Introduction

In today's rapidly evolving business landscape, organisations are increasingly turning to Artificial Intelligence (AI), Machine Learning (ML) and Generative AI to gain a competitive advantage. For FP&A professionals, the convergence of AI with human intelligence is essential to address complex financial challenges and drive strategic decision-making. In this blog, we explore how AI/ML, Generative AI and human ingenuity can complement each other to transform FP&A into a strategic partner.

As covered in our previous blogs, AI/ML is set to democratise planning. Generative AI — the ability of systems to understand and respond to natural language inquiries — will transform how end-users can take advantage of AI, thereby drastically changing the involvement of FP&A. Traditional tasks such as data management and reporting will give way to new multi-functional roles such as FP&A Data Scientist.

"By 2030, 38.6 million potential FTEs will be displaced by automation. Storytelling with data will be a skill differentiator for the FP&A professional of the future."

Saurabh Jain, CFO Endovascular Robotics, Siemens Healthineers, New York FP&A Board Member

While AI/ML can analyse exponential data volumes to perform daily tasks, human intelligence remains crucial for directing algorithms and interpreting the results.

Where AI/ML Excels

• Detailed Analysis

AI/ML thrives in analysing details, where it can deduce patterns, trends, and make predictions at a granular level (e.g. by product line or customer), provided the data is ‘clean’. This means that each dimension of a product, service or customer can have a unique forecast based on the business environment in which it exists. This is a far more accurate method to produce fact-based predictions that go beyond the capabilities of spreadsheets and older analytical tools.

• Repeatability

Like any machine, AI/ML processes are repeatable and not subject to human ‘forgetfulness’ and bias. This eliminates most manual-based errors and subconscious preferences, thereby improving the reliability of results.

• Speed

AI/ML works at the speed of light, 24 hours a day. It does not get tired and never takes breaks. The algorithms can analyse huge quantities of data, determine which model produces the best results, and run multiple scenarios in almost real-time. This lowers risk by delivering insights that allow organisations to take advantage of new market developments immediately.

Where Human Intelligence Excels

• Intuition

Humans are intuitive and great at interpreting non-digital inputs, such as body language and ‘reading between the lines.’ Often, relevant data for an AI/ML algorithm is simply not available. For example, if the government suddenly changes business rules or when a black swan event such as a pandemic occurs. These events could change business performance in a way that is not predictable from historical data.

• Defining Data Requirements

AI requires data, but its algorithms won’t necessarily know what data exists and whether it’s reliable or relevant. A human is needed to decide what data to collect, where to get it, how to ensure its accuracy, and how to count it. AI can help the process, but it needs human guidance and judgment.

• Guiding AI Analyses

AI/ML has many algorithms, but each can produce different results for the same datasets. As a result, a human is needed to set a starting point in the datasets to include, to know what results make sense and to identify the results that are purely coincidental. There are many examples of AI-based predictions and connections that have been spectacularly wrong. The system effectively relies on humans to ensure that the results make sense.

• Making Strategic Decisions

Not all decisions will have measurable results, particularly when an organisation’s goal is philanthropic. There may be conflicting priorities between the achievement of short or long-term objectives and resource availability. In these circumstances, a human is needed to make a judgment call to determine which insights take priority.

Figure 1
 

The Strength of Combination: AI and Human Intelligence Together

In a rapidly changing world filled with unpredictable black swan events, historical data may not reliably predict the future. The International FP&A Board advocates for data-informed decisions rather than purely data-driven ones. If FP&A professionals can combine data with intuition, business acumen, and vision, they can deliver more meaningful insights.

As we mentioned earlier, FP&A needs to secure additional skills to benefit from AI. This includes the role of an FP&A Data Scientist who can use and direct AI algorithms, as well as interpret outputs. It is important to remember that AI does not always get things right. Humans are needed to verify that predictions make sense. A second crucial role in FP&A is the storyteller, who clearly communicates the meaning of the results and the necessary decisions to the business.
 

• FP&A Data Scientists:

FP&A Data Scientists will be responsible for selecting and refining AI models, ensuring data inputs are clean, and interpreting outputs. They will guide AI/ML algorithms to deliver meaningful and accurate insights.

• FP&A Storytellers:

FP&A Storytellers will play a pivotal role in translating technical findings into compelling narratives. They bridge the gap between data and decision-makers by using their data visualisation and communication skills to present strategic insights.

Case Study: Product Managers and AI-Powered Rolling Forecasts

TriLink Biotechnologies, a US-based biotech innovator, enhanced its product managers' forecasts by implementing machine learning-powered rolling forecasts. By using ML in conjunction with driver-based models through the intelligent planning platform, product managers were able to rely on automated forecasts instead of manual spreadsheets. This allowed them to focus on strategic analysis rather than data processing, leading to improved data and reduced manual effort. The unified platform provided a single source of truth across operations, finance, and sales, which enabled more accurate forecasting and strategic decision-making.

Challenges:

• Data quality issues: Data needs to be accurate and consistent.
• Over-reliance on AI Models: Avoid blindly trusting AI outputs.
• Communication barriers: There is a knowledge gap between data scientists and decision-makers.

Solutions:

• Data quality management: Implement data governance frameworks to ensure data is accurate and consistent.
• Balanced decision-making: Use AI to inform rather than dictate decisions and incorporate human judgment for strategic calls.
• Facilitate collaboration: Bridge the communication gap between data scientists and decision-makers through cross-functional workshops.

Conclusion

In summary, AI increases FP&A’s value as a business partner as it enables a better understanding of the key business drivers and activates new conversations and insights that were not possible before. Nonetheless, AI is not the master; it is an assistant. While it is good at discovering underlying trends, its results and suggested actions are enhanced and rationalised by human judgment.

Key Takeaways

• AI/ML offers granular analysis, consistency, and speed.
• Human intelligence brings intuition, strategic direction, and data curation.
• Combining AI and human expertise leads to data-informed decisions and effective storytelling.

Recommendations

1. Identify Key Business Drivers: FP&A teams should start by identifying key business metrics where AI can add value.

2. Invest in Data Literacy and Storytelling Skills: Strengthen data literacy and storytelling skills within the FP&A team to unlock the full potential of AI.

3. Implement Collaborative Tools: Use collaborative tools to encourage teamwork between data scientists, storytellers and decision-makers.

4. Adopt Intelligent Planning Platforms: Leverage platforms that offer intelligent planning capabilities to unify data and improve forecasting accuracy.

This article was first published on blog.board.com