Enhancing Retailer-Brand Collaboration through Data-Driven Forecasting

Effective collaboration between retailers and consumer brands is pivotal in today’s competitive retail landscape. A crucial aspect of this collaboration is forecasting purchase orders from the retailer to the brand, a process that has evolved from simple estimations to sophisticated, data-driven strategies. This article delves into the stages of this evolution, highlighting the growing importance of point-of-sale (POS) data in shaping demand forecasts.

 
 

The Initial Approach: Simple Percent Increases or Decreases to the Wholesale Forecast

Traditionally, brands did not use retailer point-of-sale data to construct their wholesale forecasts. Instead, they relied heavily on the retailer’s guidance, often using straightforward methods like applying percent increases or decreases to their demand forecasts. This approach, while simple, needed more precision and insights for optimal inventory management.

Incorporating Point-of-Sale Data for Enhanced Forecasting

 
The next stage in this evolution involves integrating POS data and other relevant data features from the retailer’s reporting portal. This data serves as a valuable tool for forecasters in developing a more informed demand plan. The use of such data is multifaceted and can significantly enhance forecasting accuracy in several ways:

Stock market analytics vector illustration. Cartoon tiny people work on conversion of big data to infographic charts and diagrams in cone funnel with pipeline, marketing analytics and optimization

Initial Store Sets to Help with Forecasts

 
When a retailer decides to stock a new item from the brand, they determine the number of stores, or “doors,” where the product will be available. Brands can leverage this data to create a demand plan based on historical sales velocity rates, providing a more accurate initial demand forecast.

 Let’s walk through an example:

 Step 1: Determining the Number of Stores
 

The retailer decides to place the new product in 150 of its stores. These are the “doors” where the product will be available.

 
Step 2: Estimating Sales Velocity Per Store
 

Based on historical data of similar products, the brand estimates that each store will sell approximately five units of the new product per week.

 
Step 3: Projecting the Sales Period 
 

The initial sales period for this new product is projected to be 20 weeks.

 
Step 4: Calculating Total Demand Forecast – Bringing it all Together
 

To calculate the total demand forecast for the 20-week period, the brand multiplies the number of stores by the estimated sales per store per week and then multiplies this by the number of weeks.

Calculation:
 
  • Number of Stores (Doors): 150

  • Estimated Sales per Store per Week: 5 units

  • Sales Period: 20 weeks

So, the total demand forecast is:    150 stores × 5 units/store/week × 20 weeks = 15,000 units

 

Based on this calculation, the brand forecasts a total demand of 15,000 units for the new product over the 20-week period. This forecast will guide the brand in production planning, inventory management, and logistics to ensure adequate stock is available across the 150 stores to meet customer demand without overstocking.

 

Exception Reporting – Store Distribution Accountability to Ensure Forecasts are “True”

 

This process entails meticulously verifying store counts against the figures initially agreed upon during sell-in discussions. For instance, retail buyers may indicate to the brand that a new item will be featured in 75 stores. However, as point-of-sale data begins to accumulate, the brand has the opportunity to compile “distinct counts” reflecting the actual number of stores carrying the inventory. It’s not uncommon to find that the actual number of stores participating in the new product rollout differs from the number initially proposed during the sell-in phase. This discrepancy highlights the importance of ongoing monitoring and adaptation in retail distribution strategies, ensuring alignment between planned and actual product placement.

Predictive Analysis

 

When retailers supply forward-looking, predictive point-of-sale (POS) data segmented by individual items, it empowers brands with the capability to conduct predictive analysis for weeks of stock (WOS) analysis. This approach plays a pivotal role in accurately identifying overstocked or understocked SKUs. By leveraging this data, brands can achieve a more optimal balance in their inventory management. This streamlines the supply chain and ensures that product availability is finely tuned to meet consumer demand, minimizing the risks associated with excess inventory or stock shortages.

success concept, achieving goals, business vector illustration, employee studying infographics, analyzing evolutionary scale, online training vector. Modern vector flat illustration.


Purchase Order Plans Provided by the Retailer

 
When retailers share their purchase order plans, it presents an invaluable opportunity for brands to align these insights with their internal inventory levels. By conducting a thorough cross-reference of this data, brands can effectively illuminate any significant discrepancies. This critical comparison is a cornerstone for refining strategies, enabling brands to make well-informed adjustments. Such proactive measures are essential for maintaining the harmony of supply and demand, ensuring that inventory management is efficient and responsive to the evolving market needs. 
 

 

The Role of Machine Learning

 
At this advanced stage, brands incorporate machine learning (ML) to refine their forecasting further. ML models can be used in two critical scenarios: 

Planning Business Strategy and Managing Data on Laptop and Smartphone. Business Intelligence and Analysis Concept. flat vector modern illustration

 

1. Creating P.O.S. Predictive Models:
 
In scenarios where a retailer doesn’t supply a predictive point-of-sale (POS) model, it opens a strategic opportunity for brands to take the initiative. Developing an in-house predictive model allows brands to pinpoint and address potential inventory challenges proactively. While delving into the specifics of creating such models falls outside the purview of this article, it’s a venture that brands should seriously consider. Embracing this advanced approach to data analysis not only positions brands at the forefront of market trends but also provides a competitive edge in the rapidly evolving retail landscape. For brands aiming to leverage data to its fullest potential, investing in developing a bespoke predictive POS model is worth exploring. 

2. Analyzing the Relationship between POS Predictive Models and Actual Purchase Orders
 
When retailers offer a predictive point-of-sale (POS) model without an accompanying forecast for purchase orders, it presents an opportunity for brands to leverage machine learning (ML) to learn the relationship between both datasets. By applying ML techniques, brands can intricately analyze and understand the dynamics between the POS predictions and the actual orders received. This advanced analytical approach allows brands to uncover valuable consumer purchasing patterns and inventory needs. It enables brands to bridge the gap between predicted market trends and real-world sales data, facilitating more informed and strategic decision-making in their inventory and supply chain management.
 

Conclusion

 


The transition from fundamental forecasting methods to advanced, data-centric strategies signifies a major leap forward in the synergy between retailers and brands. In this evolving landscape, effectively utilizing point-of-sale data and integrating technologies like machine learning enables brands to generate more accurate forecasts, minimize inventory imbalances, and improve overall business performance. To facilitate this level of sophisticated analytics, brands can form strategic partnerships with SaaS companies such as Krunchbox. These collaborations are instrumental in establishing the necessary data pipelines, ensuring that brands can access the cutting-edge tools and insights needed to thrive. As the retail sector continues its progression, data-driven methodologies become increasingly crucial, laying the groundwork for more streamlined, profitable collaborations between retailers and consumer brands.