Artificial intelligence in finance 101: how AI can direct better CPM outcomes

Examples of AI-powered CPM processes

AI improves finance’s decision-making and efficiency, but what exactly does that look like in practice? Here are a few concrete examples of how AI can enrich CPM processes.  

Data collection

If your company acquired another organization, you’d need to add a new G/L source file to the group’s financial close process. AI can accelerate the G/L’s data mapping, ensuring that the data’s integrity is retained as it enters the group’s financial close system. In addition to mapping the G/L to the financial close system, the AI could also map data for ESG, tax, and lease reporting.  

Anomaly detection

Let’s say you were loading February actuals for your New York entity. AI could instantly read through the actuals and detect outliers in the February file. It could create a report that lists the potential outliers and the reason the figure was triggered as a deviation. You’d then be able to quickly vet the outliers to determine whether they are incorrect.  

Analysis

It’s the beginning of Q2, and you need to create a plan for a product line in the EMEA. By analyzing the region’s data, the product line sales history, and market information, AI can determine the business drivers influencing sales so you can apply that insight to your sales plan and strategy for the coming quarter.  

Planning

What was the highest-performing marketing campaign in Q4 — and how can we make it even more impactful? AI can analyze demand, marketing, and sales data in context to determine the most successful marketing campaign and provide recommendations to maximize the impact of that campaign.  

Performance analysis

When leadership needs answers, they need them fast. Using Generative AI, you can ask your CPM a question, and it can generate an answer in the form of a direct response, report, or chart on a dashboard. For example, you could receive answers to questions like:  

• Why did our overhead for New York increase year over year?
• What was our highest profit center for 2023?
• What was the line of business with the highest US sales last year?  

The AI would instantly pull results from your performance data and organize it into a report that is ready for analysis.  

What to look for in AI for finance 

While there are many different approaches to AI, there are three AI capabilities finance teams should ensure their CPM solution includes. 

1. Explainable outcomes 

How did the AI arrive at that prediction? What’s the proof that the driver the AI identified is actually influencing a KPI? Why did the AI flag a certain data point as an outlier? And, most importantly, does the CPM software give you access to these answers?  

You have to be able to trust your AI. Many AI-powered systems will provide you with predictions— but not insight into the AI’s logic. We call this “black box AI,” where the AI spits out an output, forcing you to trust its logic without any transparency.  A survey by the International Data Corporation found that a quarter of AI projects fail due to factors that include the black box phenomenon and interpretability challenges. 

We recommend looking for a solution that uses a “glass box approach,” where the finance teams can lift the hood on the outputs via explainable AI.  

The study Explainable AI: From Blackbox to Glassbox defines explainable AI as, “the class of systems that provide visibility into how an AI system makes decisions and predictions and executes its actions. Explainable AI explains the rationale for the decision-making process, surfaces the strengths and weaknesses of the process, and provides a sense of how the system will behave in the future.”  

Explainable AI is essential for finance teams that will use the AI’s outputs to make decisions, execute strategy, and build budgets. When you can see why the AI came to its conclusion, you can more confidently proceed to use its results in your decision making.  

2. Human control 

Building on the explainability factor, AI should keep finance teams in charge. AI should be a powerful tool that makes undiscernible information discernible. It should not be an autonomous decision-maker. EY writes that ultimately, finance teams need to see AI as a collaboration where AI can do the repetitive work and finance teams can do the strategic work.   

“While AI can process vast amounts of data at a rapid pace, it lacks the critical thinking and decision-making capabilities of people. The ability to identify and address bias in data and core skills such as knowing the right questions to ask stakeholders to understand their objectives means the finance professional has a significant role to play in this technological transformation.” 

While keeping humans at the wheel is true on a strategic level, it’s also true on a functional level. When the AI generates a prediction, identifies an outlier, or creates a report, finance teams should be equipped with the information to vet and check the AI’s work to ensure that the AI’s logic jives with the finance’s real-world understanding of the data.  

3. Data governance 

Data governance is a constant challenge for finance teams dealing with an influx of new requirements, including BEPS Pillar Two, ESG, and lease accounting. We recently wrote about how the scope of financial close and consolidation has expanded because of the growing data volume, data types, and reporting requirements. Mapping and formatting data across different sources so it’s apples to apples is a hefty task for finance teams to manage by hand. AI can ensure that data is streamlined and controlled.  

Acceleration Economy explains, “Today’s governance policies may call for a human to scan petabytes of this unstructured data, which would take years and be cost-prohibitive. But with AI models as part of the governance process, the task can be completed in a fraction of the time, by machines.” 

It's also important to remember that AI learns based on whatever data it receives. With that in mind, it’s important that finance teams control the data machine learning processes ingest to ensure the data is relevant and to avoid introducing biases into its analysis.  

AI can be an invaluable assistant to finance teams

By adding AI to your finance team, you’re giving them the ultimate helping hand. Not only can AI automate repetitive processes, but it can also provide finance teams with access to data trends and performance insights that would otherwise be inaccessible, buried under the enterprise’s mass of unstructured data.   

AI in CCH Tagetik runs platform-wide, augmenting the speed and accuracy of CPM processes and expanding data availability across your enterprise. Using a glass box approach, our explainable AI gives finance teams the authority to check, vet, and accept the AI’s work. You stay in control with the AI as your sidekick. Specifically, AI in CCH Tagetik can be used for data collection, anomaly detection, predictive planning, analytics, and driver-based planning.  

Learn more about the AI in CCH Tagetik and how it can support your finance team.