Inventory Forecasting: The Complete Guide (Methods, Formulas & Best Practices)

Most businesses know they should be forecasting inventory. Far fewer actually do it in a way that produces reliable results. The consequences are familiar: a warehouse full of slow-moving stock, or a phone full of angry customer messages about items that are out of stock.

This guide cuts through the complexity. Whether you’re a growing e-commerce brand, a retailer managing hundreds of SKUs, or a logistics manager responsible for a nationwide network, you’ll find practical, actionable guidance here — from the core formulas to the trickier questions like forecasting across different regions or handling a new product launch with no sales history.

Why Inventory Forecasting Is Critical to Your Business

Inventory represents frozen cash. Every unit sitting in your warehouse that isn’t selling is money that can’t go towards marketing, staff, or growth. And every stockout is a sale lost — often permanently, because the customer simply buys from a competitor instead.

Good inventory forecasting sits at the intersection of these two risks, helping you thread the needle between too much and too little. Here’s what it actually delivers:

• Prevents stockouts and overselling. By predicting demand before it arrives, you replenish at the right time — not after the shelf is already empty.
• Frees up working capital. Accurate forecasts mean you’re not tying up cash in inventory you don’t yet need. Those funds can be deployed into growth activities instead.
• Strengthens supplier relationships. Predictable, planned purchase orders replace panicked last-minute calls — giving suppliers time to prepare and giving you real negotiating leverage on price and terms.
• Improves warehouse efficiency. When you know what’s coming in and going out, space allocation, labour scheduling, and receiving processes all run smoother. Learn more about how our logistics services can support this.
• Enables faster response to supply-chain changes. When lead times blow out, freight costs spike, or a product goes viral, a business with solid forecasting can adjust its purchasing plans quickly. One without it is left scrambling.
• Reduces dead stock. Products that won’t sell at full price are identified early, giving you time to discount, bundle, or return them before they become a write-off.

The 5 Main Inventory Forecasting Methods

There’s no single “best” method. The right approach depends on how much data you have, what you sell, and how predictable your demand is. Most sophisticated operations combine two or more of the following techniques.

1. Trend Forecasting

Trend forecasting analyses past sales data to identify growth or decline patterns over time. It can be applied top-down — starting with the overall market and drilling down to individual products — or bottom-up, building from individual SKUs up to an overall business picture. Top-down is useful for understanding your best-sellers and broad market direction. Bottom-up gives you more granular, operationally useful insights. Most businesses benefit from combining both.

Best for: Established products with 12+ months of sales data.

2. Graphical Forecasting

Graphical forecasting plots your historical sales data on a chart so you can visually identify trends, outliers, and seasonal patterns. It’s a powerful complement to trend and quantitative analysis — humans spot visual patterns faster than numbers in a table.

Best for: Communicating forecasts to non-technical stakeholders, and spotting patterns that numbers alone can hide.

3. Qualitative Forecasting

Qualitative forecasting relies on expert judgement, customer surveys, focus groups, and market research rather than historical numbers. It’s essential when data is limited — such as when launching a new product, entering a new market, or reacting to a rapid shift in consumer behaviour.

Best for: New products, new markets, early-stage businesses, and situations where historical data is unreliable or unavailable.

4. Quantitative Forecasting

Quantitative forecasting applies statistical models — time-series analysis, regression, moving averages — to historical numerical data. The more data you have, the more accurate the output. This is the backbone of most professional demand planning systems and inventory management software.

Best for: High-volume operations with rich historical sales data across multiple SKUs and locations.

5. Seasonality Forecasting

Seasonality forecasting accounts for recurring demand patterns tied to the time of year — holidays, weather cycles, sporting seasons, and sales events like Black Friday or end-of-financial-year. It requires at least two full years of data to distinguish true seasonal cycles from one-off anomalies.

Best for: Retail, food and beverage, outdoor products, fashion, and any business with predictable peak periods. See our guide on choosing the right e-commerce platform to support seasonal demand.

How to Choose the Right Method

Core Inventory Forecasting Formulas

These are the building blocks of any solid demand planning process. Understand them individually first, then see how they connect into an end-to-end inventory plan.

