Digital Transformation
Jul 19, 2018

How analytics augments distribution requirement planning 

Advances in supply chain technology take the guesswork out of deployment

Imagine not being able to order a Big Mac from McDonalds because that vital slice of tasty cheese isn’t available. If you look at the supply chain behind the scenes, one reason could be that the cheese hasn’t made its way from the manufacturing plant to the distribution center (DC) on time. It’s when you can’t order your favorite meal that you appreciate the importance of accurate distribution requirement planning (DRP).

CPG manufacturing companies often face supply firefights despite their best-in-class DRP systems, mainly because they must balance inventory, service levels, and resources. Even though their existing planning systems provide a deployment solution, they don’t always hit the mark. That’s often due to manual, sequential processes and actions that rely on a complex set of interdependent variables, such as warehouse capacity, lead time, and lane capacity. Competing criteria also play a role—service level maximization versus cost minimization, for example. The result? Much productive planning time is lost because people must make manual adjustments to plans. 

What’s wrong with existing planning systems?

Existing planning systems, which see lead time as the principal constraint, can produce one-dimensional plans. Generally, they break down orders sequentially and expect demand upstream to allocate inventory (Figure 1).

In practice this creates plans that don’t consider other supply constraints, such as finite warehouse capacity, lane capacity, and transportation capacity—and that impacts upstream as well as downstream supply chain processes. Yet supply chain challenges are neither one dimensional nor sequential in nature.

Figure 1. An illustrative supply network

Why is this hard to solve?

Distribution and logistics planners encounter increasingly complex supply chain network priorities. Establishing cross-functional linkages is an uphill task, but there are other challenges as well:

  • Data availability, integrity and completeness: Cumbersome processes to work out DC storage, unloading capacity, lane capacity, and inventory availability data
  • Connecting the dots: It’s difficult to establish the link between different datasets and getting the right term view
  • Speed: Post-planning corrections using Excel-based solutions for a data-heavy supply chain take forever to provide actionable insights
  • Prediction: Unless planners understand future projections, they can't accommodate deployment decisions

What's the answer?

Recent technology advancements have produced great customizable platforms that can create the single source of truth that planners seek. They bring together different sources of data that can be viewed in real time for a rapid response to distribution problems.

Using a custom-based platform as a bolt-on that sits on top of planning systems will help create an optimized and synchronized deployment plan over the right timeframe. With the right data in play, and robust governance in place, custom-tailored optimization can shrink the demand-supply gap by assigning and prioritizing weights to use inventory stock at DCs (Figure 2).

Figure 2. Balancing service level and cost to serve

Firms can then apply predictive and prescriptive analytics to this data to spot and solve demand problems, leading to fewer cuts in orders. They can also provide planners with future projections to optimize resources such as labor, machinery and transport. This helps mitigate risks because the analytics monitor projected case fill rate and truck backlogs. Finally, machine learning can bring a deeper understanding of deployment activity, identify root causes of service failure, provide DC pipeline inventory view and generate capacity alerts to avoid stock diversions.

About the author

Anand Bihari

Anand Bihari

Senior Manager, SCM Practice

Follow Anand Bihari on LinkedIn