BUSINESS STATISTICS CASE STUDY
Case Study Background
A company wishes to improve its e-mail marketing process, as measured by an increase in the response rate to e-mail advertisements. The company has decided to study the process by evaluating all combinations of two (2) options of the three (3) key factors: E-Mail Heading (Detailed, Generic); Email Open (No, Yes); and E-Mail Body (Text, HTML). Each of the combinations in the design was repeated on two (2) different occasions. The factors studied and the measured response rates are summarized in the following table.
Introduction
In this paper we will conduct a design of experiment and analyze the data from the table above. With the experiment we will use a graphical display to present our results from the data analysis. Based off of the analysis we will make recommendations on actions the company can do to increase the rate of responses to emails sent out. Lastly, we will look at a strategy to develop a process model for the company. Let’s first look at the experiment.
Design of Experiment
There are several steps to conducting a design of experiment, the first is an analysis of the basic data (Hoerl, 2012). Based on the above data we can conclude that body type is most effective in getting a response, followed by email opened, and lastly heading type. With body type we can tell that plain text emails have a greater response then HTML bodies. In HTML there may be too much clutter causing the email to open slowly or not at all. The next factor email opened specifically looks at whether the email was opened or not, if a consumer opens the email they will be more likely to respond to it. The last factor which was the least effective on response rate was the email heading, generic headings had a higher response rate then the detailed headings. The tests that received the most responses would therefore be a generic heading plain text email that was opened and viewed by the customer. All of this information was gathered with analysis of the basic data but we can use statistical tools to gain better information and create a better response for our customer (Berenson, 2013). The best graphical display tool for this would be the Interactions Effects Chart (IEC).
Interaction Effects Chart
The interaction effects chart is a graphical display that can be created in a spreadsheet program similar to Microsoft Excel. It is easy to understand and gives management a clear, concise way to see how factors relate to each other. Below you will see the data used and the IEC graph.
Figure 1.
|
Run |
Heading (x1) |
Email Open (x2) |
Body(x3) |
x1 x2 |
x1x3 |
x2x3 |
x1x2x3 |
Repeat Rate 1 |
|
1 |
Generic(-) |
No(-) |
Text(-) |
+ |
+ |
+ |
- |
46 |
|
2 |
Detailed(+) |
No(-) |
Text(-) |
- |
- |
+ |
+ |
34 |
|
3 |
Generic(-) |
Yes(+) |
Text(-) |
- |
+ |
- |
+ |
56 |
|
4 |
Detailed(+) |
Yes(+) |
Text(-) |
+ |
- |
- |
- |
68 |
|
5 |
Generic (-) |
No(-) |
HTML(+) |
+ |
- |
- |
+ |
25 |
|
6 |
Detailed(+) |
No(-) |
HTML(+) |
- |
+ |
- |
- |
22 |
|
7 |
Generic(-) |
Yes(+) |
HTML(+) |
- |
- |
+ |
- |
21 |
|
8 |
Detailed(+) |
Yes(+) |
HTML(+) |
+ |
+ |
+ |
+ |
19 |
|
|
|
|
|
|
|
|
|
|
|
sum+ |
163 |
179.5 |
101 |
168 |
152.5 |
126 |
145.5 |
|
|
sum- |
147.5 |
131 |
209.5 |
142.5 |
158 |
184.5 |
165 |
|
|
ave + |
40.75 |
44.875 |
25.25 |
42 |
38.125 |
31.5 |
36.375 |
|
|
ave- |
36.875 |
32.75 |
52.375 |
35.625 |
39.5 |
46.125 |
41.25 |
|
|
effect |
3.875 |
12.125 |
-27.125 |
6.375 |
-1.375 |
-14.63 |
-4.875 |
|
Figure 2.
|
|
|
|
|
|
|
|
|
|
B x C Interaction B: Email Open; BLO: No, BHI: yes C: Email Heading; CLO: Generic, CHI: Detailed |
|
|
|||||
|
|
|
|
|
|
|
|
|
|
|
|
C LO |
C HI |
|
|
|
|
|
B LO |
1 |
42 |
36 |
|
|
|
|
|
|
2 |
26 |
27 |
|
|
|
|
|
|
Avg |
34 |
31.5 |
|
|
|
|
|
B HI |
1 |
57.5 |
74 |
|
|
|
|
|
|
2 |
22 |
26 |
|
|
|
|
|
|
Avg |
39.75 |
50 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Figure 3.
|
|
A x C Interaction A: Email Body; ALO: Text, AHI: HTML |
|
||||
|
|
|
|
|
|
|
|
|
|
|
|
C LO |
C HI |
|
|
|
|
A LO |
1 |
42 |
36 |
|
|
|
|
|
2 |
57.5 |
74 |
|
|
|
|
|
Avg |
49.75 |
55 |
|
|
|
|
A HI |
1 |
26 |
27 |
|
|
|
|
|
2 |
22 |
26 |
|
|
|
|
|
Avg |
24 |
26.5 |
|
|
Now that we have seen the IEC in action we can take decide on steps to increase response rate.
Recommendations
There are several ways to increase the response rate of emails within advertising. Based off of our analysis the company should focus on sending a plain text body and a detailed heading for their future marketing emails. Other actions the company can take would be to include the first and last name of the customer they are contacting (San José-Cabezudo, 2014). This would increase response because it feels more personalized to the consumer. Lastly let’s take a look at business process strategy.
Strategy
The proper strategy for the process model of this company that will be essential to continued response of its email advertising would be to continue to collect data on the factors presented, and continue to run the same experiment on the data collected regularly. Additional steps the company could take as part of a long term strategy include standardizing and creating a template for emails based on the product. The template should include: Detailed Subject Line, Text in the Body, and the Customers First and Last Name. This template can be used for different products and will help to increase the odds that the email will be opened, viewed, forwarded, or responded to. Finally, once customers have completed an order that began from an email response they should also get the chance to participate in a survey. With their gathered responses we can continue to make adjustments based off of customer preference. Not only will we be able to minimize creation time and create consistency, we will be able to make emails more user friendly and personable.
References Berenson, M. L. (2013). Basic Business Statistics: Concepts and Applications. Sydney: Peasrson. Hoerl, R. &. (2012). Statistical Thinking Improving Business Performance Second edition. Hoboken: John Wiley & Sons, Inc. San José-Cabezudo, R. &.-I. (2014). Determinants of Opening - Forwarding Email Messages. Journal of Advertising, 97-112.
C LO
B LO B HI 34 39.75 C HI
B LO B HI 31.5 50
C LO
A LO A HI 49.75 24 C HI
A LO A HI 55 26.5