In today's fast-changing consumer-packaged goods (CPG) sector, leveraging advanced analytics and artificial intelligence (AI) is crucial for staying competitive. In a recent webinar, Danielle van der Ende, Product Marketing Director at Visualfabriq, shared valuable insights into how these technologies are transforming commercial planning and revenue management in the CPG industry. This blog post explores the key points discussed in the webinar, providing concrete examples and addressing common questions.
Watch the on-demand webinar recording.
Advanced analytics involves using sophisticated techniques and tools to analyze data and extract deeper insights. It goes beyond traditional spreadsheets or business intelligence by creating predictions, generating data-driven recommendations, and using techniques like machine learning and AI to forecast, visualize, and create scenarios.
AI, on the other hand, refers to the simulation of human intelligence processes by machines, especially computer systems.
These technologies are essential because they enable businesses to gain insights from large and complex datasets, enhance commercial planning, and boost profitable revenue growth.
CPG manufacturers are increasingly turning to AI and advanced analytics to gain an edge. There are two main levers for this: automation and optimization.
Automation aims to enhance efficiency and predictability.
Optimization is all about enhancing effectiveness. While efficiency saves time and resources by doing things right, effectiveness ensures we’re doing the right things. It's crucial to drive effectiveness, but we must also recognize the benefits of efficiency and predictability. These elements are key in boosting revenue growth and profitability. Balancing both effectiveness and efficiency can lead to more sustainable and predictable business outcomes.
In essence, CPG companies need capabilities that enhance efficiency, predictability, and effectiveness to thrive in today's market.
When it comes to revenue management and commercial planning, there are three key areas where AI and advanced analytics make a significant impact: demand planning, trade pricing and contract terms management, and promotion planning.
Demand planning: Automating predictions with AI is a tremendous time saver. It allows companies to plan at a higher level of detail by processing and analyzing large datasets. This leads to better predictions of baseline volume, lift factors, and P&L forecasts. By automating these processes, companies can focus on strategic decision-making rather than manual data entry and analysis.
Pricing and trade terms management: AI is pivotal in price and trade terms planning by identifying anomalies. Game theory scenarios are particularly useful for making informed pricing decisions. One of the most common applications is pricing elasticity, which is crucial for understanding shopper responses to price changes. This is particularly challenging in the CPG industry with its high promotional pressure, as promotional price elasticity can significantly differ from price elasticity on regular sales. Using conjoint analysis, companies can better predict consumer behavior and optimize their pricing strategies.
Promotion planning: Predicting sales volume at specific retailers involves understanding both shopper behavior and account behavior. AI helps predict how shoppers will respond to promotional mechanisms and timing, as well as how retailers will behave in terms of ordering. This includes considering factors like forward buying, stock loading for displays, and phasing (timing of orders). By accurately predicting these behaviors, companies can improve their forecast accuracy and ensure that products are available when and where they are needed.
In summary, AI enhances demand planning, pricing and trade terms management, and promotion planning by providing detailed predictions and insights. This leads to better decision-making, improved efficiency, and increased profitability.
It's crucial to build the right model and parameters for your AI system. While it's tempting to include every possible variable, it's important to be selective. The goal is to have a model that provides actionable analytics and adds value at the right stage of your workflow.
When it comes to promotional analytics, there are five key factors to consider:
1. Baseline: This is the volume you normally sell without any promotions.These parameters help you predict the effectiveness of a promotion before it happens. AI-powered optimization can run scenarios to determine the best promotional strategy, not just for individual promotions but for the entire year. This helps in deciding whether to change the timing, mechanism, or other attributes of a promotion.
Some factors, like weather, can have a big impact in certain categories. However, they might not provide actionable insights if included in the prediction model. This is because promotions often require final volume sign-off months in advance. Moreover, by the time you get a reliable weather forecast, the retailer has already planned their media ads and shelf placements. Therefore, external factors like weather or competitor activity are mainly useful for post-evaluation, where you can determine their impact on actual performance. It's important to always remain critical and check for outliers in forecasts and actual results.
While AI and advanced analytics offer significant benefits for the consumer goods sector, implementing these technologies comes with its own set of challenges. Three major areas of concern are the human factor, data and the prediction model.
The human factor: Implementing AI-powered analytics requires a cultural shift within the organization. Employees need to trust and understand the technology to fully embrace it. This involves training and upskilling the workforce to work alongside AI tools effectively. Employees should neither reject nor blindly trust AI-generated insights and understand that AI doesn’t replace human intelligence but augments it. Resistance to change and fear of job displacement can be significant barriers. Therefore, it's essential to communicate the benefits of AI, such as enhanced decision-making and efficiency, and to involve employees in the implementation process to gain their buy-in. Additionally, it's important to address the paradox that more complex models, which often provide better predictions, are less explainable. Providing user-friendly interfaces and clear explanations can help build trust in the system.
