How to forecast demand for a new product in consumer goods, wholesale, and retail

Dec 30, 2024 • 9 min

In 1985, Coca-Cola released a new beverage that was predicted to take over the market: New Coke. The company had been losing market share to competitors and carefully formulated a new beverage according to the feedback from focus groups and blind taste tests. But when it was released, consumers hated New Coke, and within three months the product was pulled from shelves.

Even though taste tests predicted huge demand for a new, sweeter version of the popular drink, consumers didn’t want to buy it because they had a strong attachment to what Coke was supposed to taste like. Coca-Cola learned one of the hardest lessons of forecasting demand for a new product: It’s one of the most complex and challenging parts of demand planning.

Whether you’re in CPG, wholesale, or retail, getting new product introduction (NPI) forecasting right can mean the difference between a successful launch and costly inventory issues.

The challenges in forecasting demand for a new product

New product demand forecasting presents unique challenges because companies must make crucial inventory decisions based on limited data around market research, comparable product performance, and industry expertise. This complexity is further compounded by varying factors across different sectors, each requiring its own specialized approach to demand planning.

Coca-Cola didn’t anticipate the cultural and emotional aspect of purchasing and how that might affect consumers’ response to a new recipe for the popular beverage. In the 80s, the technology didn’t yet exist for the demand planners at Coca-Cola to dig into the kind of information that could have warned them off making such big changes to their signature drink. But today, we have sufficiently advanced forecasting technology to identify evidence of these kinds before a product release.

Beyond cultural factors, companies must also navigate the complexities of supply chain timing, market competition, and consumer adoption patterns. Success requires balancing the need for sufficient inventory against the risk of overstock, all while remaining agile enough to respond to early market signals. This delicate balance makes new product forecasting one of the most challenging aspects of demand planning.

Market acceptance is hard to gauge without demand data

One common challenge is understanding potential market acceptance. Companies must rely on various proxies and indicators to gauge demand without historical sales data. Planners getting ready for the launch of New Coke had few ways of anticipating the reaction. Releasing a new product is not exactly stabbing in the dark, but it might sometimes feel like that. Businesses have to be cautious because the indicators can be imperfect predictors of actual market performance, especially for truly innovative products that create new market categories.

Supply chains require lead time and investments

Supply chain considerations add another layer of complexity. Businesses often commit to production quantities, raw materials, and inventory levels before having concrete demand data. This creates a delicate balance between maintaining sufficient stock to meet potential demand while minimizing the risk of excess inventory. Supply chain challenges are particularly acute for produ. Supply chain challenges are particularly acute for products with long lead times or those requiring significant upfront investment in production capacity. long lead times or those requiring significant upfront investment in production capacity.

Each sector is different

Complexity is further compounded by varying factors across different sectors, each requiring a specialized approach to demand planning. The techniques that work well for fast-moving consumer goods might be ineffective for durable goods or luxury items.

How forecasting demand for new products differs by sector

The approach to new product forecasting varies significantly across different industries, each facing unique challenges and constraints. Understanding these sector-specific differences is crucial for developing effective forecasting strategies that account for distinct supply chains, customer base, and product lifecycle characteristics.

CPG sector: Balancing service levels and efficiency

This sector encounters unique challenges in new product forecasting because these brands have to work hard to meet service levels, maintain production efficiency, and minimize excess products while serving a diverse set of customer partners.

Product nature: The perishable nature of many CPG products, particularly in the food and beverage category, creates significant forecasting complexity. Companies must carefully balance inventory levels to prevent stockouts while avoiding excess stock that could lead to waste and financial losses. This challenge is particularly acute for products with short shelf lives, where even small forecasting errors can substantially impact profitability.
Supply chain complexity: The intricate nature of CPG supply chains adds another layer of forecasting difficulty. Consumer trends and preferences can shift rapidly, requiring quick adjustments to production and distribution plans. Additionally, the multi-tiered distribution network, often involving manufacturers, distributors, and retailers, means that demand signals must be accurately transmitted and interpreted across multiple stakeholders to maintain efficient operations.
Consumer behavior: You can’t predict sales until you understand consumers, and their preferences are always changing. Companies must continuously monitor and analyze changing preferences, seasonal patterns, and competitive activities. This requires sophisticated data analysis capabilities and the ability to quickly adjust forecasts based on early sales signals and market feedback.