1. Sales Velocity

Sales velocity measures how fast a product is actually selling — excluding days when it was out of stock. This is critical because including stockout days in your average artificially drags demand down and causes you to under-order. According to research by IHL Group, inventory distortion including stockouts costs retailers over $1.1 trillion globally each year — a problem that accurate velocity tracking directly addresses.

2. Lead Time

Lead time is the total elapsed time from placing a purchase order to receiving goods into your warehouse. It’s a non-negotiable input for every replenishment decision. Whether you’re shipping domestically or using international sea freight, knowing your exact lead time is what separates a good forecast from a dangerous guess.

3. Safety Stock

Safety stock is the buffer inventory you hold to protect against demand spikes and supply delays. Too little and you stockout during a busy period; too much and you’re tying up cash in stock that just sits there. Read more about safety stock principles on Investopedia.

4. Reorder Point (ROP)

The reorder point is the inventory level that should trigger a new purchase order. When stock drops to this number, you place an order — and your safety stock ensures you don’t run out while waiting for delivery. For businesses using door-to-door courier services, factoring in accurate delivery windows makes ROP calculations significantly more reliable.

5. Economic Order Quantity (EOQ)

The Economic Order Quantity (EOQ) tells you the most cost-efficient quantity to order each time — the point where the combined cost of ordering frequently and holding inventory is at its lowest.

6. Inventory Turnover Ratio

Inventory turnover tells you how many times you sold through your entire stock over a given period. A higher ratio generally signals efficient, healthy inventory management. A low ratio often means too much capital is tied up in slow-moving stock. The Australian Bureau of Statistics publishes business indicator benchmarks that can help you contextualise your turnover ratio against industry averages.

How to Forecast Inventory: A Step-by-Step Process

Here’s a practical, repeatable framework for building an inventory forecast — whether you’re doing it in a spreadsheet or dedicated software.

Step 1: Choose your forecast period

Decide whether you’re forecasting for 30, 90, or 365 days. Shorter windows are more accurate but require more frequent updates. Most operations run a rolling 90-day forecast with monthly reviews, and a separate annual planning cycle for big-picture supplier and budget decisions.

Step 2: Gather and clean your historical data

Pull at least 12 months of sales data — ideally 24. Remove or annotate outliers: one-off promotions, viral product moments, a competitor going bankrupt, or COVID-era demand swings. Your forecast should reflect sustainable, repeatable demand — not anomalies that are unlikely to recur.

Step 3: Calculate your sales velocity per SKU

Use the sales velocity formula (units sold ÷ days in stock) for each product. Never use simple averages — they undercount demand by treating stockout days as though demand was zero, when in reality customers just couldn’t buy.

Step 4: Factor in known variables

Apply adjustments for planned promotions, upcoming marketing campaigns, seasonal cycles, new product launches, and any supply chain changes such as new suppliers or changed delivery lead times. These are all knowable in advance — build them into your forecast before they catch you off guard.

Step 5: Calculate safety stock and reorder points

Use the formulas above for each SKU. Store these in your inventory management system so reorder alerts fire automatically when stock hits the threshold — removing the need for constant manual monitoring.

Step 6: Build your replenishment plan

Based on your forecast, safety stock levels, and current on-hand quantities, calculate what needs to be ordered and when. Always factor in stock that’s already on order and in transit — double-counting incoming supply is a common error that leads to overstock. If you’re managing imported parcels, keep a separate log of shipments in transit by expected arrival date.

Step 7: Monitor, compare and reforecast

Compare actual sales against your forecast on a weekly basis. When reality diverges — either above or below your prediction — investigate the cause and update your assumptions. A forecast that’s never revised isn’t a forecast; it’s a guess that gets progressively less accurate over time.

How to Forecast Inventory Needs Across Regions

One of the most common — and most costly — inventory mistakes is treating your entire customer base as a single homogeneous group. A product might sell strongly in Queensland but barely move in Western Australia. A shipping delay from an Asian supplier might hit your Melbourne DC two days before your Sydney DC. Regional forecasting accounts for all of this.

