How to Embrace the New Normal in Demand Planning

Shipra Sharma, Head of AI & Analytics

It’s no secret that demand volatility caught supply chain professionals off-guard during this pandemic. Over the past six months, we have seen demand patterns ranging from huge spikes to deep troughs, and everything imaginable in between. Suddenly, what had been adequate planning tools have now become entirely defunct because of their reliance on historical trends.

A side effect of this situation is that companies that relied heavily on statistical forecasts have been unable to find an alternative approach to demand planning and get their demand planning back on track in such a short span of time. Excel-based planning spreadsheets have once again assumed center stage as planners try to adjust forecasts based on local and regional trends to the best of their ability, with the limited information and tools available to them.

Not surprisingly, one of the biggest outcomes of this pandemic is that business conversations have become a lot more customer-centric – and all the basic assumptions around consumer consumption and buying behaviors are being challenged daily. This has prompted planners to go back to fundamentals and rethink and redesign their demand planning processes.

Before overhauling the planning process, it’s imperative that we first understand the impact of COVID-19 on consumption patterns.

  • Spending has decreased in necessity categories like automotive additives because people are driving their vehicles less, and higher-end luxury and discretionary categories like electronics
  • Consumers are buying fewer items in certain categories like cosmetics because they’re staying at home more, and school supplies because of the move to online learning
  • Rate of consumption on items like household products and cleaning supplies has seen a significant increase; consumers are buying more and consuming these products faster
  • Fear-based buying, or hoarding, of goods like packaged food and paper products has led to product shortages and even caused some retailers to enact rationing measures

How long will these consumption patterns last? And how can you enhance your planning processes to enable forecasting accuracy – even in times of extreme demand volatility?

Start by asking if your traditional statistical forecasts are enough, or if the current systems allow you to expand features and capabilities beyond the norm.

Organizations that have invested in intelligent tools that consider not only traditional statistical forecasting models but also environmental variables to drive planning have managed demand volatility better – and there is data to prove that.

increase in
forecast accuracy

reduction in
operational costs

reduction in



Here’s how to approach it:

Step 1. Gain Visibility into True Volatility

Visibility into end customer demand has become critical during these times. This means getting access to PoS (Point of Sale) information. If this is not feasible or economical, then the second-best option is to triangulate orders, shipments and forecasts. The end goal is to infer volatility.

Getting your hands on relevant data is a step in the right direction, but it’s not enough.

Step 2. Analyze your Data

There are demand sensing tools in the market that help with sharing and analyzing customer demand data with robust underlying platforms. These bring the power of a platform to ingest and analyze massive amounts of data. Their focus is on analyzing short-term trends and determining if any changes are needed to supply plans.

Imagine you’re in the business of selling spare auto parts. Intuitively you know that newer cars are less likely to be sold during a pandemic due to shrinking spend on luxury items. Through PoS data analysis, you notice that a major auto parts retailer is selling more carburetors by the day. In the traditional planning model, you would see that spike – but not until after a week or even a month depending on when your systems refresh or maybe when the retailer gives your supply manager a call. Demand sensing removes the dependencies and delays.

What you gain is the ease of a timely alert that prompts immediate action. With clarity on the uptick in carburetor spending, you can now commission more carburetors or start to pull in safety stock to meet demand and beat volatility.

Step 3. Reinforce your Forecast

Environmental factors impact your forecast, and most planners intuitively know which causal environmental factors drive demand more than others. But in the case of COVID-19, even the most experienced planners couldn’t have guessed or estimated the impact with a high degree of accuracy. As a result, at the onset of COVID-19, most supply chain leaders and planners were blindsided by the demand volatility.

In more recent months, it’s been helpful to know what areas are lifting their lockdown, where the economic activity is resuming, what the current employment index is, etc. This information is readily available and helpful in its own right – but what planners may not accurately be able to infer is the extent and potential impact of these factors.

These leading indicators could vary widely based on industry and use case, but they may directionally give a boost to the accuracy top-down. Correlating the indicators with demand drivers and running scenario analysis could help you build out a few best- and worst-case scenarios. Such an analysis, once layered with PoS volatility, can give you a more complete and data-backed picture of true demand.


Some other considerations:

What is the right granularity – daily, weekly, monthly?

Does this approach work for products with long lead times?

What if my supply plans aren’t flexible enough?


Interested in learning more or talking through your specific needs with a demand planning expert? Simply email with your availability and we’ll get a meeting scheduled.