A crystal ball sometimes wouldn’t go amiss in inventory forecasting. But in the absence of a magical tool, your best bet is to manage your stock by using industry techniques and technology.
Here we explain the methods and formulas you can use to improve your inventory forecasting, and best practices for your business when designing and carrying out this process.
We also take a look at the technology being used to make this process easier and more accurate – and how new IT solutions are evolving to make this process even more efficient.
What is inventory forecasting?
Inventory forecasting is when a business uses historical data and analyses trends to predict and plan its future inventory requirements. This ensures it can produce the goods it manufactures or sells in the right quantities to meet demand.
A manufacturer of ice cream, for instance, might slow production during the winter months and increase it in spring in order to be well stocked for summer. Of course, this is a straightforward scenario, and many businesses face far more complicated forecasting challenges.
For most modern companies, inventory forecasting requires a multi-faceted approach that relies on a mix of science, technology and experience.
Businesses must understand what quantity of stock is required, what external factors could influence demand, and how the business can adapt to changes.
And then there are major events that are impossible to predict – like a global pandemic – which throw inventory forecasting into complete disarray.
A simple example of external factors affecting demand is how ice cream becomes more popular in summer. Most inventory forecasting involves multiple, and much more complex, factors
Why is forecasting important in inventory management?
Accurate inventory forecasting is critical for manufacturers and suppliers.
Too little and you may be unable to continue with production or meet consumer demand.
Too much, and you have a number of issues related to oversupply. You may be unable to move it all, have increased storage expenses, have to sell below cost or - in some cases - have to dispose of it.
In short, incorrect forecasting for inventory management will be costly for your business.
That’s because inventory forecasting – as part of inventory management – is essentially business forecasting. It’s how businesses calculate and allocate budgets, develop growth strategies, assign workforce personnel, ensure adequate cash flow and plan for capacity.
The advantages of accurate forecasting include:
- Minimising the risk of lost sales due to product being unavailable
- Less chance of perishable stock spoiling before selling
- Optimising workforce capabilities - i.e. not having more staff than needed or too few
- Better profit margins by not oversupplying products
- Less strain on storage facilities by ensuring appropriate fulfilment flow through warehouses and distribution centres
What are the types of inventory forecasting?
There are four basic types of inventory forecasting, each with their own unique ‘spin’ on how to predict demand for goods.
The type of inventory forecasting you choose will depend on the type of business you run.
1. Demand forecasting
Demand forecasting is one of the most widely used inventory forecast methods.
This type of forecasting involves predicting customer demand by taking into account five main factors:
- Seasonality: What peaks and troughs are there in demand throughout the year?
- Competition: Are there new competitors in the market – or are some dropping out?
- Geography: Does the location of your supply chain allow you to supply goods where they are in demand?
- Economy: Is there recession, which can have an impact on the type and quantity of goods in demand?
- Types of goods: Is your product perishable, meaning you must forecast demand accurately? Or is it a non-perishable product that has a more flexible shelf-life?
Demand forecasting can be broken down into six types. A business can use more than one of these to produce a more accurate forecast:
- Passive: Primarily focused on using past sales data to predict future demand
- Active: Usually adopted by startups or businesses growing rapidly
- Short-term: When a business is gearing up for a sale and/or promotion
- Long-term: A view of 12+ months that helps identify seasonal and annual patterns
- Macro & micro: Considers external factors like economic conditions and competition, then looks closely at consumer demand for the specific industry
- Internal: Forecasts what resources – e.g. staff and equipment – will be needed to meet projected demand
Forecasting inventory helps businesses make critical planning decisions - including how they assign workplace personnel
2. Quantitative forecasting
Quantitative forecasting involves analysing statistics on current and past sales performance to predict future demand. This type of forecasting uncovers patterns by using a series of calculations and well-known methodologies – we'll cover these in more detail below.
The Time Series method and Causal methods are two types of quantitative forecasting.
3. Qualitative forecasting
Qualitative forecasting relies on the opinions and insights of industry observers.
The Delphi method is the most widely used type of qualitative forecasting. This involves collecting opinions from a panel of experts by using several rounds of questionnaires. After each round, the results of each questionnaire are summarised, and the panel gives feedback on the other participants’ answers. This process continues until consensus is reached.
Other qualitative methods include market research, surveying, and consulting with experts and senior professionals in a specific industry.
In qualitative forecasting historical data is still important, but so is consideration of other factors that can disrupt the normal course of events. For instance, demand can be affected by an innovation or a law change – or even a pandemic.
4. Artificial intelligence and machine learning forecasting
As we move through 2021 and into 2022, there’s no ignoring the impact that artificial intelligence and machine learning are having on inventory forecasting.
Market fluctuations that have been enhanced by Covid-19 mean that businesses are looking to make their supply chains more resilient.
As supply chains become more complex, AI and machine learning solutions are being offered that are able to process large amounts of data and take hundreds of variables across different locations into account. These advanced tools are able to analyse trends and make predictions more quickly and accurately than humans.
However there is still room for development. AI and machine-learning inventory software requires a significant capital investment, and takes time and manpower to establish properly.
Recent research also suggests that AI inventory management systems have some way to go before they are adaptable enough to cope with rapid market changes such as those experienced during the Covid-19 pandemic.
AI-driven inventory software is a powerful tool for inventory management that is still developing
Inventory forecasting formulas and methods
Here we take a look in more detail at some of the formulas and methods of analysis inventory forecasting uses.
