
“As it matures AI should lead to far greater returns because of this access to better information. For example if you're buying oil, access to more information about the direction of the oil price, supply levels and future demand will automatically lead to better buying decisions."
Ultimately AI may make the chief financial officer's role more interesting and rewarding, in addition to delivering improved business outcomes.
- Allow businesses to make better use of data for enhanced business decisions
- Lead to a cheaper, more streamlined finance function
- Allow for less expensive product development
- Produce better customer outcomes