Coronavirus series part 2 – Overcoming demand forecasting challenges due to Covid-19

Covid-19 has had a profound impact on global marketplaces, with customer demand and inventory supply more erratic and complex than ever before. The effects of Covid-19 has put supply chain management (SCM) teams well and truly in the spotlight, as they play a crucial role in ensuring business continuity for many organisations across the UK.

Many businesses are turning to technology to support their operations and embracing digitalisation to help overcome Covid-19 challenges. In this blog series we look at how inventory optimisation tools, such as EazyStock, can support inventory management teams.

This is our second post of the series, in which we review the demand forecasting challenges brought about by Covid-19 and how SCM teams can use inventory optimisation to find solutions.

Covid-19 and demand forecasting challenges

Within weeks of Covid-19 being reported, supply and demand across the globe was affected and SCM teams had to move fast to manage the crisis. Whilst some businesses saw dramatic rises in the demand, others saw their sales disappear overnight. Unfortunately, unpredictable demand is here to stay, as Governments across the globe modify social distancing policies (easing and re-introducing lockdown as they go).

Without a crystal ball to predict the future, demand forecasting will remain very difficult. Demand planners therefore need to ensure they have the best tools and forecasting processes for the job.

A good starting point is to invest in a statistical demand forecasting tool, that can continually analyse and adapt to the latest market conditions. Such software will help you understand how volatile your demand is and adjust your replenishment parameters as needed. It can also quickly identify trends in customer behaviour and help you respond accordingly.

Why 30-day rolling average forecasts will fail

Covid-19 has caused erratic customer behaviour. Any inventory manager, therefore, using a basic 30-day moving average to calculate demand will find it almost impossible to generate an accurate forecast. This is because a 30-day rolling average forecast is only suitable for inventory items that have stable demand, where the previous 30 days demand is a good indicator of the future forecast period.

Right now, however, very few items have stable demand. Whether it’s PPE, fitness equipment, DIY items or home furnishings. Even goods with virtually no demand, such as those linked to aviation, will experience fluctuating demand patterns as the industry begins to recover.

Forecasting models therefore need to account for a range of demand patterns and treat items differently, depending on the type of demand they’re experiencing.

EazyStock does this by analysing the demand patterns of all your stock items and assigning each of them to one of nine different demand types. These are based on an item’s position in its product lifecycle. In today’s climate, many items will move around this lifecycle for example, some items in the automotive industry may have moved overnight from growth and maturity to decline, but very soon may return to the growth phase, as sales pick back up.

With demand types assigned, the software then uses the most appropriate statistical algorithm to calculate demand, so every item can be treated differently. For example, items with stable demand will be assigned one algorithm, whilst those with lumpier or slow moving demand will use a more suitable formula.

Reacting to demand trends caused by Covid-19

Going forward, having the ability to monitor demand trends will be key to ensuring stock levels match your customers’ requirements. For example, when restaurants reopen the demand for napkins will begin to rise, whilst as the threat of Covid-19 diminishes, the sale of hand gels will fall.

Demand trends will move fast, and inventory managers using manual forecasting will get left behind. In contrast, a demand forecasting tool will review projections regularly to ensure they reflect the marketplace. With up-to-date forecasts, inventory management teams can then react quickly, maximising sales opportunities, but equally preventing a build-up of excess stock.

Combining demand forecasting technology with human insights

When demand is unstable it’s wise to use qualitative forecast data, or human insights, to augment quantitative inventory forecasts, for the most realistic demand projections possible. A good forecasting tool, such as EazyStock, will provide the base demand forecast. To this you can then add qualitative information, such as feedback from your sales teams, market intelligence and insights from your customers.

Alleviating inaccurate demand due to Covid-19

Unfortunately, with so many unpredictable events taking place, producing forecasts with 100% accuracy is going to be extremely challenging. Demand planners therefore need to find ways to spot when forecasts are inaccurate and put safety nets in place to help prevent consequential stockouts or increases in surplus inventory. Here’s some ways inventory optimisation software can help:

1. Ongoing demand tracking

EazyStock’s demand-to-date alerts track your actual demand throughout the forecast period and alert you when it’s deviating significantly from its projection. You can then investigate the cause, keep an eye on the items and adjust reordering if needed.

2. Demand outliers

At the end of a forecast period, EazyStock compares actual demand against its forecast and highlights extreme variances e.g demand outliers. You can then investigate the reasons behind the data and decide whether to include the outlier in your next forecast, which, in turn will update replenishment parameters.

3. Risk-of-runout

EazyStock generates a daily risk of runout report, that compares the forecasts of all stock items to current stock levels (including stock in transit) and supplier lead times and alerts you to potential stockouts. You then have the information to manage by exception, delving into the detail of each item and confirming if an order should be placed.

It’s also important to identify items that are the most critical to your supply chain e.g those that you rely on to keep production flowing, or that your customers can’t be without. In EazyStock this list can be uploaded and flagged so items can not only be more closely scrutinised, but different replenishment parameters can be assigned to help ensure availability e.g increased safety stock levels, more sensitive stockout alerts etc.

4. Safety stock

Safety stock is a great way to help alleviate issues with demand and supply. Utilising an inventory optimisation tool, such as EazyStock, safety stock levels will dynamically be set, based on a wide range of factors, such as an item’s demand type, it’s forecast, supplier lead times and your target service levels (stock availability). During times of unstable demand, safety stock levels are often increased, to help mitigate unexpected rises in demand or delays to supply

Summary

The Covid-19 pandemic is here to stay for the foreseeable future, affecting consumer behaviour like never before. There’s no denying that even with inventory optimisation software, demand forecasting is going to be challenging. However, inventory management teams that have automation to support their efforts will no doubt be best prepared for the task ahead. Using EazyStock, your business will have a powerful demand forecasting engine that will do all forecasting calculations in the background, while your team focuses on your customers and suppliers. In addition, you’ll have an arsenal of functionality to help prevent stockouts from occurring and stop inventory building up, helping protect your cashflow.

If you’d like to know more about how EazyStock can support your inventory management processes, please contact the team – we’d be happy to discuss your needs.

First published in May 2020. Updated January, 08, 2021