Key Differences to Account for by Region

• Demand patterns. Consumer behaviour, demographics, and climate differ significantly by region. A winter product forecasted at a national level will overestimate demand in tropical northern Australia and underestimate it in cooler southern states.
• Supplier lead times. If you operate multiple distribution centres, the lead time from the same supplier can vary by 2–5 days depending on which DC is receiving. Whether you’re shipping to a Sydney warehouse or a Perth depot, build region-specific lead time data into your ROP calculations for each location.
• Local events and seasonality. State-based public holidays, local sporting events, and regional promotional calendars create sharp, localised demand spikes that a national forecast misses entirely.
• Freight and logistics constraints. Remote and regional areas in Australia often face different delivery frequencies, higher minimum order quantities, and less predictable service windows — all of which affect the safety stock you need to hold at each location. Our domestic delivery service covers the full Australian network.

How to Build a Regional Forecasting Model

1. Segment your sales data by location — by state, postcode, distribution centre, or store. This is the foundation of everything that follows.
2. Calculate regional sales velocity for your top-performing SKUs in each area. Some products will have dramatically different velocities by region.
3. Set region-specific reorder points and safety stock based on local demand variability and the actual lead time to each specific location.
4. Coordinate inventory across DCs. If one location is forecast to stockout while another has surplus, transferring stock between warehouses using a Sydney to Perth courier or Sydney to Adelaide courier is often faster and cheaper than placing an emergency supplier order.
5. Calculate seasonal indices at a regional level for your most important SKUs. A national seasonal index will be systematically wrong for regions with different climate patterns or different peak periods.

Forecasting Inventory for Seasonal Products

Seasonal products need a fundamentally different approach from evergreen items. Their demand profile isn’t a trend — it’s a cycle. And the consequences of getting the forecast wrong are particularly severe: too little stock means missing the entire selling window; too much means heavily discounting at the end of the season or writing stock off entirely.

The Seasonal Index Method

The most reliable way to account for seasonality is to calculate a seasonal index — a measure of how much demand in a given period deviates from your annual average. Here’s how to do it:

1. Collect at least two years of sales data and organise it by period (month or quarter).
2. Calculate the average sales for each period across all years.
3. Calculate the grand average — the average of all your period averages.
4. Divide each period average by the grand average and multiply by 100 to get the seasonal index for that period.

In this example, Q1 sees 42% above-average demand. To forecast Q1 next year, multiply your expected baseline demand by 1.42. For Q2 and Q3, multiply by 0.66 — you’ll need significantly less stock. Plan your purchase orders and supplier shipments around these numbers well in advance.

Inventory Forecasting Best Practices

The difference between businesses that forecast well and those that consistently struggle is rarely the formula they use — it’s the habits and processes built around the forecast.

Make it a team sport

Inventory forecasting can’t live in one person’s spreadsheet. Your marketing team knows about upcoming campaigns, your sales team hears customer signals, your warehouse staff see what’s actually moving on the floor. Build a regular cross-functional review into your calendar and actively bring those inputs into the forecast.

Compare like-for-like time periods

When benchmarking against historical data, always compare the same period in prior years. Q3 of this year should be benchmarked against Q3 of last year — not a rolling average that smooths over seasonal variation and gives you a misleading baseline.

Use sales velocity, not average sales

This is the single most common forecasting error. Averaging sales over a period that includes stockout days systematically underestimates true demand. Always exclude out-of-stock periods from your velocity calculations.

Set safety stock at the SKU level

A blanket policy of “30 days of safety stock on everything” will leave you with enormous overstock on slow-moving items and dangerously thin buffers on your fastest sellers. Calculate safety stock individually for each SKU, based on its own demand variability and lead time variability.

Use actual lead times, not quoted ones

Pull your actual purchase order history and calculate your real average and maximum lead times for each supplier. Whether you’re using DHL, FedEx, or a local carrier, suppliers consistently give optimistic estimates. Build in the real numbers.

