Welcome to Chapter 14 of our inventory management guide. This chapter is fully dedicated to inventory forecasting, which we have mentioned a ton of times throughout our guide.
On a basic level, forecasting helps businesses get the right amount of inventory in the warehouse and not waste space on products they don’t need. However, as we dive deeper into forecasting, you’ll see the benefits are rather significant.
When forecasting inventory, we want to get ahead of rising or falling demand and react accordingly.
The market can change quickly, new ways of reaching customers can boost sales, and your inventory suddenly evaporates.
Inventory forecasting a little too advanced? Head back to the inventory management guide homepage.
What Is Forecasting? A Quick Refresher
Inventory forecasting can also be called ‘demand planning,’ and it is key to a business’s efficiency. Some of the biggest benefits of inventory forecasting are:
- Reduce stockouts and back orders—Forecasting reduces the chance of running out of inventory.
- Reduce holding costs (and overstocking)—Forecasting enables companies to store less inventory.
- Save money and time on labor and warehousing costs—Because you are better prepared for demand changes, you can be more specific with your ordering, meaning you can spend less time dealing with incoming orders.
- Reduce the chance of dead and obsolete stock—Again, because you can order more closely to your demand, you are less likely to order excess inventory and have unsellable stock.
It’s important to point out the difference between inventory replenishment and inventory forecasting, as people can think that the two are one and the same.
Replenishment is the act of reordering, while forecasting is the ongoing process of using data to estimate how much to order. Forecasting enables replenishment.
Large corporations may need to rely on teams focusing solely on inventory forecasting, and some may hire data scientists.
How Does Inventory Forecasting Work?
You need a healthy mix of data analysis, industry experience and awareness, and customer insight to generate useful forecasts.
While it’s not an exact science, the more information you have available, the better your forecast will be—think of forecasting as an ‘educated guess.’
It’s also important to acknowledge that forecasting is influenced by both internal (related to your business) and external (related to the wider world) factors.
Internally, you will need metrics such as lead times and sales trends, set reorder points, calculate safety stock, average lead time, demand, sales history, trends, and much more, which we will get into.
Head back to Chapter 9 if you need a refresher on the different KPIs and metrics for inventory management (many of them are helpful for inventory forecasting).
To forecast inventory, you must know your current inventory levels, factor in any outstanding purchase orders, know your maximum possible stock levels, and sales velocity (how quickly products leave shelves).
It also helps if you know your company’s goals, which may change how much you need to order and any local or global supply chain challenges.
For example, the supply chain disruption in 2022 changed the way many companies ordered from suppliers—many started making fewer orders but with larger quantities.
And if you really want to dig in deep, you can factor in any marketing activities and how they might drive up demand—yours and your competitors.
Whatever your approach to inventory forecasting, effective forecasting should always start with tracking data automatically using software to ensure you have accurate information available.
What Are the Types of Inventory Forecasting?
The four most common ways to forecast inventory are:
- Quantitative forecasting.
- Qualitative forecasting.
- Trend forecasting.
- Graphical forecasting.
Its best practice is to combine these four methods to create a more comprehensive forecast. Combining these approaches enables you to look for points of ‘confluence,’ where multiple signals suggest the same thing.
Let’s look at how each of these methods works.
What Is Quantitative Forecasting?
The bulk of your forecasting may involve quantitative factors. They are easier to measure, consider many of the KPIs and metrics mentioned above, and are a great place to start before looking at other forecasting methods.
Some of the most useful metrics for quantitative forecasting include:
- Replenishment data—The timing of replenishment, the availability of products, and delivery speed.
- Base demand—What is the known customer need for the product you are forecasting—look at sales for the previous period.
- Average inventory—How much inventory do you have at any given time? You may want to keep to this level as it could be optimal for your business.
- Lead time—This can vary due to supplier reliability or external disruptions. Addressing lead time variability is crucial for maintaining optimal inventory levels.
- Manufacturer/supplier lead time—How long does it take to process your order? Is your supply chain vulnerable to delays?
- Sales velocity vs. average sales—Informs you of the sales rate, omitting stock outs.
Quantitative forecasting also involves taking into account seasonality. Seasonal forecasting typically references trends from the last 12 months, though the further back, the better.
