Ordering the right amount of inventory is a balancing act. Too little stock and you risk turning away customers who can’t get what they came for; too much and you’re stuck with high carrying costs.
Inventory forecasting gives you the foresight to stay ahead of demand, ensuring you have exactly what you need before your customers even ask for it. By analyzing sales data, seasonal trends, and market shifts, you can transform your supply chain from reactive to proactive.
This guide shows you how to forecast inventory, with models and software to consider.
What is inventory forecasting?
Inventory forecasting is the process of predicting your future stock needs using historical sales data, seasonal trends, and market conditions. Also known as demand forecasting, its goal is to have the correct number of products on hand to meet customer demand while avoiding excess stock and high carrying costs that can slow cash flow.
Accurate inventory forecasting relies on accurate data in these key areas:
- Historical sales
- Seasonal trends
- Market conditions
- Competitor activity
- Supply costs
Using this data for inventory forecasting takes the guesswork out of your supply chain.
Inventory forecasting vs. replenishment
Inventory forecasting and replenishment are closely related, but they do different jobs. Effective inventory management depends on both: forecasting for the demand outlook and replenishment to turn that outlook into purchasing and allocation decisions. Together, they help your business keep enough stock available without carrying more inventory than you need.
Inventory forecasting
Inventory forecasting comes first. It predicts future stock needs using historical sales data, seasonality, and other demand signals. Here are some questions it answers:
- How much demand is likely next week, next month, or next season?
- Which products are likely to see changes because of seasonality, trends, promotions, or market shifts?
- Where could stockouts or excess inventory happen if demand changes?
Inventory replenishment
Replenishment follows by using that forecast, along with current inventory levels and inventory policies, to decide when to reorder, how much to reorder, and where stock should go next. It answers these questions:
- When should you place the next order or transfer?
- How much stock should each location receive?
- Which purchase orders or transfers are needed to keep inventory within target thresholds?
Why is inventory forecasting important for businesses?
Here are some ways accurate inventory forecasting can help your business:
- Minimize product waste. Inventory forecasting pinpoints when demand for certain products will increase or decrease, reducing the chance you’ll have to throw out old, unsold products.
- Increase savings. Every unnecessary product costs money to store—especially if it’s bulky or oversized. By ordering the right products in the right quantity at the right time, you reduce the costs incurred by overstocking low-demand inventory.
- Optimize cash flow. In a 2025 survey of 500 established Shopify merchants, 43% of Home & Garden businesses reported struggling with cash flow management in their first year, due to bulky inventory, high cost of goods, and seasonal demand swings. Better inventory planning can free up cash that would otherwise be tied up in excess stock.
- Improve customer satisfaction. If you’ve ever tried to buy a shirt only to find that it wasn’t available in your size, then you know the value of having popular products on hand when your customers want them.
- Enhance supplier relationships. More accurate forecasting can help you place more predictable orders and give suppliers better visibility into your needs.
- Reduce stockouts. Any time you have a popular item out of stock, known as a stockout, you are losing revenue. Predicting future demand and understanding when to restock reduces the likelihood of running out of a product when demand is at its highest.
“We know exactly what we need based on real sales data, so we’re not tying up cash in excess inventory or missing sales due to stockouts,” says Tyler Angelos, CEO of Angelus Direct.
5 types of inventory forecasting
- Trend forecasting
- Graphical forecasting
- Qualitative forecasting
- Quantitative forecasting
- Causal forecasting
There are several approaches you can take to forecast how much inventory to have on hand:
1. Trend forecasting
The trend forecasting model predicts buying trends based on how a product’s demand has historically fluctuated. It uses historical sales data and market analysis to project future customer demand.
“Going into the holiday season, I’ll look at what I sold last time,” says Jonathan Grahm, owner of Compartes Chocolate, in a Shopify Masters episode. “Whether it’s Easter, Mother’s Day, Christmas, Valentine’s Day, it’s important for me to take a look and see how many heart-shaped boxes I sold last year on Shopify."
There are two types of trend forecasting:
- Long-term forecasting. Also referred to as macro trends, this method analyzes broad indicators like societal or cultural shifts that affect your customer’s buying habits while excluding seasonal impact and unsubstantiated irregularities (an event that can not be traced to a specific cause).