Annotate your data

When you see a spike or a slump in sales data, add a note explaining why it happened. Next year, when you’re looking at the same period and wondering whether to plan for a repeat, that context is invaluable.

Account for promotional pull-forward

A promotion doesn’t just increase sales during the event — it often pulls demand forward from the weeks that follow. Build the pull-forward effect into your post-promotion replenishment plans from the start.

Never stop adjusting

A forecast is a working hypothesis, not a finished document. Review actual vs. forecast weekly for your most important SKUs and monthly for everything else. When reality consistently diverges from your model in one direction, find the broken assumption and fix it.

If managing the operational side of inventory is stretching your team, our end-to-end logistics and fulfilment services can take the burden off so your team can focus on the planning side.

Choosing Inventory Forecasting Software

Spreadsheets can get you started, but they have hard limits. They’re static snapshots, error-prone when formulas break, require constant manual updating, and can’t automatically respond to changes in real-time sales velocity or supplier lead times. Once you’re managing more than a few dozen SKUs across multiple locations, purpose-built software becomes a genuine competitive advantage. For guidance on the broader technology stack, see our post on choosing the right e-commerce platform.

What to Look for

Frequently Asked Questions

What is inventory forecasting?

Inventory forecasting is the process of using historical sales data, market trends, supplier lead times, and known variables such as seasonal demand and planned promotions to predict how much stock you’ll need over a future period. It ensures you have enough product to fulfil customer demand without tying up excess cash in unsold inventory.

What are the four main types of inventory forecasting?

The four main approaches are trend forecasting (identifying demand patterns over time using historical data), qualitative forecasting (using expert judgement, customer surveys, and market research — especially useful when data is limited), quantitative forecasting (applying statistical models to historical numbers), and graphical forecasting (visualising data on charts to identify patterns). Most effective operations combine two or more of these methods.

How do you forecast inventory needs across different regions?

Regional forecasting requires segmenting your sales data by location — state, postcode, or distribution centre — and calculating separate sales velocities, seasonal indices, safety stock levels, and reorder points for each area. The most common mistake is applying a national forecast to regional operations, which predictably results in overstock in some locations and stockouts in others. Our domestic courier network spans all major Australian regions, making regional replenishment significantly faster.

How do I forecast inventory for a new product with no sales history?

Start with comparable existing products in the same category or with a similar customer profile. Analyse the launch trajectory of those products — how quickly did sales ramp up, and what was the month-by-month pattern? Apply that trajectory as your initial forecast, then adjust as real sales data arrives. Complement this with qualitative research: customer surveys, supplier insights, and competitor analysis.

What is the difference between inventory forecasting and replenishment?

Forecasting is the planning exercise — predicting what demand will be over a future period. Replenishment is the operational execution — actually ordering and receiving stock to meet that forecast. A replenishment plan takes the forecast as its input and combines it with current on-hand stock, stock already on order, supplier lead times, and safety stock requirements to calculate exactly what needs to be ordered and when.

What is the reorder point formula?

The reorder point formula is: ROP = (Daily Sales Velocity × Lead Time in Days) + Safety Stock. When your inventory for a product drops to the reorder point, it is time to place a new purchase order. The formula ensures you have enough stock to cover ongoing demand during the supplier lead time, with a safety buffer to handle higher-than-expected demand or a slower-than-usual delivery.

How often should I update my inventory forecast?

As a minimum, review your forecast monthly and update key assumptions — sales velocity, lead times, promotional calendar — whenever something changes materially. For your top 20% of SKUs by revenue, weekly actual-vs-forecast comparisons are worth the investment. Any major external change — a supply chain disruption, a viral product moment, a new competitor — should trigger an immediate review outside your normal cycle.

How do promotions affect inventory forecasting?

Promotions distort your sales data in two directions. During the event, demand spikes — often 2–5 times normal velocity. In the weeks after, demand typically dips as customers who stocked up have no immediate need to buy again. When building your forecast: strip historical promotions out of your baseline so you’re not inflating normal demand; add the expected promotional uplift back in when the next promotion is scheduled; and plan for a post-promotion dip in your replenishment orders so you don’t overstock immediately after the sale ends.