The role of finance departments is no longer confined to producing accounting statements; they must also analyze, interpret and clarify strategic decisions, and support business units in their operational management. Management control plays a central role in this mission: budget construction, variance monitoring, reporting, scenario modeling... all activities that mobilize large volumes of information and determine the quality of overall management.
This is precisely where artificial intelligence (AI) is gradually establishing itself as an essential ally.
From skepticism to the first concrete use cases
Just 2 years ago, AI remained a distant notion for the majority of controllers. Less than 10% put it at the top of their list of concerns (PwC, 2023). However, half of them already envisaged it becoming central within three years. In the meantime, a great deal of experimentation has taken place: Proof of Concept (POC), recruitment of data scientists, exploration of predictive models based on machine learning or deep learning.
These experiments have produced contrasting results, but they have enabled an essential step to be taken: acculturation. Finance departments now have a better grasp of AI concepts, from machine learning to generative models, and understand the crucial importance of data quality in producing convincing results. They are now looking for use cases that are truly industrializable and deliver rapid returns on investment.
Towards an integrated, sustainable offering
As was observed when regulatory non-financial reporting was first introduced, the market initially saw the emergence of a multitude of specialized solutions, each meeting a specific need. The complexity and cost of this dispersed approach soon prompted companies to turn to integrated solutions.
Planning software publishers have grasped this trend: they are now enriching their platforms with AI functionalities, capable of interfacing directly with existing environments. The challenge is therefore no longer to test isolated tools, but to identify the relevant, easy-to-deploy components of the planning platform that will really improve the performance of finance departments.
Major application areas for AI in EPM
Today, the use of AI in financial performance management is organized around three main dynamics.
The first is augmented analytics. It aims to automate repetitive and time-consuming tasks: generating reports, analyzing variances between forecasts and actual, detecting anomalies... These applications, enabled by AI, free up time for higher value-added analyses. Progressively, more advanced functionalities are emerging, such as "intelligent prompts", which will soon enable financial databases to be queried via a conversational agent, in the manner of a specialized ChatGPT.
Augmented forecasting represents the second dynamic. This involves projecting future activity based on historical data, enriched by internal or external data. AI reinforces this capacity with sales forecasting models, automated "what if " scenarios and early detection of anomalies. While these models currently serve primarily as points of comparison with human forecasts, they should eventually accelerate budget construction, reducing a process that can sometimes take several months.
Last but not least, intelligent modeling opens up a still emerging field. It makes it possible to simulate different economic or organizational models and anticipate their effects on business, with the promise of value based on a limited need for technical skills. In the short term, the benefits envisaged are lower development and maintenance costs, more resilient and robust models, and more effective quality control.
But above all, it opens the way to more agile management, capable of anticipating the impact of strategic decisions on the whole company, and eliminating the number of intermediaries.
Towards a renewed role for the management controller
If AI is attracting so much interest, it's because it goes to the very heart of management control's mission: transforming data into performance drivers. But far from replacing the human element, AI reinforces its role.
When faced with a predictive model, the management controller must be able to understand its mechanisms, assess its relevance and challenge the results. His value lies in his ability to combine the power of algorithms with a detailed knowledge of the company's context.
A gradual but inevitable revolution
AI applied to performance management is not a sudden breakthrough. Its adoption will be gradual, in step with the maturity of organizations and the consolidation of technological offerings. But the direction is clear: tomorrow, finance departments will have at their disposal tools capable of making data more reliable, automating a significant proportion of processes and bringing new depth to management scenarios.
One thing is certain: the finance departments that seize on these technologies to strengthen their steering role will gain a decisive competitive advantage.
Yoni Cadosch
Director - Finance Transformation & Performance
Micropole, a Talan company