- Short-term forecasting. Also referred to as seasonal forecasting, this method looks at specific times of the year to forecast for the next six months.
Use your available data to decide if you should choose long- or short-term forecasting.
If your data reveals that your products are susceptible to broader cultural shifts—like the push for organic products influencing the packaged snack industry—then long-term forecasting would serve you better.
If demand for your products fluctuates because of the time of the year—like demand for pool floats peaking in the summer—consider leaning more heavily on short-term forecasting.
2. Graphical forecasting
Graphical forecasting uses historical data to identify market trends and sales patterns on a chart. The visual representation of data makes it easier to identify seasonal patterns, demand spikes around promotions, and long-term growth or decline trends that might not be obvious as rows of numbers on a spreadsheet.
You’d use the same data as trend forecasting. The only difference is you represent it visually. Choose this forecasting type if you prefer to visually discern patterns rather than review numbers as line items.
3. Qualitative forecasting
If your business lacks sufficient historical data for trend forecasting, qualitative forecasting can provide actionable insights. Instead of looking at historical sales data, forecast future demand using:
Qualitative forecasting can also be useful if you’re making major product changes or shifting your business model—instances when historical data won’t be able to reliably predict future demand.
You can use qualitative forecasting to supplement graphical and/or trend forecasting.
4. Quantitative forecasting
Quantitative forecasting projects future demand using numerical data. As a general rule, you’ll need enough historical data to identify reliable patterns over time, especially if your business has seasonal demand.
There are multiple ways to create a quantitative forecast, such as:
Time-series forecasting
This approach uses historical data to map out seasonal patterns and predict future cycles based on the time of year. There are several time-series forecasting techniques:
- Moving averages.This method filters out random blips to reveal broader, long-term trends.
- Exponential smoothing. This technique prioritizes your most recent sales data.
- Linear regression. This method connects the dots between different variables, like how a specific marketing campaign or price change directly impacts your sales over time.
Demand-driven forecasting
This approach uses real-time inventory data captured through point-of-sale (POS) features or similar tech to generate accurate forecasts.
5. Causal forecasting
Causal forecasting predicts demand by looking at the specific events that drive it, rather than just repeating past sales, using factors like:
- Promotions and advertising
- Holidays and shifting seasonal events
- Weather
- Economic trends
- Stockouts and other external signals
Where trend-based methods answer what demand will be, causal forecasting answers why: how much a promotion affects sales; how a late Easter will shift purchase timing; how a weather event or economic shift will move expected volume.
Causal forecasting usually requires more advanced analysis than trend-based methods because it tracks how specific variables actually drive sales. Use it alongside trend-based mapping as part of your overall inventory forecasting process.
How to calculate an inventory forecast
To calculate an inventory forecast, you’ll need to determine key inventory metrics such as sales velocity, average sales, economic order quantity (EOQ), safety stock, reorder points, and lead time demand. Together, these metrics help determine when to reorder inventory and how much to reorder.
Here are the steps to this process:
1. Measure sales trends
A sales trend indicates a pattern in increases or decreases in sales over time. You can analyze this data in the micro (one product over a short period, like 30 or 90 days) or the macro (a range of products over a more extended period, like the last 12 months) to get insight into buying patterns.
For items with high seasonal demand, calculate your average daily sales using the same or a comparable period from previous years.
Start by finding your average daily sales over the past year:
Total number of sales last year / 365 = average daily sales
Look for days with sales that are noticeably over or under this daily average.
2. Calculate lead time demand
Lead time is the time it takes for a supplier to fulfill an order. Lead time demand is the number of products you want to have on hand to avoid running out before your next order comes in.
The inventory formula to calculate lead time demand is:
Average lead time in days x average daily sales = lead time demand
For example, if it takes a water bottle company 10 days to get a new order in and the company has average daily sales of five water bottles, they want to have 50 water bottles on hand to be able to complete expected customer orders.
3. Calculate your safety stock
Safety stock is the extra inventory required to mitigate the risk of stockout. The amount of safety stock you keep should also align with how risk tolerant your business is.
The formula for calculating safety stock is:
Maximum daily sales x maximum lead time in days – lead time demand = safety stock
For example, if our hypothetical water bottle company sold at most 10 water bottles in one day last year and it took a maximum of 15 days to get a new shipment in, their safety stock would be 100 water bottles (10 x 15 – 50 = 100).
