Machine-assisted corporate performance management: great potential and poised for takeoff

In the summer of 2016, we examined the application of artificial intelligence (AI) to corporate performance management (CPM), highlighting the potential role that machine-learning algorithms could play in the planning, forecasting and financial analysis process. In this follow-up report, we take a look at the current state of play and future outlook for machine-assisted CPM.

Artificial intelligence and performance management are a good union, but we have yet to see it blossom. However, the business conditions are right and the underlying technology foundation is in place. It is now a matter of vendor execution. While there are myriad benefits to be had from injecting machine learning into CPM processes, we would recommend caution. The more decisions that are automated using machine learning, the higher the risk of finance leaders running afoul of laws and regulations, and having other unintended consequences arise from letting learning algorithms make decisions. Therefore, CFOs should understand how the algorithms are operating and what decisions they are making – and the ramifications of them – to avoid this situation.

Early entrants

Last year, the beginnings of a shift in performance management started to occur as vendors began to embrace artificial intelligence within their offerings. Microsoft-based performance management player deFacto Global was an early player to take advantage of AI. The company integrated Microsoft's Azure Machine Learning framework into its offering to provide an alternative to manually derived forecasts, making forecasting more predictive and automated via the use of machine learning.

MindBridge Analytics has also entered the machine-driven performance management sector. The startup is applying AI to one aspect of CPM: auditing. MindBridge's offering essentially uses machine-learning algorithms to detect anomalous patterns of activities, unintentional errors and intentional misstatements that can occur in the auditing process.

However, we have yet to see AI-driven performance management offerings proliferate, although we expect that a significant pickup in vendor activity is coming. This expectation is based on business considerations as well as technical factors. It is also underpinned by the reported direction behind certain vendors' product strategies – but it is not meant to be exhaustive.

CPM + AI: the drivers

When it comes to business drivers, speed is a key one. Companies are under escalating pressure to adapt very quickly to volatile market conditions. Finance teams therefore need to be able to gather data from multiple sources and apply analytics in a more real-time manner than ever before. Moreover, finance professionals also face increasing pressure to shave days off planning, budgeting and forecasting processes.

AI-driven performance management could help by automating some of the drudgery, such as matching invoices and purchase orders together. Furthermore, it could aid CFOs with decision-making by enabling them to glean forward-looking indicators, which could help them uncover new opportunities faster. Additionally, it could help finance professionals address issues more proactively, enabling them to detect unusual expense items recorded by an individual, for example, before they become a problem.

The promise of machine-assisted CPM is likely to be realized by the acceptance of the cloud as an environment for planning, forecasting, budgeting, and more. Learning algorithms live or die depending on the amount of data they are fed. Cloud CPM is therefore ideally placed to handle one of the biggest obstacles to successful machine learning – feeding the algorithms with sufficiently large volumes of data to make them learn successfully.

A few years ago, cloud CPM was an evangelical sale. CFOs were worried about placing highly sensitive financial information into a cloud platform. Many regarded it as a risk not worth taking. However, cloud CPM is now mainstream and trust in using a SaaS delivery model is no longer an issue. As a consequence, vendor activity is likely to gather pace, as just about every purveyor of performance management now provides a cloud service.

Upcoming activity

Oracle is an existing performance management veteran that we anticipate will move into machine-assisted CPM. Why? The company is already delivering applications of this description for other business arenas such as customer experience management as part of its Adaptive Intelligent Apps strategy. Infusing AI into its performance management offerings in the Oracle Cloud seems like a logical next step.

SAP, which has also adopted an application-driven approach to AI, is another member of the performance management establishment likely to move deeper into machine-assisted CPM. Its Analytics Cloud already features some predictive machine-learning functions. But SAP is looking to bring to market others as AI is a high priority for the company.

BOARD International, which offers an all-in-one BI and CPM stack, is another firm that we expect to embrace machine learning as it is a roadmap item for the company. Additionally, we believe cloud CPM pure plays Adaptive Insights, Host Analytics, Anaplan and Vena Solutions will deliver machine-assisted planning and analysis capabilities at some point. These vendors have always used innovation as a growth driver, and for competitive differentiation. With proven success in this product strategy behind them, they seem likely to innovate in other areas, including machine learning.

Lastly, we think SaaS-based enterprise application specialist Workday could be ripe for making a machine-assisted CPM play. The company has already noted that it will focus on machine learning to help improve its HR-oriented offerings. Infusing learning algorithms into Workday's financial performance management service, which is another lynchpin in its cloud portfolio, would also make sense.