At least one year’s data is needed to recognize seasonal trends. Seasonality can also take into consideration things like the weather. The longer a product and business have been around, the easier it is to forecast sales.
Lastly, your forecast may need to take into account inventory costs. You’ll want to ensure that you’re not spending a fortune on inventory and adopt forecasting techniques that ensure those costs are as low as can be.
Head back to Chapter 12 for more on the costs of inventory management.
What Is Qualitative Forecasting?
Qualitative forecasting brings together all other factors that you cannot necessarily quantify. It relies more on ‘market intelligence,’ which can be hard to measure and acquire. Many of these are external factors.
A quantitative forecast will consider things, such as current events that will take place in your forecasting period.
For example, is a company (a competitor or a supplier) about to go on strike? Is an election around the corner? Is a potentially disruptive product about to launch that may make some of your stock obsolete?
Determining qualitative factors requires awareness of current events—watching the news, economic factors that could impact people’s shopping habits, competitors, and suppliers.
It’s important to ask good questions and think outside the box—what could disrupt the normal flow of inventory exiting (and entering) the warehouse? You may uncover upcoming events that might decimate your quantitative forecast.
However, measuring how these qualitative factors impact your forecast can be hard. If you need to order more, how much more? So, it is important to find an agreeable way to measure the impact of these factors.
On an internal level, inventory managers need to work with marketers to be aware of potential spikes in demand generated through marketing campaigns, promotions, and company goals, as mentioned earlier.
Aside from communicating with your internal teams and keeping a watchful eye on the market, you can do things to gain qualitative data for forecasting, such as conducting surveys or other methods where you speak directly with consumers, but these are not always accurate.
What Is Trend Forecasting?
In inventory management, trend forecasting uses historical data and predictive analytics to project potential consumer demand.
Specifically, trend forecasting aims to identify patterns. For example, if the long-term demand for a product is trending downwards or upwards. If a product is trending upward, you may need to order more than the previous month.
Trend forecasting differs from quantitative forecasting as it looks for patterns and directional movements in data.
In contrast, quantitative forecasting uses specific numerical models and data analysis techniques to make precise numerical predictions.
Trend forecasting isn’t just watching sales trends. You can also observe consumer buying behaviors. For example, you can research social media and see if any products are trending—if so, demand may be about to spike.
You can also look into customer purchasing habits—do they buy the same products repeatedly? Are there certain products that consumers typically buy together?
There are also geographical trends that you can look at, such as a product selling better in a certain part of the country than another.
There are two primary types of trend forecasting:
- Top-down forecasting—Good for giving you an idea of your best sellers, but misses out on lots of detail
- Bottom-up forecasting—Provides big insights but misses out on broader details.
It’s important to note that trend forecasting doesn’t take seasonal or external events into account or sales—this is quantitative data.
What Is Graphical Forecasting?
While often touted as the fourth way to forecast inventory, graphical forecasting is more of a way to visualize data you have gathered, specifically trend data.
Graphical forecasting can be useful to view trajectories, peaks, and falls more clearly than simply looking at numbers. It can help speed up the inventory forecasting process.
How to Forecast Inventory? 6 Steps
The best way to forecast inventory is to use the data you have gathered from the above four forecasting methods and weigh the best number. The best forecasting method is based on data and statistics, not what your instincts tell you.
Before you start forecasting, consider what data you have and what data you need to gain. Are there any risks? Are there any wild cards? And always ensure you have high-quality and accurate data.
You will also need forecasting tools and inventory management software to enhance efficiency.
You should also bear in mind that the forecasting process will be different for different companies because not every company will have access to the same types of data.
For example, if you’re an established company, you will probably start by looking at quantitative factors such as historical data and analyzing your current stock levels, then move on to quantitative forecasting.
Meanwhile, smaller, younger companies will have less access to data (particularly historical data) and may start with qualitative forecasting.
1. Decide on a Period You Want to Forecast
Whatever your approach, the first step to any inventory forecasting should be to decide what period you are forecasting.
Most companies forecast for 30 or 90 days or annually, which should include seasonal fluctuations.
The further into the future you forecast, the less accurate it will be. So, you may prefer to keep the date range shorter.
2. Analyze Your Quantitative Data
Once you have decided on your date range, collect all the quantitative data you have related to sales for the same number of days prior.
