The company manufactures a range of horticultural
products and sells them through various distribution channels including
the major DIY outlets and garden centres.
There had been a belief in some parts of the business that sales were
affected by weather variations, but this understanding was not shared
in all departments. Indeed some individuals expressed the
view that bad weather was used by sales executives as an excuse for
poor sales performance.
A project was instigated to determine if weather effects really existed
and to quantify them so that they could be of benefit for demand
analysis and sales forecasting. Forecast Solutions was
selected to carry out the work.
"With
the help of Forecast Solutions we now understand the considerable efect
that weather variation has on sales of our products. As a result
we are better placed to interpret recent events and make the best
possible forecast of future demand"
Steve Harper, Sales Director, William Sinclair Horticulture
Historical weather data was sourced in the form of population weighted
national data for the UK. Statistical analysis was carried
out in order to examine the influence of weather variation on product
group sales.
In view of the inherent seasonal patterns of the markets concerned, the
natural seasonality of the weather and the strong interdependence of
the weather elements involved, a great deal of care was necessary in
the analysis. Various time lags in the weather data were
explored in terms of the explanatory capability and the statistical
robustness of the results.
shifts in sales volume or market
share.
Integration of results with the sales forecasting system
Not only was it was established beyond doubt that that
weather variations affected the company's sales, but it also proved
possible to quantify the relationships for most product
groups. Simple models were developed to explain the
percentage change in sales caused by unit deviation from average
seasonal weather patterns, with temperature, rainfall and sunshine
hours featuring in the models.
The results of the weather sensitivity analysis have been incorporated
into the demand forecasting system to give a more effective
interpretation of historical variances and an improvement in forecast
accuracy. Careful use is also made of the short-term weather
forecast in considering possible forecast outcomes.
