A prescriptive approach for AWS Supply Chain proof of values (PoVs)

Introduction

The COVID-19 pandemic prompted organizations to reevaluate their supply chain strategies to ensure they could meet rapidly changing customer needs. Companies are shifting away from traditional push or pull inventory approaches toward more dynamic, complex models that can better match customer demand. In this rapidly evolving landscape, the ability to experiment and validate new technologies swiftly has become essential for businesses to stay ahead of the competition.

Companies are investing heavily in large language models (LLMs) and machine learning (ML) to augment their demand planning processes and better match production with dynamic customer needs. AWS Supply Chain’s Demand Planning capability empowers you to leverage sophisticated ML algorithms for generating accurate forecasts. Notably, you can take advantage of the free tier to determine the feasibility and conduct a PoV for your business processes before committing further resources.

A common approach for adopting new supply chain technologies or implementing major changes is to conduct proof of value (PoV). A PoV allows organizations to demonstrate a solution’s functional feasibility, value, and return on investment (ROI) before embarking on full-scale implementation. It involves hands-on experience, gathering user feedback, identifying and rectifying potential issues, measuring against defined success criteria, and providing insights to influence data-driven decision-making. By mitigating risks, justifying investments, and ensuring alignment with business goals, a PoV plays a crucial role in making informed strategic choices and enabling continuous improvement.

This blog post outlines a three-step process based on an actual PoV to help you effectively scope and conduct a PoV for AWS Supply Chain Demand Planning. It addresses common questions like:

1. What is a proof of value (PoV) and what does it provide?
2. Is there a supply chain management proof of value (PoV) template or framework?
3. What does a supply chain proof of value (PoV) example look like?

A prescriptive approach provides a robust framework that can be applied to evaluate other AWS Supply Chain capabilities and functional processes, serving as a valuable guide for effectively conducting any type of PoV.

Correctly scoping your PoV

It’s essential to scope the PoV properly to ensure you can accurately assess the solution’s capabilities and performance. This section outlines the recommended steps to scope a PoV and provides an example from a food and consumables distributor conducting a PoV on AWS Supply Chain Demand Planning.

Step 1: Establish PoV success criteria

The first step is to determine how you will measure the success of the PoV. This includes identifying the processes you want to evaluate and the performance metrics you will use. For example, the food and consumables distributor wanted to evaluate demand planning, so they used industry-standard demand planning accuracy metrics like WAPE (Weighted Absolute Percentage Error) and MAPE (Mean Absolute Percent Error).

MAPE is a widely used statistical measure to evaluate the accuracy of forecasting methods. It calculates the average absolute percent error between the forecasted values and the actual values, expressed as a percentage. A lower MAPE indicates better forecast accuracy. In demand planning and inventory management, MAPE is commonly used as a key performance indicator (KPI) to evaluate and compare the accuracy of different forecasting models or methods, set acceptable forecast error targets (the industry standard is typically 10 percent or less), and monitor forecast accuracy over time to identify areas for improvement. You can read more about MAPE in our Amazon Pharmacy case study about how AWS Supply Chain improved forecasting accuracy.

The distributor also evaluated the overall user experience and whether the process was simpler and reduced processing and operational time.

Step 2: Identify the best product mix

Next, identify the products or product groups you want to include in the PoV. This helps ensure you can test the solution’s capabilities effectively across different product types and demand patterns. The selection of products to test in a Proof of Value (PoV) depends on the problem you’re trying to solve and the most effective ways to test the limits of the PoV. Ideally, your PoV can cover all your products and test demand planning effectiveness across various products and product combinations. However, this approach would be time-consuming, require significant resources, and increase complexity. A more practical and feasible approach is to group your products into a smaller subset. You can base this subset on factors such as profit margin contribution, rate of sales/consumption, or a combination of these factors. Testing a smaller dataset is more effective and faster than testing a larger one.

The food and consumables distributor segmented their products based on margin contribution and inventory turns. They created a PoV test subset from the following product groups:

1. Fast movers and high margin contribution (making 60-75% of their margin)
2. Medium movers with lower margin contribution (10-15% of their margin)
3. Infrequent or low-velocity movers

Step 3: Identify the correct inventory locations

Determine the inventory locations you want to include in the PoV, such as distribution centers, warehouses, flow-through sites, or stores. The recommendation is to pick 1-3 locations based on factors like business volume (high, medium, low), operational significance (covers tier-A customers, tier-B customers, etc.), and the product mix identified in Step 2. Another variable to consider is the channel hierarchy, which is useful if you have broad categorizations of customer tiers based on volume and business importance (e.g., tier-A, tier-B). The recommended approach is to model the PoV based on how your current business process is managed. For instance, if your demand planners adjust forecast numbers based on customer clusters, then that would be an appropriate level to model.

By following these steps, you can effectively scope a PoV to accurately evaluate capabilities and performance based on your business needs.

Conclusion and next steps

Conducting a well-scoped PoV is crucial for organizations to validate the feasibility and value of adopting new supply chain technologies or processes before full-scale implementation. The three-step process outlined in this blog post provides a prescriptive approach based on real-world customer PoVs, enabling businesses to accurately assess the capabilities and performance of AWS Supply Chain Demand Planning against their specific requirements. By establishing clear success criteria, identifying the optimal product mix, and selecting the appropriate inventory locations, organizations can mitigate risks, justify investments, and ensure alignment with their business goals. This systematic approach not only facilitates informed decision-making but also paves the way for continuous improvement and a competitive edge in the rapidly evolving supply chain landscape.

Getting started with AWS Supply Chain is simple and doesn’t require any upfront licensing fees or long-term commitments. Begin your journey with the following three steps:

1. Learn about AWS Supply Chain: Visit the AWS Supply Chain website to understand the product’s features and capabilities.
2. Get a technical overview: Explore the AWS Workshop Studio for a self-paced technical walkthrough. You’ll learn how to create an instance, ingest data, navigate the user interface, create insights, and generate demand plans.
3. Start using AWS Supply Chain: Once you’re ready, access the AWS Console and begin streamlining your supply chain operations with AWS Supply Chain’s efficient, data-driven management tools. You can also access the user guide for detailed setup instructions and additional guidance.