On top of that, analyze the variety of other KPIs and metrics that will impact ordering, including things like manufacturing replenishment.
You will need software to measure quantitative data accurately and in a timely manner.
3. Identify Trends for That Period
Next, research any trends that may have arisen over the past several months (and even years) that are likely to impact the date range for your forecast. Avoid bias by neglecting downward trends.
Observe yearly recurring trends such as Black Friday, Cyber Monday, and Christmas, and any periods where sales slumps are likely, like in January when demand usually falls following the holiday season.
Remove any enormous spikes or declines that are irregular and not likely to happen again.
These are called ‘outliers’ and can be caused by promotions you did or even freak weather occurrences that are not likely to happen again any time soon.
You can then visualize your trends data as a graph.
4. Clean Your Data
Review your quantitative and trend forecasting data for any anomalies, irrelevant data, or unexpected factors that aren’t helping you forecast and may be incorrect.
Look for patterns and correlations in the data you have collected so far. Does it all make sense, or is it vastly different from what was forecasted for the previous period? Is there a chance something went wrong?
Doing this will help you make more accurate forecasts and prevent problems when replenishing your inventory later.
It’s also a best practice to forecast for best and worst-case scenarios. For example, if sales were to slump over the next 30 days, what would be the lowest amount of inventory you would need to order?
5. Consider Any Upcoming Plans and Events (Qualitative Data)
Find out if there are any upcoming plans and events that may impact your forecasting and encourage more sales, such as campaigns, promotions, etc., or anything that may do the exact opposite.
You should then calculate the need for safety stock and add that number to the forecasted inventory for the period.
Safety stock is more important during periods of large fluctuations.
6. Check the Quality of Your Forecast and Reforecast Frequently
A genuinely effective forecast should be ongoing and evolve as the data you use as the foundation for your forecast changes.
When you habitually stay on top of forecasting, you can get more accurate, valuable data and can react faster to any sudden and unpredictable changes.
As you enter the forecasted period, use real-time data to compare the actual performance with what you originally forecast. (Highlighting once more the importance of a perpetual inventory system over a periodic inventory system.)
Data obtained closer to the forecasted period will always help produce more accurate forecasts.
When you actively update your forecast, ensure communication with other departments, as this information may also impact their processes.
In the long term, large, well-funded companies may establish teams solely on inventory forecasting.
Companies can also adopt metrics like Mean Absolute Percentage Error (MAPE), which can be used to measure the accuracy of forecasts by calculating the difference between predicted and actual demand.
The lower the percentage difference between the two, the more accurate the forecast. A company should evaluate its forecasting process if there is a significant percentage difference.
Bonus: Advanced Forecasting
On a more advanced level, you can look into automated forecasting using computer algorithms and software to analyze historical data to forecast without requiring manual intervention.
Like many things we can automate, automated forecasting speeds up the process and reduces human error.
Other advanced forecasting methods include generating computer simulations, such as the Monte Carlo simulation, which can “predict the probability of a variety of outcomes when the potential for random variables is present,” according to Investopedia.
How to Forecast Demand for New Products?
Forecasting for new products can be challenging. Often, the best way to forecast demand for new products is to use data from similar products and products with similar life cycles.
If you don’t have anything like this available, it’s even more challenging.
You should also consider how previous product launches have gone. Look at the trends for new products—do they start slow and then pick up, or the opposite, they leave shelves quicker?
Key Points From Chapter 14
You’re now a master of inventory forecasting. Remember these key points.
- Inventory forecasting is a central part of inventory management. Forecasting helps reduce stockouts, back orders, holding costs, and time and money on labor.
- Inventory forecasting uses data analysis, industry experience and awareness, and customer insights to gauge how much inventory to order for a predetermined period.
- The four primary types of inventory forecasting are quantitative, qualitative, trend, and graphical forecasting. Using a combination of these methods will yield more effective forecasts.
- Start an inventory forecast by determining the period, then analyze the quantitative and trend data, ensure your data is accurate, and then consider qualitative factors.
- Forecasting demand for new products is tricky, but it can be done using data from similar products with similar life cycles.
In the next chapter, we will tell all there is to know about inventory management contingency planning.