What is Business Forecasting?
Business forecasting is the process of using analytics and experience to make predictions about future customer/consumer demand. The goal is to go beyond knowing what has happened to arrive at the best assessment of what will happen in the future so a company can make optimal business decisions, whether that be operational or strategic. Business forecasting incorporates a lot of different data and viewpoints, uses forecasting tools for modelling, and generates numbers (forecasts) that be used in multiple areas of the business.
What is Demand Planning?
Demand planning is the process of identifying and managing customer/consumer demand for a company’s goods or services and formulating responses to meet that demand. The idea is to balance demand and supply, i.e. serving the customer with the products they want while optimizing the operational elements that go into it.
People use the terms ‘demand planning’ and ‘forecasting’ almost synonymously but there are some differences. Demand planning is the process that drives operational supply chain activities like resource planning, production, logistics, and inventory policies. Forecasting generates the numbers used to inform those activities.
Demand planning is typically manifest in cross-functional processes like Sales & Operations Planning (S&OP) or Integrated Business Planning (IBP) that bring different functions together to decide on what the company can deliver and manage the trade-offs between Production, Supply Chain, Finance, Sales & Marketing etc.
Whatever you call it, you’re trying to predict what a company will sell in the future to successfully be able to supply it when it’s needed.
What Happens When a Company Doesn’t Have Good Forecasts?
If you have bad demand forecasts you may make poor decisions. If you underestimate demand, it can result in lost sales or, even worse, lost customers. If you overestimate demand, it can mean wasting money on inventory you can’t sell and tying up capital that could be better utilized elsewhere.
With a good forecast you give the customer what they want, when they want it, thereby maximizing sales and helping deliver on the strategic goals of the company. With an idea of what’s going to happen before it occurs, you can set inventory policies, set production schedules, determine investments, predict market impacts, control costs, and understand the lifecycles of your products.
What are the Key Steps in Demand Planning?
Demand planning is about more than just a number – it’s a process with a lot of different elements.
Data Collection: Data can come from multiple sources. We must understand what exactly is out there as far as inputs and insights and know how we can bring those into the forecast. Data typically includes historical sales data and qualitative information from Sales about key customers and from Marketing who can reveal how promotional activity will impact demand.
Data Analysis: The data you get won’t always be clean and usable in its current format it will require some preparation before analysing it. We need to look for anomalies in the data as well as formatting issues, determine what data is relevant and what isn’t, and make sure we’re using the right amount of data.
Forecast modelling: Multiple time series methods can be used to take the data, extrapolate it forward, and arrive at a forecast. Increasingly companies are turning to advanced systems to do machine learning and AI which use a wider range of data and automate much of the process.
Gaining Consensus: A challenging part of the process for a lot of companies is arriving at one number used by the different functions. You need everyone on the same page in terms of what you think is going to happen in the future – and collaboration is fundamental to this. This where collaborative planning forums like S&OP and IBP come in.
Communicating the forecast assumptions: This is often overlooked. We need to explain the expected result (forecast) and the reasons behind as this is key to those forecasts being trusted and therefore used across the business.
What Data is Used in Business Forecasting?
It can be internal data such as sales orders, or external data which a lot of companies are starting to look at now. External data includes customer information, macro information, and demographic data, as well as causal information like sales promotions or weather data.
Data is either structured (easily managed in a spreadsheet and easily accessible) or unstructured (not easily managed in a spreadsheet and often difficult to access). Unstructured represents over 85 percent of the data out there and includes data from social media comments, product reviews, and audio and video content.
What Forecasting Models are Used in Business Forecasting?
There are a lot of different models available. This is because there’s a lot of different types of data out there which require different forecasting approaches. At one extreme we have pure qualitative and knowledge based judgements. This could be a sales team giving their own estimate of sales and then you’re aggregating those things up. At the other extreme you have pure quantitative approaches like machine learning with less human judgement and intervention.
In the middle there are various types of Time Series methodologies and causal models. There’s no right or wrong model or approach – rather we must choose the best approach for the data we have and the resources and time we have to generate a forecast. According to IBF research, right now most companies use Time Series types of data in their modelling and their preferred method is exponential smoothing. Does that mean exponential smoothing is the best? Not necessarily, but it is versatile method and it’s good for a lot of Time Series data.
What is Bias in Forecasting?
Bias is consistent over-or under-forecasting. It can be conscious or unconscious. For example, Sales may always forecast higher sales numbers because they want the inventory on hand in case they make the sale, or Finance may always push the number down to to avoid tying up cash in inventory. Whether it is high or low, bias is dangerous and gives a false picture of the future. It creates bad decisions and deteriorates the trust in the forecasting process. Bias is actually often worse than uncertainty.
To learn the fundamentals of business forecasting and demand planning, join us for IBF’s Chicago Demand Planning & Forecasting Boot Camp from March 15-17, 2023. You’ll learn how to forecast demand and balance demand and supply from world-leading experts. Click here for more information.