The Business Problem

One of the biggest challenges for business executives today is demand volatility in relation to demand forecasting. Whereas data availability continues to increase, customer purchase patterns are becoming increasingly complex, and therefore harder to detect or predict (Symphony Retail, n.d.).

There are too many factors influencing demand, ranging from weather fluctuations to posts by social media influencers, causing customers to frequently changing their minds. Even worse, things that will reshape customer intentions will mostly happen quite unexpectedly (Alexsoft, 2019). Traditional Forecasts are only as accurate as the data, models, resources and people that have to interpret them (Symphony Retail, n.d.). So, how can we respond to these challenges?

The Solution


AI-powered demand forecasting

A response to demand volatility is demand forecasting using Artificial Intelligence. Traditionally, demand forecasting is a form of predictive analytics, where the process of estimating customer demand is analysed using historical data (Dilmegani, 2021). Using AI, organisations can make use of Machine Learning algorithms to predict changes in consumer demand as accurately as possible. These algorithms can automatically recognise patterns, identify complicated relationships in large datasets and capture signals for demand fluctuation. See figure 1 for a side-by-side comparison between traditional forecasting and ML forecasting.

Typically, organisations use this form of AI to avoid inefficiencies caused by misalignment of demand and supply throughout the operational process. Honestly, this will never be 100% accurate (Alexsoft, 2019). Yet it can offer companies the opportunity to significantly reduce supply chain costs and make improvements in financial planning, workforce planning, profit margins and risk assessment decisions (Dilmegani, 2021).

Traditional
forecasting
Machine Learning
forecasting
Ability to consider
numerous variables
and data sources
Adding extra variables
and sources requires
substantial effort
Multiple variables
and sources can be
smoothly
incorporated
thanks to the high
level of automation
Volume of manual
work
High Low
Amount of data
required
Small Large
Maintenance
complexity
Low High
Technology
requirements
Low High
Best fit Mid / long-term
planning
Established products
Stable demand
Short / mid-term
planning
New products
Volatile demand
scenarios

Figure 1: Comparison of Traditional and Machine Learning Forecasting solutions

Source: (Alexsoft, 2019)


Example

A real-life example can be found at Danone Group, a French manufacturer of food products. Danone implemented a machine learning system to make better demand forecasts. The company required more accurate and secure demand forecasts, due to the short shelf-life of its fresh products and volatile demand (Brosset, et al.). Danone uses many promotions and media events. More than 30% of the total volume is sold through promotional offers such as discounts and leaflets, so the demand forecasts were somewhat ad hoc (Best Practice AI, n.d.).

The implemented machine learning system did not only improve forecasts, but also improved planning between different departments such as sales, supply chain, finance and marketing. This system improved efficiency and inventory balance, allowing Danone to achieve its target service levels for channel or store-level inventories (Brosset, et al.).

Benefits

Several advantages of AI in demand planning have already been mentioned. Below are the key benefits that demand forecasting with AI can bring to any company in the manufacturing industry:

  • Improvements in accuracy over time: Better forecasts will be made over time as machine learning algorithms learn from existing data.
  • Higher customer satisfaction: When products are ‘out of stock’, this will decrease customer satisfaction, whereas customer satisfaction will increase when products are always available. This improves customer loyalty and brand perception.
  • Improved workforce planning: Demand forecasting can support the HR department in making efficient considerations between full-time or part-time staff mix, thus optimising HR costs and effectiveness.
  • Improved markdown/discount optimisation: Cash-in-stock is a common situation for retail companies, where products remain unsold for a longer period than expected. This often causes higher expected inventory costs and the risk of products becoming obsolete and losing value. In this scenario, products are sold at lower selling prices. With demand forecasting, this scenario can be minimised.
  • Overall efficiency: With demand forecasting, teams can focus on strategic issues instead of trying to reduce or increase inventories and staffing levels (Dilmegani, 2021).

Impact

What value can AI add to a company? Let’s look at the numbers.

  • Errors in supply chain networks can be reduced with 30 to 50% with AI-powered demand forecasting.
  • Warehousing costs decrease with around 10 to 40%.
  • The loss in sales due to inventory out-of-stock situations can be reduced up to 65% with the improved accuracy.
  • In general the estimated impact of AI is between 1.2 and 2 trillion dollars in the manufacturing and supply chain planning (Dilmegani, 2021).

For Danone Group, AI in demand planning ultimately led to a 30% reduction in lost sales, a 30% reduction in product obsolescence, a 20% reduction in wrong forecasts and a 50% reduction in the workload of demand planners (Brosset, et al.).

