Small retailers should consider the use of a control chart
to help determine whether a marketing campaign, such as offering discounts or
gift cards, has any effects on sales.
Control charts can be used to show where a statistically significant
variation from a normal pattern of events exists. Daily sales totals are an example of a normal
pattern of events. If the daily sales
totals increase by a statistical significant amount above the normal variation
in sales from day to day, at the same time that a marketing initiative is
implemented, the increase is likely due to the marketing initiative. If no significant variation is seen, then the
marketing initiative likely has had no effect on increasing total sales.
Control charts creation and use do involve some complexity,
and therefore time is required in learning about and dealing with the complexity. Certain aspects of the use of control charts
and how best and how not to apply them can be beyond the abilities within a
small retailer’s shop. However, I
believe other uses of control charts, for example, in charting daily sales
against statistical significant limits variability for sales, could be within
the resources that a small retailer might want to expend. One reason for this is that accounting software,
such as QuickBooks, can easily export daily sales data to Excel. Once in Excel, a control chart, based on that
daily sales data, should be relatively easy to create, after the initial
understanding of how to do so is learned.
I did some Internet research to find a few websites that
provide introductions to a basic use of control charts, using Excel to create
control charts, and how control charts can be used with daily sales data to
show effects, or lack of effects, of an external factor, such as a marketing
camping, on the sales data.
This link (click here) takes you to a CPA Journal article
that provides a basic definition of a control chart and what can be shown on a
control chart. Data points falling
outside of the upper and lower control limits show that likely something is
affecting the data other than normal variation.
This link (click here) takes you to a ProfTDubYouTube video
that demonstrates how to create a control chart in Excel.
This bizmanualz site (click here) uses a series of weekly sales
data to demonstrate the various computations that are used in creating the
average line, the upper and lower control limits, and the moving range
average. Graphs are used to show how
these computed values are used on a control chart to reach conclusions about
the graphed data.
Two websites show well how a control chart can demonstrate
that graphed data is being affected by external factors. Click here to go to a BPI Consulting site. Go to the bottom to see how a control chart
shows that on-time airline arrivals and departures have improved as of a
certain date due to changes made by the airlines. That several consecutive data
points fall on the same side of the average line indicates a non-normal
variation. The biamanualz site above
provides more on this consecutive rule.
The second website (click here) has an article written by Australians
Phil Cohen and Onno van Ewyk that provides a clear explanation of how data points
on a control chart might show unusual affects on the data.
My blog article here focuses on one use of a basic control
chart – to show how daily (periodic) sales might be affected by an external
event such as a marketing campaign. Many
more uses of control charts exist. Their
uses to show information about processes is only limited by the number of
processes that exist and the need to gain some insights about those processes.
This article also deals with a very basic statistical used of
a control chart, such that smaller retailers might be able to use a control
chart with available resources. Much
more sophisticated uses of control charts, requiring much more sophisticated computations
and statistics, have been developed, which also greatly extend the uses of
control charts.
A lot can be found on the Internet about these additional uses
of control charts.
No comments:
Post a Comment