Wholesale sector: Managing bulk orders and long-term planning

Managing the wholesale supply chain presents aifferent sorecasting challenges, primarily centered around managing large-volume orders and maintaining efficient inventory levels across extended time horizons. The business-to-business nature of wholesale creates more stable demand patterns but requires careful attention to customer ordering cycles and supply chain coordination.

Product nature: Wholesale operations typically deal with longer shelf-life products and larger order quantities than CPG companies. The primary challenge lies in managing high-volume orders efficiently while maintaining optimal inventory levels across warehouses and distribution centers. To avoid costly issues, wholesalers must carefully consider storage capacity, handling capabilities, and transportation logistics.
Business model: The focus of wholesale operations creates more predictable demand patterns than consumer-facing businesses. Large customers typically follow established ordering cycles and quantities, allowing for more structured planning processes. However, disruptions in major customer ordering patterns can significantly impact forecast accuracy.
Supply chain structure: The emphasis on long-term planning in wholesale operations means the wholesale forecasting approaches differ fromretail or CPG. Forecasters must consider minimum order quantities, lead times for bulk shipments, and warehouse capacity constraints. Coordinating with both upstream suppliers and downstream customers is essential for optimizing order timing and quantities.

Retail sector: Responding to consumer behavior

Retail forecasting combines elements of both CPG and wholesale challenges while adding its unique complexities. The direct interaction with end consumers, coupled with the need to manage diverse product categories and seasonal variations, makes retail forecasting particularly challenging for new products.

Product lifecycle: Retailers face significant challenges with inventory obsolescence. This is particularly true for fresh products, fashion, electronics, and seasonal items, where products can quickly become outdated or lose value. Accurate forecasting helps prevent markdown losses while ensuring sufficient stock for expected demand periods.
Market factors: Retail demand patterns are highly influenced by a complex web of factors, including shopping seasons, economic conditions, and local market dynamics. Store traffic patterns, weather conditions, and competitive activities can all significantly impact sales performance. Successful forecasting requires understanding and accounting for these influences on consumer purchasing behavior.
Data dependencies: The retail sector’s position at the end of the supply chain creates unique forecasting challenges. Retailers must collaborate closely with their CPG and wholesale suppliers to understand product availability, lead times, and potential supply constraints. This level of multi-tiered collaboration requires sophisticated data sharing and analysis capabilities to ensure forecast accuracy and inventory optimization.

Seven steps to successful new product forecasting

Successfully forecasting demand for new products requires a systematic approach that combines data analysis, market understanding, and strategic planning. Fortunately, there are some essential steps that can help organizations build more accurate forecasts for their new product introductions.

1. Understand launch scope and objectives

Start by defining the fundamental parameters of your launch:

What is the planned launch date?
Is this product part of a known category or a completely new offering?
Who is the target market, including geographic location, age, gender, and other demographics?
What are the key attributes and characteristics of this new product?

The level at which you plan to forecast can significantly impact accuracy and usefulness. Different industry verticals require different levels of aggregation – wholesale operations might focus on regional-level forecasts for bulk ordering, while retailers often need store-level granularity. Your forecasting granularity should align with your operational capabilities and business objectives.

2. Identify similar products

Find comparable items based on your defined category, market, and product attributes. Modern forecasting systems can automatically detect these reference products, allowing planners to review and adjust selections as needed.

Data quality and comprehensiveness are paramount at this stage. Look for reference products with reliable, clean historical data that can serve as meaningful comparisons. Consider factors like:

Similar market positioning.
Comparable price points.
Related seasonal patterns.
Matching distribution channels.

