Home › Learning Hub › IB DP Business Management › 4.3 Sales forecasting
4.3

Sales forecasting

Unit 4 · Marketing · Higher level only

This topic is higher level only. A sales forecast predicts future sales so a business can plan production, staffing, stock, cash flow and budgets. The syllabus asks you to evaluate the benefits and limitations of forecasting; this page also shows how a moving-average trend is built, because it underpins most forecasting questions, and simple linear regression is in the HL toolkit.

🎯What you need to be able to do

  • HL Evaluate the benefits and limitations of sales forecasting.
  • HL Interpret a sales trend and seasonal variation (and, with the toolkit, extrapolation and linear regression).

📚The business management

What a sales forecast is

A sales forecast is a prediction of future sales volume or revenue over a given period, based on past data, market research and expected changes in the market. Past data usually contain:

  • a trend: the underlying long-term direction;
  • seasonal variation: regular ups and downs within a year (tourism peaks, festive seasons such as Ramadan and Lebaran, school terms);
  • cyclical variation: longer swings with the business cycle;
  • random variation: one-off events (a volcanic eruption closing an airport, a viral post).

Finding the trend: moving averages

A moving average smooths out seasonal and random variation. For quarterly data, add four consecutive quarters (a four-quarter moving total), then average two neighbouring totals and divide by 8 to centre the result on a quarter. The seasonal variation for a quarter is actual sales minus trend.

Quarterly sales over three years swing up and down seasonally, from 40 to 99 thousand dollars. A red centred moving-average trend line runs from 60.75 in quarter 3 to 73.00 in quarter 10, rising about 1.75 per quarter.
The trend removes the seasonal pattern and shows the underlying growth.

The trend can then be extrapolated (extended) into the future, and the average seasonal variation added back for each quarter, to produce a forecast.

Benefits of sales forecasting

  • Operations: plan production capacity and stock levels, avoiding shortages or waste.
  • HR: plan recruitment, temporary staff and training (workforce planning).
  • Finance: produce cash flow forecasts and budgets; support loan applications.
  • Marketing: time promotions and set targets.
  • Strategy: reduce uncertainty in decisions such as expansion.

Limitations

  • Based on the past, which may not repeat: new competitors, changing tastes, technology and shocks (a pandemic) break trends.
  • The further ahead, the less reliable the forecast.
  • Data quality: new products and new businesses have little history.
  • Qualitative factors (a competitor’s campaign, a change in regulation) are hard to quantify.
  • Bias: managers may be over-optimistic to justify plans.
  • Costs time and money; complex methods give an illusion of precision.

✏️Worked example HL

A beach café’s quarterly sales ($000) for three years were: 40, 62, 55, 83; 46, 70, 61, 91; 52, 76, 69, 99. (a) Calculate the centred trend for Q3 of year 1. (b) The trend rises by about 1.75 per quarter to 73.00 in Q10 (Q2 of year 3), and the average seasonal variation for Q4 is about +21. Forecast sales for Q4 of year 3 (Q12). (c) Evaluate the reliability of the forecast.

(a) Four-quarter totals: 40 + 62 + 55 + 83 = 240; 62 + 55 + 83 + 46 = 246. Centred trend = (240 + 246) ÷ 8 = 60.75.

(b) Trend for Q12 ≈ 73.00 + 2 × 1.75 = 76.5. Forecast = 76.5 + 21 = about $97 500 (actual turned out to be 99, close to the forecast).

(c) Fairly reliable in the short term: a clear, steady trend and a regular seasonal pattern. But it assumes tourism continues to grow at the same rate; a new competitor, a drop in arrivals or bad weather could break the pattern, so the café should combine it with market knowledge and update it each quarter.

Check it. Seasonal variations for the same quarter should be similar across years; here Q4 gives +20.5 in year 1 (83 − 62.5) and +21.75 in year 2 (91 − 69.25), averaging about +21, a consistent pattern.
Forgetting to centre a four-quarter average, which leaves it between two quarters. And treat any forecast as an estimate, not a fact.

📝Practise

All HL.

1. [2 marks] Define the term sales forecasting.
Predicting future sales (volume or value) over a period, using past data, market research and expected changes in the market.
2. [2 marks] Distinguish between the trend and seasonal variation.
The trend is the long-term underlying direction of sales; seasonal variation is the regular, repeated rise and fall within each year around that trend.
3. [2 marks] Sales in four consecutive quarters are 120, 150, 110, 180, and the next quarter is 130. Calculate the two four-quarter moving totals and the centred average.
Totals: 120 + 150 + 110 + 180 = 560; 150 + 110 + 180 + 130 = 570. Centred average = (560 + 570) ÷ 8 = 141.25.
4. [4 marks] Explain two ways a hotel could use a sales forecast.
(1) Staffing: hire seasonal staff before forecast peaks and reduce hours in quiet periods, controlling labour costs. (2) Pricing and cash flow: raise rates in forecast peak periods and plan an overdraft or promotions for quiet months.
5. [4 marks] Explain two reasons why a sales forecast for a new product may be unreliable.
(1) There is no past sales data, so the forecast relies on market research and comparisons that may not reflect actual behaviour. (2) Customer response, competitor reactions and the effect of promotion are uncertain; early adopters may not represent the wider market.
6. [10 marks] Discuss the view that sales forecasting is of little use in a fast-changing market such as mobile games.

For the view: trends change quickly, hits are unpredictable, product lives are short, and past data quickly become irrelevant.

Against: forecasts still guide server capacity, marketing spend and cash flow; short-term forecasts using real-time data (downloads, user retention) can be accurate; scenario planning (best, expected, worst) manages uncertainty.

Judgment: long-range forecasts are unreliable, but frequent, short-range, data-driven forecasts remain valuable if updated and used alongside judgment.

🔗Go deeper — other people’s work

These are external resources, not mine. If one stops working, tell me and everything above it on this page still stands.

  • BPS tourism statistics — monthly foreign arrivals to Bali, to see seasonality in real data.
  • Khan Academy — Fitting a line to data and extrapolation.
  • Tutor2u — moving averages and sales forecasting.