Data challenges: The effectiveness of AI and advanced analytics also heavily relies on the quality and quantity of data available. In the CPG industry, data can be fragmented and siloed across different departments and systems, making it difficult to gather a comprehensive dataset necessary for accurate analysis. Additionally, data quality issues such as missing, inconsistent, or outdated information can hinder the performance of AI models. Ensuring data integrity and establishing robust data governance practices are crucial steps in overcoming these challenges. Internal data, such as product master data, sell-in, and demand forecasts, need to be accurate and well-maintained. External data, like sell-out, weather information and competitor activity, must be integrated effectively to provide actionable insights.
Read more about Visualfabriq’s data integration engine Bifrost.
Model challenges: Building effective AI models requires being selective. You should focus on key variables to ensure actionable analytics. Factors such as baseline sales, promotional mechanisms, placement, media support, and timing are crucial. Short-term influences like unseasonable weather have little practical use, as reliable weather information becomes available too late to make any significant changes to planned activities. One problem that AI models can have is overfitting, which means that while you were testing, you were using a dataset that was too harmonious, or you were adding so many parameters that your model is not capable to generalize anymore. While the model seemed to work fine on a specific dataset, in a ‘laboratory environment’ so to say, it is not adequate ‘out in the wild.’
In summary, addressing human, data, and model challenges is essential for successful implementation of AI-powered analytics. By ensuring high-quality data and fostering a culture of trust and collaboration, companies can fully leverage the potential of AI to drive growth and profitability.
Leveraging advanced analytics and AI can seem daunting, but Visualfabriq enables a phased approach that makes the transition smoother and more manageable. This approach allows companies to gradually integrate AI into their workflows, ensuring a seamless transition from manual processes to fully automated systems. Right from the start, customers benefit from improved efficiency and productivity because the system seamlessly combines and harmonizes multiple data sources, automates data flows to eliminate error-prone manual data entry, and enhances cross-functional collaboration through configurable workflows and audit-proof approval flows.
Phase one: Starting small
The first phase in a step-by-step approach to leveraging AI involves starting small, beginning, for example, with a single plan account and one product group or all products for that one plan account. This allows companies to evaluate the performance of the AI model on a limited scale. During this phase, it's crucial to continuously assess the model's accuracy and make necessary adjustments. Internal training can also begin at this stage, helping teams get accustomed to the new system. Additionally, companies can start adapting their reporting to incorporate more advanced analytics, which can provide valuable insights even at this early stage.
Phase two: Parallel implementation
In the second phase, AI modeling is used in parallel with manual forecasting. This dual approach enables companies to compare the results of AI predictions with manual forecasts, building trust in the AI system. Visualfabriq's software excels at enabling manual forecasting while running AI models in the background. By comparing the AI predictions—which typically get better over time—with manual forecasts, companies can see the accuracy and reliability of the AI model, which helps in gaining confidence in the system.
Phase three: Full automation
The final phase involves moving to full automation of predictions and zero-touch planning. At this stage, the focus shifts to manual forecasting of outliers and identifying anomalies in the data. This phase also sees meetings evolve to focus on understanding the drivers behind the numbers rather than questioning their reliability. With reliable data as a single source of truth, meetings can become more productive, focusing on strategic decisions and actions to achieve targets. Understanding the impact of promotions and price changes also becomes easier, leading to more informed decision-making.
Throughout all phases, data availability and quality are prerequisites for success. Selecting the right solution that supports phased adoption and provides user-friendly interfaces is crucial. Visualfabriq's approach ensures that companies can gradually transition to AI-powered analytics, enhancing their planning and decision-making processes while building trust and confidence in the new system.
For CPG manufacturers, leveraging advanced analytics and AI is no longer optional—it's essential for staying competitive. By automating routine tasks, enhancing predictability, and optimizing effectiveness, these technologies empower CPGs to make data-driven decisions that drive revenue growth and profitability. While the journey to full AI integration comes with challenges, a phased approach, as enabled by Visualfabriq, ensures a smooth transition. By addressing data quality and fostering a culture of trust and collaboration, companies can fully harness the power of AI to transform their commercial planning and revenue management processes.
To assess the potential for your company, request a consultation with a Visualfabriq expert.