The safety stock formula
Safety stock is extra stock held by a business to ensure it doesn’t run out of goods for sale in the event of an emergency or a supply chain failure. This ensures manufacturers and retailers can always supply goods to their customers, which maintains customer satisfaction and maximises profits.
The safety stock level for a given product is calculated using figures for daily use levels and lead times:
- Safety stock = (Maximum daily use x Maximum lead time) – (Average daily use x Average lead time)
The safety stock formula is intended to work in conjunction with the reorder point formula (below).
The reorder point formula
Setting a reorder point for your most important SKUs helps you determine when to replenish your stock, so that you have neither too much nor too little of each item.
To find the reorder point for a product, multiply the average daily usage and the average lead time in days to get the demand lead time.
You then add the demand lead time to safety stock (above) to get the reorder point:
- Reorder point = (Average daily use x Average lead time in days) + Safety stock
Bear in mind that demand may increase due to the business expanding or seasonal trends, so your reorder points may change.
The economic order quantity (EOQ) formula
The economic order quantity formula is used to identify the right amount of inventory that needs to be ordered. This minimises the number of times the business has to make orders and ensures that excess inventory is not being stored – which keeps costs down.
The formula to calculate a product’s EOQ is:
- D: The annual demand for the product
- K: The cost of an order per purchase, including handling and shipping
- H: The holding (or ‘carrying’) cost per unit per year
Time series method
Time series analysis takes into account repeating patterns of demand and long-term trends. This method uses historic information to predict the future by looking at averages, patterns and other relevant data. The data can then be used to predict what future demand will look like.
This method is best used by businesses that experience cyclical trends or are trying to understand what external factors cause variations in demand.
ABC method
ABC classification is used to rank products from A to C based on their value in a business. By identifying their most important products according to demand, cost and risk, businesses can decide which of these to prioritise.
Category A is the smallest category, and made up of the most valuable stock. Category B is usually larger, and has products of less value. Category C is usually the largest category and is made up of stock which are of least value.
By using this method manufacturers can focus on ensuring high-demand products are available. It also helps businesses understand customer demand to improve their inventory forecasting.
Methods like time series analysis will give your business the data to make informed decisions about what stock needs to be reordered and when
Inventory forecasting best practices for your business
As with anything in business, you’ll see better results if you follow best practices in inventory forecasting.
Here we list some of the steps you should take to make your inventory forecasting efficient for your business.
Understand the forecast period that is ideal for your business
Should you forecast for one month, 90 days or 12 months?
This time period is the basis for all your planning and without it you’re likely to miss the boat when it comes to achieving the right stock levels at the right time.
Use the right software
Inventory forecasting can be difficult without software fit for purpose.
There are many options available on the market, so ensure you choose the right one for your industry and your business.
Have the right staff
Inventory forecasting relies in part on having experienced personnel to analyse data and make predictions.
You will need staff who can understand and implement the right forecasting methods for your business - and most importantly, interpret data so the right decisions can be made.
Make ecommerce fulfilment a priority
People are shopping online more than ever, and this has seen the emergence of the ‘delivery economy’.
What does this have to do with inventory forecasting? Everything. If you’re unable to meet the needs of your online customers quickly because you don’t have stock on hand, you’ll miss out on sales - and alienate your customers.
Know your ideal reorder point
Make sure you know what the magic number for reordering is by calculating your reorder points (formula above).
Why? Because if you don’t have something in stock immediately, then you’ll lose business. If you order stock too early, you use up cashflow unnecessarily and will have to store it at cost.
Use a variety of methods for inventory forecasting
Every business is different and so is every time period – as demonstrated by the Covid-19 pandemic. For this reason it’s a good idea to use a variety of methods to achieve a more well-rounded inventory forecast.
You might combine quantitative analysis of historic data with other forecasting methods – like market research and consulting experts in the industry – to understand future demand.
Unexpected events – like a global pandemic – can result in rapid changes in demand for products
Include merchant-specific analytics, not just big data
There’s no doubt that big data is powerful, but sometimes this can mean you miss out the smaller – but equally important – details.
For example, there may be a seasonal pattern that suggests demand for ice cream increases in summer. But there may be other factors at play too, such as market position, geographic location, what kind of marketing initiatives are being rolled out and new competitors. These all need to be taken into account before planning for production.
Expect the unexpected
If Covid-19 has taught us anything, it’s that unexpected events can result in radical fluctuations in demand for products – toilet paper, face masks and hand sanitiser are obvious examples.
On a smaller scale, there are plenty of other external factors that can have an effect on demand – something might go ‘viral’ online, a new fad emerges or paid publicity impacts on a product’s popularity.
This demonstrates that manufacturers and suppliers cannot rely on the past alone to predict the future. A more robust inventory forecasting process will mitigate some of the effects of rapid changes in demand.
Inventory forecasting and inventory management software
Inventory management software is a must for any manufacturer or supplier wanting to optimise their stock holdings through inventory forecasting. It gives you the ability to manage your stock without time-consuming and inefficient manual processes.
Unleashed software provides stock alerts to let you know when each product is reaching its reorder point, preventing the risk of stockouts and poor customer satisfaction. You can also access a reorder report for your goods, which allows you to review your stock, check what is available, what has been allocated to an order, and what is being purchased.
Having software as a single source for your business’ inventory also means you can easily access and analyse historic data as part of your inventory forecasting process.
Looking to 2024 and beyond, we will no doubt see inventory management software evolve further. AI and machine-learning software have undergone some exciting developments in recent years, and these will continue to be refined and become more accessible.
