📉 Time series analysis
Big Idea: A time series is a set of data points recorded over time (e.g. monthly sales for the past 3 years). By studying the pattern, businesses can predict what might happen next. 📊
Components of a time series
- Trend — the general direction of sales over time (upward, downward or flat)
- Seasonal variation — regular, predictable patterns (e.g. higher sales at Christmas)
- Cyclical variation — longer-term patterns linked to the economic cycle (boom/recession)
- Random variation — unpredictable one-off events (e.g. a viral social media post)
If an exam question shows a sales graph, look for the overall trend first, then identify any seasonal peaks and troughs.
📐 Moving averages
Big Idea: A moving average smooths out short-term ups and downs in the data to reveal the underlying trend. It calculates the average of a set number of periods, then 'moves' forward one period at a time. 🔄
How to calculate a 3-period moving average
Example data: Jan = $10k, Feb = $14k, Mar = $12k, Apr = $16k, May = $13k
- Average of Jan–Mar = ($10k + $14k + $12k) ÷ 3 = $12k
- Average of Feb–Apr = ($14k + $12k + $16k) ÷ 3 = $14k
- Average of Mar–May = ($12k + $16k + $13k) ÷ 3 = $13.7k
- Plot these averages to see the smoothed trend line
The 'moving' part means the window slides forward one period each time. A 3-period average uses 3 data points; a 4-period average uses 4. More periods = smoother line! 📏
Learn what examiners really want
See exactly what to write to score full marks. Our AI shows you model answers and the key phrases examiners look for.
Try AI Feedback Free7-day free trial • No card required
➡️ Extrapolation
Big Idea: Extrapolation means extending the trend line into the future to predict upcoming sales. You assume the past pattern will continue. ➡️📈
- Draw the trend line based on historical data
- Extend it forward to predict future values
- Works well when conditions stay similar
- Very unreliable if the market is changing rapidly
- ✅ Simple and visual — easy to show on a graph
- ✅ Uses real historical data
- ✅ Quick to do
- ❌ Assumes the future will be like the past — dangerous!
- ❌ Cannot predict sudden changes (new competitors, economic shocks)
- ❌ Less reliable the further you extrapolate
Extrapolation is only as good as the assumption that the future follows the past. If something changes, the forecast breaks down!
🧰 Other forecasting methods
- Market research — surveys, focus groups and test marketing to gauge future demand
- Expert opinion — managers, salespeople or industry experts give their estimates
- Correlation analysis — looking at how one variable relates to another (e.g. temperature and ice cream sales)
- Consumer trends — tracking social media, search data and lifestyle changes
Exam tip: The best forecasts combine quantitative methods (numbers, trends) with qualitative methods (opinions, research). One approach alone isn't enough!