Cognitive computing refers to the ability of IT systems to understand, learn, reason and interact. As well as being able to make sense of unstructured data (such as the imagery, natural language and even sounds found in emails, tweets, journals, blogs, images, sound and videos), they can extract ideas, learn from outcomes and interact with humans.
Data analytics provides useful insights, but cognitive systems can turn those insights into knowledge, explains Linnet Sen, performance management leader at IBM.
“Cognitive systems are probabilistic; in other words they learn systematically, are not dependent on rules and can handle disparate and varied data,” she says. “They unlock meaning because they can reason and they learn continually, helping to hone human expertise so people can immediately make more informed actions. They also interact with users in a natural way, removing some of the barriers between humans and machines.”
In a financial planning and analysis context, these systems can be used to increase forecasting accuracy, investment optimisation and revenue growth by observing how market forces interact. The result is a forecasting mechanism that can help CFOs identify hidden opportunities and risks to the business.
Sen cites the example of a large pharmaceutical company that used machine learning to get a more accurate understanding of how market forces affect its market share, enabling it to predict the impact of competitive events and undertake complex scenario planning.
Another example of how finance teams have benefitted from cognitive computing is a multinational consumer goods company that was forced to shed millions of dollars of operating costs following a takeover approach, says IBM global finance leader, Devesh Nayel.
“As part of the process of cutting supply chain costs, the company used logistical and algorithmic modelling to identify where it could reduce the cost of getting its products from the point of manufacture to the point of sale,” he explains.
IBM has also implemented a blockchain solution for the company that will reduce the cost of processing accounts payable invoices.
By creating a common platform and using smart contracts (programmes that control the transfer of money or assets between parties) to settle invoices, companies can create a completely transparent process that reduces opportunities for customer disputes. It is estimated that using blockchain or shared ledger technology for invoice processing could enable companies to reduce the cost of processing invoices by as much as 80%.
According to Shelley Davies, IBM's global offerings lead – accounts payable and cloud ERPs, blockchain is the next disruptor of financial operations. “We believe our work with one major consumer packaged goods company will redefine the benchmark for cost per invoice. Many of the activities that take place in shared services centres relate to reconciling activities between the buyer and supplier’s books of record. In a blockchain environment, with its shared ledger providing a single version of the truth, these inconsistencies disappear.”
Another benefit for finance teams is that suppliers will not have to implement new technology. Their invoices would simply be processed in a different way.
Nayel observes it is no longer enough for a CFO to run their business. They have to be change agents to modernise it, and digital technology can help. In this age of disruptive technology, companies are encouraged to embrace the technologies and opportunities on offer, otherwise they risk being disadvantaged by lagging behind competitors. This is also where companies can improve ecosystems by partnering with trusted advisers able to help navigate a path to cognitive computing.