Demand forecasting: what it is, why it matters, and how it relates to Process Intelligence

It feeds into and shapes overall business strategy. But predicting what consumers are going to want is notoriously difficult.
Well over a third (37%) of supply chain leaders consider difficulties forecasting demand as among the biggest challenges they’re facing today, according to recent process trends research.

However hard it may be to pull off, organizations capable of accurate demand forecasting give themselves a competitive edge by optimizing resources, reducing costs, and ensuring they consistently meet customer demand.

This post will explore what demand forecasting is, why it’s important, how it works, and how it relates to Celonis Process Intelligence.

What is demand forecasting?

Demand forecasting is the process organizations undertake to estimate the future customer demand for its products or services. Through demand forecasting, enterprises are able to ensure they have the time, capacity, and resources to meet customers’ requirements over a set period.

Demand forecasting can help businesses ensure their supply chains are able to satisfy customers in the short term.
Take businesses that operate with seasonal peaks. Demand for Halloween items like costumes, decor, and candy, for example, is predictably highest around Halloween (October 31st), with consumer spending on Halloween-related items ~5x higher in mid-October than in late July.

This isn’t a shocking statistic, but it shows the need for demand forecasting — businesses that make and distribute Halloween products have to be ready to meet this massive increase in demand every October, and to do so, they need accurate forecasting.

And because having a data-driven forecast for future demand can inform a business’s future growth plans, organizational architecture, operational requirements, customer engagement plans, and more, demand forecasting can also be used to help businesses approach their longer-term decisions more strategically.

What factors and inputs are used to help predict demand?

Businesses use many methods to forecast demand based on a variety of factors and data points. The actual inputs enterprises use for demand forecasting typically include a combination of qualitative and quantitative factors, such as:

• Historical sales data
• Market research data (such as sector growth trends)
• Sales team intelligence (like subjective feedback from teams that directly interact with customers)
• Direct customer feedback
• Economic outlook indicators, news, and current events
• Company growth ambitions (understanding the demand levels the company will need to hit in order to satisfy expansion targets)
• Econometric and statistical modeling

Is demand forecasting accurate?

If you have to ask the question, you probably know the answer. Demand forecasting accuracy inevitably varies.
Even the most extensive, considered forecasting model can be susceptible to error. Technology has made data-driven demand forecasting easier in many ways, but still — no person or program can reliably predict the future, even if AI is helping us get closer than ever.

Customer markets are frequently volatile, subject to unpredictable shifts or unexpected events. Demand planning error jumped to 59% during COVID due to the cascading chain of events caused by pandemic shortages and staffing issues, compared to a pre-pandemic error rate of 45%.

And error isn’t just caused by catastrophes like pandemics or geopolitical events — demand forecasting accuracy issues can also come from sudden changes in consumer behavior. Tiktok, for example, has a pattern of driving sudden shortages in trendy foods that are featured in viral videos or recipes, from cucumbers to Swedish candies.

Even minor shifts in consumer demand can impact the accuracy of a sales forecast. But, while accurate forecasts are the most valuable, even broad indications of likely demand will be helpful.

Why is demand forecasting important to business success?

Fail to plan, plan to fail: There are few areas where this axiom is more applicable than having a clear picture of future customer demand.

An accurate prediction of future sales is key to ensuring the business can give customers what they need – when, where, and how they need it. It’s also central to being able to meet these needs efficiently while controlling costs.
Inaccurate forecasts can mean a business has to respond to unexpected demand spikes by buying materials at an inflated price, paying more to rush shipping and delivery, and subjecting customers to delays that damage the business’s reputation. Faulty forecasts can also over estimate demand, leaving the business holding tons of extra inventory, like the luxury fashion retailers reportedly left with billions of dollars of unsold products in 2023.

Viewed from a more positive angle, accurate demand forecasting can drive greater efficiency that cascades through a variety of business processes, such as:

• Production: With production levels aligned to forecast demand, production costs can be controlled through smarter scheduling, with advance notice of when production should speed up or slow down.
• Inventory Management: Inventory planning is easier when demand forecasting helps businesses understand the right levels of merchandise or raw materials to have on hand to avoid lost sales due to stockouts, or the costly storage of excess inventory.
• Distribution: Demand forecasting can help companies better understand when they’ll have products that need to be delivered to customers, and, subsequently, when they’ll need delivery and logistics services. This foresight can help them maximize on-time delivery without over-commitment of their own resources or contractual obligations to third-party logistics providers (3PLs).