4. Set concrete reorder points
A reorder point is the inventory level at which you need to reorder more of a product. The formula to calculate a reordering point is:
Lead time demand + safety stock = reorder point
If the water bottle company has a lead time demand of 50 water bottles and a safety stock of 100, they’d want to reorder when their inventory reaches 150 water bottles.
As your product catalog grows, setting up automated reorder alerts can help reduce the risk of manual tracking errors. Doe Beauty does this with unified data inside Shopify.
Inventory forecasting tools and software
The inventory forecasting software market is divided into two main categories based on business size and complexity:
- For small to mid-sized businesses. If you need core features like low-stock alerts and basic demand tracking, look at Shopify-native apps or standalone tools like syncX, Inventory Planner by Sage, or Prediko.
- For larger enterprises. Businesses with global supply chains or complex manufacturing need advanced ERP systems like NetSuite or Microsoft Dynamics 365. These platforms handle multilocation inventory, scenario planning, and automated replenishment across entire organizations.
Shopify’s inventory management system works with Sidekick, the AI and machine learning feature baked into Shopify. “I am able to pull and move data across stores—instantly comparing customer rates, loyalty program participation, and generating week-by-week performance reports,” says Jamie Evans, head of ecommerce at Jaded London.
Inventory forecasting best practices
Inventory forecasting can be challenging. “We try to do the bare minimum but just enough so that we never go out of stock, and it’s a really really tough thing to do,” says Anaita Sarkar, cofounder of Hero Packaging, in a Shopify Masters interview.
To make forecasting inventory easier and potentially reduce holding costs, follow these tips:
- Use comparable time periods. Compare apples to apples when using sales data for inventory forecasts. For example, if you’re forecasting sales for the second quarter of this year and your business sold 500 units during the second quarter last year, use 500 units as the base for your forecasting model.
- Review trends and marketing variables. Consider if a trend or marketing initiative, like a promotion, affected demand, and factor in these trends and variables into your inventory forecasting. For example, if a marketing initiative ran one year but is not scheduled to run again, then your sales might be lower, even if every other variable remains the same.
- Review all future marketing activities. Align with your marketing team as you put together your inventory forecast. Consider whether you need to carry extra stock to coincide with a promotion or advertising campaign.
- Account for external factors. Consider outside variables such as seasonality, supplier disruptions, economic conditions, or market shifts that could affect demand or inventory availability.
- Segment your inventory. Group products by factors such as sales velocity, seasonality, or profitability so you can apply forecasting methods more appropriately across your product catalog.
- Document your assumptions. Keep a record of the inputs and assumptions behind your forecast so you can review what changed and improve future accuracy.
- Continuously adjust. Forecasting is based on assumptions. As real-life events happen and sales accumulate, adjust your original forecasting parameters (such as time of year, sales data, and lead-time demand) accordingly.
Shopify’s POS features include delivering accurate and automated inventory forecasting. Business owners can forecast customer demand with inventory reports, which track quantities and percentage of inventory sold per day, and retail sales reports, which provide information about POS orders.
“It enables us to understand what sells, how it sells and when it sells. That granular insight is very important,” says Mario Toscano, technology and innovation manager at Bathu.
Inventory forecast FAQ
What role does technology play in inventory forecasting?
Inventory management software provides real-time tracking and ongoing visibility into your stock levels. Technology also provides automation and digitizes an inventory and its processes. Shopify POS, Streamline, and Inventory Planner are leading software for retailers to forecast their inventories.
How do you choose the right forecasting method for your business?
Start by considering how much available data you have. If you have an established company that’s been collecting data for several years, you could produce accurate forecasts with a quantitative approach. Qualitative forecasting would be more appropriate for newer businesses with less data to work with. Consider using a combination of both approaches to produce more informed forecasts.
Can inventory forecasting be automated?
Inventory forecasting automation is possible, so you can forecast more accurately and ensure that forecasting is part of your regular inventory management.
What factors affect inventory forecasting accuracy?
Financial or economic factors, supply chain issues, lead time, and product type (perishable or non-perishable items, for instance) can all impact inventory forecasting accuracy.
How often should you update your inventory forecasts?
Update your inventory forecasts quarterly or when events—such as a busy season or a marketing initiative—require you to adjust your initial predictions.
