Home› Unit 4› 4.3 Sales Forecasting
4.3 HL Only

Sales Forecasting

Why forecasting matters, how it's done, and the consequences of getting it wrong.

Learning Goals
  • Explain why businesses carry out sales forecasting HL
  • Describe methods of forecasting: extrapolation, time series analysis, simple linear regression HL
  • Evaluate the benefits and limitations of sales forecasting HL

Why Forecast?

A sales forecast is an estimate of future sales over a given period. If a business can make a good forecast, it can:

  • Plan capital needs — how many stores, factories, or staff will be needed
  • Plan production — how much to make or stock, including seasonal variation
  • Plan staffing — including part-time and seasonal hires
  • Control finances more effectively — more accurate budgets and cash flow forecasts
  • Secure external finance — lenders and investors want to see credible forecasts

Methods of Sales Forecasting

Market research

Primary and secondary research can be used to estimate future demand — asking potential customers about their intentions, or analysing industry trend data.

Extrapolation

Extending a historical trend into the future using a line of best fit. Assumes the past is a reliable guide to the future.

Time series analysis

A more sophisticated form of extrapolation that identifies and adjusts for patterns in data:

  • Trend — the overall direction of the data over time
  • Seasonal variation — regular fluctuations within a year (e.g. higher sales in December)
  • Cyclical variation — longer-term fluctuations linked to the economic cycle
  • Random variation — unpredictable, one-off fluctuations

Simple linear regression

A statistical method that finds the line of best fit through a scatter plot of data, using the formula y = mx + c. Key concepts:

  • Scatter diagram — plots two variables to look for a relationship
  • Correlation — the strength and direction of a relationship (positive, negative, none)
  • Extrapolation — using the line to predict values outside the range of existing data (less reliable the further you go)
  • Correlation vs causation — two variables can be correlated without one causing the other
  • Dependent vs independent variable — the independent variable (x) is what you control or observe; the dependent variable (y) is what you're predicting

Benefits of Sales Forecasting

  • Reduces uncertainty and supports more confident decision-making
  • Enables proactive planning rather than reactive crisis management
  • Improves resource allocation across production, staffing, and finance
  • Supports external stakeholders (investors, lenders) by demonstrating planning rigour
  • Helps identify seasonal patterns that the business can prepare for

Limitations of Sales Forecasting

  • Garbage in, garbage out — forecasts are only as good as the data and assumptions behind them
  • Inaccuracy of predictions — the future is inherently uncertain; external shocks cannot be predicted
  • Overreliance on historical data — past trends may not continue, especially in fast-changing markets
  • Lack of data for new businesses or new markets — extrapolation requires historical data that may not exist
  • External influences — economic change, competitor actions, regulatory shifts, or pandemics can invalidate forecasts quickly
Recap — what you should know
  • Forecasting helps plan production, staffing, capital, and financing
  • Three main methods: market research, extrapolation/time series, simple linear regression (y = mx + c)
  • Time series analysis decomposes data into trend, seasonal, cyclical, and random variation
  • Correlation does not imply causation
  • All forecasts carry limitations: data quality, external shocks, rapidly changing markets
Practice Exercises
Why forecasting matters:

Think about a business preparing to open a new restaurant.

  1. List five specific decisions the owners would need to make before opening. For each, explain how a sales forecast would help.
  2. What data would they use to build their forecast? Where would this data come from?
  3. What could make their forecast significantly wrong?
Case Study — Nintendo Wii U: Over-forecasting:

Nintendo's Wii U console was launched with significant optimism — the company initially forecast sales of 9 million units in its first financial year. However, the console failed to capture the mass-market appeal of its predecessor, the Wii. Nintendo was forced to slash its forecast dramatically to just 2.8 million units, and warned it would fall to a net loss. The over-optimistic forecast had led to high production commitments, marketing spend, and inventory build-up that could not be unwound quickly.

  1. Why might Nintendo have over-estimated demand for the Wii U?
  2. What are the financial consequences of over-forecasting for a manufacturing business?
  3. What would better forecasting have required Nintendo to do differently?
Case Study — Target Canada: Forecasting without data:

In January 2015, Target Canada announced it would close all 133 stores, less than two years after opening. A central cause was catastrophic inventory management rooted in poor forecasting. Target Canada had no historical Canadian sales data to guide purchasing decisions. Instead, it used forecasts developed at US headquarters — based on US consumer behaviour — rather than Canadian market conditions. As a result, Target Canada ordered far more products than it could sell, leading to distribution centres at maximum capacity while shelves were simultaneously under-stocked in the wrong items.

  1. Why is applying forecasts from one market to another country particularly risky?
  2. What primary market research could Target have done before entering Canada?
  3. What does this case tell us about the importance of the source of forecast data?
Case Study — Zara: Using technology to improve forecasting:

Zara implemented RFID (Radio Frequency Identification) technology across its stores and warehouses from 2014. Each product is tagged with an RFID chip embedded in the security label, allowing Zara to track every item from production through to sale in real time. Before RFID, Zara's staff spent significant time manually counting stock. After implementation, replenishment decisions became faster, more accurate, and data-driven — directly improving the accuracy of demand forecasting and reducing overstock and stockout situations.

  1. How does real-time stock data improve the quality of a sales forecast?
  2. What are the costs and benefits of implementing RFID at the scale Zara operates?
  3. How does Zara's approach contrast with Target Canada's failure? What is the common underlying lesson?