3. Analyze reference product history

Use the historical performance of similar products as your starting point. Providing a baseline for understanding potential demand patterns and seasonal trends. Consider multiple data types simultaneously — from historical sales data to POS data and external market research. This multi-source approach provides a more complete picture of potential demand.

Order patterns from reference products reveal how customers typically purchase similar items, providing insights into potential ordering behavior. Inventory turnover rates can help predict how quickly your new product might move through the supply chain, while waste rates indicate potential risks and necessary safety margins.

4. Generate initial forecast

Run forecasting models using the historical data from your reference products to create a baseline forecast that can be refined with additional inputs and market intelligence. The complexity of your forecasting methods should match your business environment – sophisticated AI/ML algorithms might be necessary for rapidly changing markets, while simpler statistical models often work well for more stable environments.

Geographic distribution plays a crucial role because consumer preferences and demand patterns can vary significantly by region. For instance, a new beverage product might see different adoption rates in urban versus rural areas, or in different climates.

5. Incorporate team knowledge and market trends

Modify your initial forecast by incorporating:

Cross-functional team expertise.
Current market trends.
Scaling factors to adjust for expected sales magnitude differences.
Advanced modeling techniques to select the optimal forecasting approach.

Even the most sophisticated forecasting model won’t be effective if it requires data you don’t have or produces outputs your operations team can’t act upon. Consider factors like forecast horizon, update frequency, and required accuracy when selecting your forecasting methods. Different sales channels (e-commerce, brick-and-mortar, direct-to-consumer) often show distinct demand patterns that need to be considered separately.

6. Monitor and adjust performance

Once your product launches, continuously monitor sales performance and adjust forecasts as needed. Performance indicators such as early sales data serve as vital feedback mechanisms — by closely monitoring these metrics from day one, you can quickly identify whether your new product is performing as expected and make necessary adjustments to your forecast and production plans.

These metrics are crucial because they directly impact your ongoing production and ordering decisions. For example, if early sales show different patterns than predicted, you might want to adjust production runs until demand patterns become clearer.

7. Transition to actual sales history

As your product builds its own sales history, gradually shift from relying on reference product data to using actual performance data. Modern forecasting systems can automatically manage this transition, ensuring your forecasts become increasingly accurate over time.

This transition period is critical for long-term forecast accuracy. While reference products provide valuable initial insights, your product’s actual performance data will ultimately yield the most reliable forecasts. Continue monitoring both your product’s performance and reference product trends to identify any diverging patterns that might require forecast adjustments.

Optimizing new product forecasting with modern solutions

Modern demand planning solutions have revolutionized new product forecasting through:

Granular forecasting: Generation of forecasts at SKU-Location-Day level, providing flexibility for different industry needs.
Advanced analytics: Use of pooled model approaches to select optimal algorithms based on industry requirements.
Comprehensive data integration: Ability to incorporate various causal factors, including customer trends, seasonality, and economic indicators.
Promotional impact assessment: Capability to evaluate the impact of promotions and sample sales on new product performance.
Inventory optimization: Integration of key metrics like inventory turns, spoilage history, and fill rates to ensure optimal ordering levels.

Successful new product demand forecasting requires a deep understanding of your industry’s unique challenges and the application of appropriate forecasting methodologies. Whether in CPG, wholesale, or retail, the key is to select the right tools and approaches that align with your specific business needs while maintaining flexibility to adapt as demand evolves and changes.

Building a foundation for successful product launches

The complexity of new product forecasting demands both art and science, combining sophisticated analytics with deep industry expertise. While each sector faces unique challenges, success in demand forecasting ultimately comes down to understanding your market, choosing the right level of analytical complexity, and maintaining the flexibility to adjust as new data becomes available.

Businesses that excel at new product forecasting share common traits: they take a systematic approach to data collection and analysis, they choose forecasting methods appropriate to their industry and capabilities, and they maintain the agility to respond quickly to early market signals. This rigorous methodology and adaptable execution makes the difference between successful launches and costly misses.

Level up your demand planning with RELEX

Written by

Pantea Carlton

Field Presales Solution Principal - CPG