.png)
A sales forecast is a prediction of the quantity of a product that will be sold at a specific location during a specific period. As simple as the definition is, its impact is significant: The sales forecast forms the basis for nearly every business decision—from daily production volumes to purchasing and workforce planning.
This article explains the definition, common methods, and what distinguishes good forecasts from bad ones—with real-world examples from the food industry.
Definition: What is a sales forecast?
A sales forecast (or demand forecast) estimates future demand based on historical data and known influencing factors. The three dimensions of any forecast are: what (which product), where (which location or channel), and when (which day, which week). The more granular the breakdown, the more useful the forecast—a monthly figure for the entire product range is of little help to daily production, whereas a daily quantity per item and store is very helpful.
An Overview of the Methods
1. Estimation based on experience. The oldest method: A person makes an estimate based on experience. Strength: draws on contextual knowledge. Weakness: does not scale well across many products, is person-dependent, and is systematically biased—for example, by overemphasizing the most recent days.
2. Naive Methods and Averages. “Like last week” or the average of the last four weeks. Easy to implement, but blind to weather, holidays, and trends—any deviation from the average becomes a planning error.
3. Classical statistics. Methods such as exponential smoothing or seasonal time-series models identify trends and recurring patterns. They are reliable for stable demand, but less effective when many external factors are at play simultaneously.
4. Machine Learning. Modern AI models learn from large amounts of data how dozens of factors—weather, day of the week, holidays, season, location—collectively affect demand. They provide granular forecasts for each item and location and continuously adapt. The article “What Is Machine Learning?” explains how this basic principle works.
Real-world example: the bakery
Hardly any other industry illustrates the value of sales forecasts as clearly as the baking industry: perishable products, daily pre-production, and highly fluctuating demand. A good forecast answers hundreds of questions every day—how many croissants for Store A on a rainy Tuesday, how many braided loaves on the Saturday before the holidays?
GoNina provides exactly these forecasts: daily, by item and location, for the next seven days, as specific ordering recommendations within the usual workflow. Businesses can thus reduce their excess inventory by up to 52%, increase sales by up to 6%, and save up to 12 hours of planning time per month.
How to Recognize a Good Forecast
Breakdown: by product, location, and day—not just monthly averages. External factors: Weather, holidays, and school breaks are factored in. Adaptability: New sales data continuously improves the forecast. Integration: The figures are incorporated into the workflow as order suggestions, not in a separate report. Flexibility: Humans retain the final say in special situations.
Frequently Asked Questions
What is the difference between a sales volume forecast and a revenue forecast?
A sales volume forecast estimates quantities (units, kilograms), while a revenue forecast estimates monetary amounts. For production and ordering, the quantity is what matters—revenue is derived from that based on prices.
How accurate can a sales forecast be?
No forecast is perfect—demand always has an element of randomness. The relevant benchmark is a comparison with the previous method: Compared to experience-based estimates and averages, ML forecasts measurably reduce errors, especially on atypical days.
Does every company need a sales forecast?
Every company has one—the question is simply whether it exists implicitly in someone’s mind or is explicit, data-driven, and verifiable. The more perishable the product and the more volatile the demand, the more valuable an explicit forecast becomes.
Conclusion
A sales forecast is the data-driven answer to the most fundamental of all operational questions: How much of what, for whom? The methods range from gut instinct to machine learning—and the shift to AI is most worthwhile where many products, locations, and influencing factors come together.
You can read about exactly how AI sales forecasts work in a bakery in our comprehensive guide.