Accessibility and Requirements

In order to implement machine learning in demand planning and forecasting, the ideal AI system is trained using data from different sources such as: weather data, financial data and third-party data (e.g. social media, historical sales data and macroeconomic data). The AI system makes predictions on how event combinations in the past effected demand for future consumer demand (Brosset, et al.)


Figure 2: Example of data types, structures and sources

Source: (Alexsoft, 2019)


Example

L’Oréal, a French multinational and one of the world’s largest beauty products manufacturers, used information from various sources to anticipate on trends, optimize sales and predict customer demand. The sources they used included social media, weather and financial market indicators and data gathered at point-of-sales, such as collection, reception and inventory. Combining these datasets, and identifying which variable combinations affected consumer demand, allowed L’Oréal to target end-customers more effectively and respond to the challenge of demand volatility (Brosset, et al.).

However, no matter how smart the forecasting solution may be. Human logic is still needed to evaluate the relevance of the outcomes produced by AI solutions. To draw conclusions with common sense and domain expertise. Companies should invest in industry specialists to determine what factors should be taken into account in predictive algorithms/models. Only by combining the strengths of both human and artificial intelligence can a company foresee and plan for a better future (Alexsoft, 2019).

Want to know more what AI can do for your supply chain?

Contact us.

More AI Case Studies:

AI case study 1: Demand Forecasting using Artificial Intelligence

The Business Problem One of the biggest challenges for business executives today is demand volatility in relation to demand forecasting. Whereas data availability continues to increase, customer purchase patterns are becoming increasingly complex, and [...]

Bibliography

Alexsoft. (2019, November 11). Demand Forecasting Methods: Using Machine Learning and Predictive Analytics to See the Future of Sales. Retrieved April 6, 2021, from Alexsoft

Armstrong, M. M. (2020, December 4). Cheat sheet: What is Digital Twin? Retrieved April 21, 2021, from ibm.com

BCG. (2015, November 17). Think Outside Your Boxes: Solving the Global Container-Repositioning Puzzle . Retrieved from Boston Consulting Group

Best Practice AI. (n.d.). Danone reduces forecast error and lost sales by 20 and 30 percent respectively and achieve a 10 point ROI improvement in promotions with machine learning. Retrieved April 6, 2021, from Bestpractice

Brosset, P., Patsko, S., Khadikar, A., Thieullent, A., Buvat, J., Khemka, Y., & Jain, A. (n.d.). Scaling AI in Manufacturing Operations. Retrieved April 06, 2021, from Capgemini

Caulfield, B. (2019, December 16). What’s the Difference Between a CPU and a GPU? Retrieved April 22, 2021, from nvidia.com

Dilmegani, C. (2021, January 7). Demand forecasting in the age of AI & machine learning [2021]. Retrieved April 6, 2021, from AImultiple

Europawire. (2020, April 15). Shipment planning and inventory management improved with AI and big data on Kuehne + Nagel’s new version of SeaExplorer. Retrieved from Europawire

Hyeongchan, K. (2020, December 16). Detecting Sounds with Deep Learning. Retrieved April 21, 2021, from towardsdatascience.com

Kvartalnyi, N. (2021, May 11). 6 TIPS OF HOW TO IMPROVE INVENTORY MANAGEMENT USING ARTIFICIAL INTELLIGENCE. Retrieved from Inoxoft

Majumdar, D. (n.d.). Case Study-How SynergyLabs AI solutions Brought Efficiency in warehouse Inventory management. Retrieved April 19, 2021, from Synlabs

Micron Technology. (2021). Case Study: Micron Uses Data and Artificial Intelligence to See, Hear and Feel. Retrieved April 21, 2021, from sg.micron.com

Opperman, A. (2019, November 19). What is Deep Learning and How does it work? Retrieved April 21, 2021, from towardsdatascience.com

Serheichuk, N. (2020, December 15). Inventory management automation: How you can benefit from it. Retrieved from N-ix

Subramanyam, J. (2019, July 8). Prediction Models: Traditional versus Machine Learning. Retrieved May 20, 2021, from Gartner.com

Supply Chain 247. (2017, March 28). Coca-Cola Leverages AI for Inventory Management. Retrieved from SupplyChain247

Supply Chain Dive. (2018, August 17). Two-thirds of companies consider Excel a supply chain system. Retrieved from Supply Chain Dive

Symphony Retail. (n.d.). demand forecasting ai. Retrieved April 06, 2021, from symphonyretail

TradeGecko. (2019, December 4). What is inventory management? Retrieved from tradegecko

Tuv, E., Murat, G., Enis, P., & Lee, D. H. (2018, November). Faster, More Accurate Defect Classification Using Machine Vision. Retrieved April 21, 2021, from Intel.com