Frequently Asked Questions

Demand Planning FAQs

Home » Demand Planning Faq’s
FAQs

Demand FAQs

  • How is AI/ML used in demand planning?

    ML models such as regression, time series analysis, and deep learning can be trained on historical data to predict future demand accurately. These models take into account various factors like seasonality like holidays, trends, promotional activities, macro indicators, and external events that influence demand. Demand sensing techniques leverage real-time data such as point-of-sale transactions, as well as external and enterprise data that helps in creating demand forecasts predictions close to the actual expected demand. AI-driven demand forecasting systems can identify uncertainties in demand patterns, such as sudden spikes or drops in demand, supply chain disruptions, or changes in consumer preferences. This enables businesses to proactively mitigate risks and adapt their demand planning strategies accordingly.

  • What are the parameters that impact the demand planning process?

    Historical Sales Data: Past sales data is a fundamental parameter for demand planning. It helps in identifying patterns that can be used to forecast. Market Trends: Understanding market trends, including economic conditions, industry trends, competitor actions, and consumer preferences. Promotional Activities: Promotions, discounts, and marketing campaigns can influence demand levels. Demand planners need to consider the timing, duration, and effectiveness of promotional activities when forecasting demand. Seasonality: Many products experience seasonal demand variations influenced by factors such as holidays, weather, and cultural events. New Product Launches: Introducing new products or product variants can impact demand for existing products. Supply Chain Factors: Factors such as lead times, production capacity, supplier performance, and transportation constraints can affect the availability of products and influence demand patterns. External Events: External events like natural disasters, regulatory changes, geopolitical tensions, and pandemics can disrupt supply chains. Customer Behavior: Understanding customer behavior, preferences, purchasing habits, and buying cycles is crucial for demand planning.

  • What is decision science?

    Decision science is an interdisciplinary field that combines techniques from various disciplines, including mathematics, statistics, economics, psychology, computer science, and management science, to study and analyze decision-making processes. It focuses on understanding how individuals and organizations make decisions and developing methods to improve decision-making outcomes. Key aspects of decision science include: Probability and Statistics: Probability theory and statistical methods are used to quantify uncertainties, assess risks, and analyze data relevant to decision-making. Optimization: Optimization methods are employed to find the best possible solutions to decision problems given constraints and objectives. Linear programming, integer programming, dynamic programming, and simulation optimization are used. Behavioral Economics: Decision science incorporates insights from behavioral economics to understand how cognitive biases, heuristics, and emotions influence decision-making behavior. AI Modeling: Developing robust decision support systems (DSS) and software tools to aid in analyzing complex patterns in existing systems, training itself, and generating solutions. DSS often integrate data analytics, modeling, visualization, and optimization capabilities to facilitate decision-making.

  • Where to use Time-series models and where to use ML based models?

    Time-series models are best suited for forecasting tasks where the data points are collected over time and exhibit patterns such as trends, seasonality, and autocorrelation. They are particularly effective when the focus is on capturing the temporal dependencies and dynamics inherent in the time series data. Time-series models like ARIMA, SARIMA, Holt-Winters, and LSTM are commonly used in the following scenarios: Demand Forecasting, Financial Forecasting, Energy Consumption Forecasting, Weather Forecasting and Healthcare Forecasting. Machine learning (ML) based models, on the other hand, are suitable for forecasting tasks where the data exhibit complex relationships, non-linear patterns, and high-dimensional feature spaces. ML models like random forest, XGBoost, and neural networks excel in capturing intricate patterns and making accurate predictions based on diverse sets of input features. ML based models are commonly used in the following scenarios: Demand Modelling, Customer Churn Prediction, Marketing Analytics, Predictive Maintenance, and Risk Assessment

  • What is a demand planning software?

    Demand planning software is a type of business software designed to help organizations forecast future demand for their products or services accurately. These software solutions typically integrate advanced forecasting algorithms, data analytics tools, and collaboration capabilities to facilitate the demand planning process. Demand planning software offers various features and functionalities to support different aspects of demand planning, including: Forecasting Models: Demand planning software provides a range of forecasting models, such as time series analysis, regression analysis, and machine learning algorithms. Data Integration: Demand planning software allows users to integrate data from multiple sources. Demand Collaboration: Demand planning software enables collaboration among different stakeholders involved in the demand planning process. Scenario Planning: Demand planning software allows users to create and analyze multiple demand scenarios based on different assumptions and factors that may impact demand, such as changes in market conditions, pricing strategies, or promotional activities. Reporting and Analytics: Demand planning software provides reporting and analytics tools to monitor forecast accuracy. Integration with ERP Systems: Many demand planning software solutions integrate seamlessly with enterprise resource planning (ERP) systems, such as SAP, Oracle, and Microsoft Dynamics.

  • How forecast accuracy typically varies for Efficient and Responsive Supply Chain?

    Efficient and responsive supply chains represent two different approaches to managing the flow of goods and services. Each approach has its own set of characteristics and priorities, which can affect forecast accuracy in different ways. Efficient Supply Chain: Forecast accuracy is typically higher in efficient supply chains because they prioritize stability and consistency. Since the emphasis is on minimizing waste and maximizing efficiency, accurate forecasts are essential for planning production schedules, managing inventory levels, and optimizing resource allocation. Responsive Supply Chain: Forecast accuracy may be lower in responsive supply chains compared to efficient ones because they prioritize agility over precision. Since the focus is on quickly responding to changes, such as sudden spikes in demand or shifts in market trends, forecast accuracy may take a backseat to the ability to adjust plans and operations rapidly.

  • What are the key challenges in demand planning for products with short product life cycles, and how can they be addressed?

    Demand planning for products with short product life cycles presents challenges such as limited historical data, high demand variability, and time sensitivity. To address these challenges, companies can implement strategies like collaborating closely with partners to share market insights, and using historical data of similar product in past using predecessor-successor linkages for improved forecast accuracy thereby building agile supply chains to quickly adapt to changes.

  • What are the potential risks associated with relying too heavily on automated demand forecasting algorithms, and how can these risks be mitigated?

    Over-reliance on automated demand forecasting algorithms can lead to inaccuracies due to their limitations in capturing sudden shifts, overlooking qualitative factors, and potential biases. To mitigate these risks, organizations should balance automation with human expertise with the help of collaborative demand planning, continuously validating algorithms, and promote data literacy among stakeholders to make informed decisions. The planners should be actively involved in the forecasting processes to ensure that they can override the system generated engine if their expertise says that the forecast is not accurate. On achieving this type of data-driven decision making culture only, organizations can rely on the automated demand forecasting algorithms.

  • How can companies incorporate factors, such as market trends and consumer behavior, into demand forecasting models?

    To incorporate factors like market trends and consumer behavior into quantitative demand forecasting models, companies can utilize a combination of internal and external datasets. Internal data sources include sales data and enterprise external data, such as information on holidays and promotions. External datasets encompass broader factors like supply chain crises and geopolitical events. By analyzing these datasets alongside traditional quantitative historical data, companies can identify patterns and correlations that reflect market trends and consumer behavior. For instance, they can track sales performance during specific promotions or holidays and correlate it with external events to understand consumer preferences and behaviors better. This approach enables companies to enrich their quantitative models with qualitative insights, enhancing the accuracy and effectiveness of demand forecasting.