Finally, demand forecasting plays a critical role in organizations’ ability to scale. A reliable forecast minimizes the risk of being unable to meet demand or overcommitting resources for insufficient sales orders. Overproducing or underproducing products can be a huge problem for smaller companies trying to grow with limited resources.

How do demand forecasting solutions work?

Forecasting tools come in many shapes and sizes, but at this point, they’re usually in the form of software. From standalone solutions to broader demand planning platforms, there are cloud-based, on-premise, and integrated options for analyzing demand drivers and data.

Typically, these platforms and solutions assist with demand forecasting by combining and processing:

• Qualitative forecasting data, including forecasts from sales teams, market research data, surveys and interviews.
• Quantitative forecasting data, such as sales data, market research data, economic data, trend projections, and seasonal demand averages.

Having consolidated and validated this information, the software analyzes the data to identify trends and insights that shape robust (and, hopefully, accurate!) demand forecasts.

Does demand forecasting software use AI?

Demand forecasting software is now frequently incorporating artificial intelligence and machine learning to increase the speed and accuracy of projections by:

• Rapidly processing huge volumes of data
• Running sophisticated algorithms and predictive analytics to forecast future demand
• Providing scenario planning, enabling businesses to explore different potential situations (such as shifts in market conditions) and their likely impact on demand

The more pertinent data these systems can successfully integrate, the more accurate and resilient their demand forecasting methods tend to be. And more businesses are actively seeking out AI to help with demand forecasting, with almost half (48%) deploying AI-based systems to help them overcome issues with demand forecasting accuracy.

How does Celonis differ from demand forecasting solutions?

Celonis and demand forecasting solutions are fundamentally different. The primary focus of the Celonis Process Intelligence Platform is process optimization, whereas the goal of demand forecasting solutions is to predict future demand for a product or service over a specific period.

However, in-house supply chain experts at Celonis are frequently asked by customers whether the platform can be used for demand forecasting. Celonis is not designed specifically for demand forecasting, but the Platform can be used in conjunction with demand forecasting tools to bridge the gap between planning and execution.
Because it’s system-agnostic, Celonis can integrate third-party demand forecasting data with other dataflows from across the tech stack, allowing the platform to be used as part of a supply chain command center approach.

The Celonis Platform can monitor all supply chain KPIs and workflows, including all customer demand data. With the capability to ingest existing forecasts, Celonis users can track and compare actual customer demand versus forecast demand.

Demand forecasting solution

• Helps predict future customer demand
• Can use a variety forecasting-related inputs for more accurate demand forecasting outputs
• Shows what demand might be
• Actions and decisions based on forecast output generally take place outside of the forecast solution

Celonis Process Intelligence Platform

• Helps optimize business processes across teams, functions, and roles
• Can use demand forecasts themselves alongside input from any software, system, or program for a real-time look at how processes across the business are running
• Shows what the impact of predicted demand might be across all business processes, and suggests and enables actions
• Can use the Platform itself to act on insights generated by the Platform, uniting visibility and information with actual action

The Platform can analyze the impact demand levels will have on sales targets, Procurement, inventory levels, manufacturing capacity, and Order Management. Rather than predict future demand, Process Intelligence provides the ability to translate demand data into actionable insight. Celonis enables businesses to understand the operational and financial implications of demand, and to react more quickly to unexpected fluctuations in it.

How can Celonis improve demand forecasting?

Celonis isn’t only great for making better use of demand forecast data. The Platform also provides a more robust understanding of how end-to-end processes actually work, which can then be used to inform future demand planning processes.

One major manufacturing client, for example, improved their demand forecasting by using Celonis to identify planning gaps in materials management and to optimize lead time strategies. This not only allowed them to better meet customer demand fluctuations, but also enhanced overall Inventory Management.