RGM Innovation: Disruption in Forecasting

In the face of socio-economic change, dynamic consumer demands, shifting markets and lurking uncertainties, the recent months have brought to the forefront the volatility of the consumer packaged goods industry. Predicting shopper decisions in times of conscious and resilient consumer spending, is proving to be a big challenge for RGM managers. 

 

The growing competition, and the pressing need for innovation has also made it imperative for organizations to acknowledge the drawbacks of their current pricing, promotion, and trade investment processes. The bar is rising and traditional methods are no longer enough. It’s time for businesses to shift to AI-powered revenue growth management for the future.

Störer Blogs  (2)-1

The Future of RGM - Artificial Intelligence and Machine Learning

AI and ML have been key topics for years now and their future importance is undeniable in the CPG industry. Since CPGs are heavily reliant on decision-making efficiencies, working with the right technology becomes even more important. We have to keep in mind here that AI and ML are not synonyms as is often suggested, they are closely intertwined. Hence, before identifying the right application or tool, it is important to understand how they function.

AI (Artificial Intelligence) describes the capability of a computer system to reproduce human cognitive functions. With AI, computers use math and logic to simulate a stage of reasoning that replicates the human ability of learning and decision-making with the benefits of computational capacity. With the help of AI, organizations can replicate virtual customer behavior or models by persona to predict how a customer might act on, for instance, a discount or price offer. Read more on why algorithms price differently from humans, in our recent blog post.

On the other hand, Machine Learning is an application of AI that enables a computer to continue learning and improvising on its own. ML has the strong capability to process large amounts of data, factors, and detect patterns. In the case of CPGs, this can be crucial in unlocking new insights especially those that are locked within the chains of transactional data.

Since RGM is essentially profit-optimization based on human decision-making, there are numerous opportunities to apply AI- and ML-driven processes, e.g. in demand forecasting, price optimization or promotion planning, thereby allowing CPGs to implement a data driven approach to their decision making processes.

Störer Blogs  (1)-1

Impact of AI on Revenue Growth Management

Today, CPGs are effectively using AI in their RGM systems to get deeper insights on productive customer discounts and segments in order to stop revenue leakage into suboptimal investments. Recently, Boston Consulting Group reported that improving pricing optimization with advanced AI techniques has the potential to deliver a 10% increase in total revenue. However, another recent study that stated, ‘incorporating AI is able to reduce errors in forecasting by 20-50%’, also said that, ‘it leaves a huge disparity in the quality of ML algorithms’. Evidently, the sheer use of AI neither increases profitability nor enables higher efficiency or resilience, unless it is implemented well. Simply using it is not going to create the expected impact. The implementation process, however, seems to be harder than many CPGs might have expected given that only 50% of CPGs that have made digital and analytics investments are achieving returns above the cost of capital.  

So, how does a business leverage AI to its maximum potential?
In order to maintain a long-term and sustainable competitive advantage in the sector, CPGs must integrate AI into their RGM that can specifically respond to their shoppers' and retailers’ needs and deal with potential shocks and market trends. 

Important Factors for Successful Integration of AI in RGM

AI is often described as the silver bullet for the industry, but the successful application of AI-driven processes is dependent on many factors and can lead to deterioration if poorly integrated. As outlined, the choice and the development of a sufficient and working ML model is crucial to establish an AI-driven RGM. Here are four main components that determine the success of your implementation:

1. Synergy of code and method

The fast and reasonably accurate results of "bad" models often suggest that there is simply a lack of data or data quality to achieve better results. However, in revenue growth management in particular, the core challenge lies in the general understanding of shopper behavior. The purely rationalistic view of the algorithm will never be able to reflect the exact needs and attitudes of real shoppers without the inclusion of behavioral models that allow the algorithm to understand and learn the irrationality of the human mind. 

 In this respect, it is the interplay of extremely good code and development, and a comprehensive methodology that enables the long term success of AI solutions.

2. The quality and quantity of your data

The truth is, if your input is bad your output will be horrible. Despite the fact that AI conveys a sense of autonomy and intelligence, it is still based on the input that it receives. In this case mostly from the people working with the tool. The reliability of a “human constraint” bears a huge limitation in optimizing AI solution performance that is highly dependent on the preceding creation of data curation and management. 

Although the selection of data depends on the application, there are some general features that should be considered. The input data should be complete, comprehensive, consistent, accurate, valid, and unique. Modern RGM therefore has to become more interdisciplinary to ensure a holistic view on the implementation of AI. 

3. Automation of advancing data and model

The third component is the regular inflow of data and automated updating of the applied model. Especially in fast-changing industries such as consumer goods, it is not possible to apply the same model over time and expect it to deliver high accuracy results, consistently. 

 The chosen AI solution should be able to provide a solid statistical model that not only aggregates or simplifies, but also provides accurate results in the long term and reflects the individual needs of shoppers. An efficient AI solution in this case needs to be updated at most every 6 months. This is true even despite, for example, rising inflation rates or economic fluctuations.

Störer Blogs -1

 

Applying the Future of AI in Consumer Goods - The Virtual Shopper Technology

Buynomics integrates all of the above challenges in the Virtual Customer technology. The proprietary ML algorithm, in combination with behavioral economics models included in the method, is able to mirror the entirety of CPG customers on an individual level. Learn how you can simulate the effects of pricing, portfolio changes, and promotions in our insightful free webinar on Virtual Customer Technology

The predictions provided by our SaaS tool are not based on aggregated estimates of all customers but on the individual decisions of each Virtual Customer. Each Virtual Customer has its own value drivers and behavioral characteristics that change with respect to, for example, the optimization of prices within a product portfolio, depending on the price point. 

Our automated AI solution enables the integration of competitors and cross-functional effects within their own portfolio. Finally, the always-on platform not only allows you to generate and optimize endless scenarios in minutes, but also updates itself in real-time to cope with potential shocks, such as inflation. 

Buynomics’ virtual customer technology is applicable in a wide range of use cases that cover the majority of challenges of todays’ RGM. 

  • Continuous Price Optimization: Optimize prices for every pricing strategy considering behavioral pricing and competition
  • Promotion Optimization: Identify promotion offers and optimize type and discount 
  • Price Pack Architecture (PPA): Compare infinite scenarios of product variations and optimize PPA based on your products and competitors
  • Product Mix Optimization: Decide on product rage and develop new innovative product mixes 
  • Trade Term Optimization: Create full transparency about offering changes and develop trusting relationships with your retailers 

Transforming your Business Future with AI

The successful integration of AI and ML in the consumer goods industry is inevitable for the challenges of the coming years. However, it is not just the mere use of new technologies but the specific optimisation of one's own needs and goals in combination with advanced analytics that will allow CPG companies to achieve highly optimized results. Over time, AI solutions have entered the CPG industry, but have perhaps been subjected to skepticism. Not many companies have taken the step to implementing them. The experiences and insights from this year's EPP conference also illustrated how most CPGs are still hesitant to implement AI solutions even though they are aware of the possibilities and necessity because most solutions bear the problem of  low accuracy, bad usability, and lacking topline results.

Buynomics virtual shopper technology is able to support and guide CPGs on their journey into the future of RGM, empowering teams and managers quickly and efficiently while increasing profitability through the SaaS solution's user-friendly interface and automation. 

 

Are you interested in how machine learning might benefit your RGM?

Talk to our experts