1-SALES-FORECASTING.pptx

Edexcel International A Level Business 333 Decision-making techniques 1 QUANTITATIVE SALES FORECASTING Revisionstation Calculators will be needed for this topic From the specification a) Calculation of time-series analysis: moving averages (three period/fou...

Edexcel International A Level Business 333 Decision-making techniques 1 QUANTITATIVE SALES FORECASTING Revisionstation Calculators will be needed for this topic From the specification a) Calculation of time-series analysis: moving averages (three period/four quarter). b) Interpretation of scatter graphs and line of best fit: extrapolation of past data to future. c) Limitations of quantitative sales forecasting techniques. What is quantitative sales forecasting? (QSF) QSF is a statistical technique which uses data to make predictions about the future (in terms of sales not the weather etc.) The method that your exam board would like you to know about is called “time series analysis” What can a business do with QSF information? Once a business has carried out time series analysis they will use this information to; 1 ） Organise production 2 ） Organise resources in the business e.g. employees, premises, raw materials 3 ） Organise marketing to back up the sales predictions What can a business do with QSF information? The four main components are: 1)Trend 2)Seasonal fluctuations 3)Cyclical fluctuations 4)Random fluctuations Calculations Identifying the trend Calculate three period moving average Step 1 – smooth the raw sales data from Date Sales the table by calculating a 3 year moving average. Take the first 3 months data and 2007 400 calculate an average: 2008 500 2009 770 2007 400 2008 500 2010 900 2009 770 2011 600 2012 700 400 + 500 + 770 1,670 _____________ = ____ = 556.6 2013 1,100 3 3 2014 1,500 Calculate three period moving average Step 2 – Now leave out the first year and Date Sales calculate the average for the next three years 2007 400 2008 500 2008 500 2009 770 2009 770 2010 900 2010 900 2011 600 2012 700 500 + 770 + 900 2,170 2013 1,100 _____________ = ____ = 723.3 3 3 2014 1,500 Calculate three period moving average Step 3 – Now add all your calculated averages to the table: Date Sales 3 year moving average 2007 400 2008 500 556.6 2009 770 723.3 The new calculation goes 2010 900 756.6 next to the 2011 600 733.3 middle year 2012 700 800 2013 1,100 1,100 2014 1,500  Plot a graph from your 3 period moving average Date Sales 3 year moving average 2007 400 2008 500 556.6 2009 770 723.3 2010 900 756.6 2011 600 733.3 2012 700 800 2013 1,100 1,100 2014 1,500 Notice that the data is now “smooth” and so you can see trends occurring This will help you to make more accurate sales forecasts for your business as it smooths out any large fluctuations in data which may be down to weather or recession etc. 4-year period moving average Centring is comparatively hard when it’s 4-year period moving average. 4-year period moving average  4 quarter moving average Sales Look how the process smooths out the line so predictions can be made and production scheduled correctly Dates E Commerce Global Sales Figures 2010-2014 2010- 1,336 2016- 4,248 2022 – 7,476 2011-1,548 2017 – 4,988 2023 - 8,034 2012 -1,845 2018 – 5,311 2013 – 2,382 2019 – 5,784 2014 – 2,982 2020 – 6,330 2015 - 3,352 2021 -6,876 Kambala Farm Graph Predicting the line of best fit from the trend Plotting a graph –Line of best fit Variation from the trend Variation We can find how much variation there is from the Date Sales 3 year Variation moving in each trend by calculating: average year 2007 400 2008 500 556.6 2009 770 723.3 And then calculate the 2010 900 756.6 average. 2011 600 733.3 Finally, predicted value 2012 700 800 2013 1,100 1,100 may be more accurate by 2014 1,500 adding the average variation. Seasonal variation Seasonal variation If based on trend, you predict the sales for the fourth quarter is 💲470000, a more accurate prediction might be to calculate the seasonal average variation in the fourth quarter: And subtract this figure. Limitations Limitations Relies on what has happened in the past continuing to happen, and historical data is not always a good indication of what might happen in the future In high technology markets change happens rapidly and products have a short product life cycle, therefore extrapolation can be misleading It is time-consuming and complex and is only as reliable as the data put in Use of moving averages doesn’t take into account how recent the data is Doesn’t link with corporate objectives Interpretation Scatter graphs Time-series analysis only describes what is happening. Casual modelling tries to explain data, usually by finding a link between a set of data and another. As a business director you need to find out if you need to put up the budget in a marketing department. Does more advertising guarantee more sales? If it does we call that correlation. Finding a correlation  Some examples of scatter graphs Sales Sales Spend on advertising Spend on advertising Sales Sales Spend on advertising Spend on advertising  Interpretation of scatter graphs The correlation depends on the angle of the data points and how close together they are Sales Sales Negative correlation No budget increase! Strong positive correlation Raise the marketing budget Spend on advertising Spend on advertising Sales Sales No correlation More research needed Weak positive correlation Keep budget same as last year Spend on advertising Spend on advertising Sample practice questions Case study for question 1  Sample question 1 Evaluatio Knowledg Applicati Analysis n e2 on 2 3 3 Answer sample question 1 Answer sample question 1 How to level sample questio n1 Glossary Time series analysis; a management tool to make predictions from past data Scattergraph; a graph which shows the performance of one variable (sales) against another (spend on marketing) Cyclical; changes in the data due to large changes such as a recession Correlation; a relationship between two sets of variables Line of best fit; a line that can be drawn through a series of data to look for a trend Extrapolation; when data is stretched out over a line of best fit to predict what will happen in the future