Professional Documents
Culture Documents
Design of Experiments For Orelbuy
Design of Experiments For Orelbuy
Design of Experiments For Orelbuy
Introduction
We are designing an experiment with Orel Corporation employees, persuading them to download,
use, and promote the OrelBuy app. OrelBuy marketing team is planning a campaign to run a day
event with fun and game and also communicating the user benefits for Orel employees using the
app. Some of the benefits are employees are entitled to buy their day-to-day stuff at discounted
rates, they could promote the app among their friends and family, we call it the golden circle and
receive reward points. The experiment will be a before-after design experiment. Not going in for a
full factorial as it would be difficult to quantify as multiple parameters may be altered.
Experiment Design
In this experiment, the independent variable is the promotional banner that is developed for the
campaign; the dependent variable is the app downloads that ultimately relates to sales. The
experiment will run in most of our locations in Meegoda, Nawinna, Galle, Kadawatha, Kurunagala.
The control location has been identified as Boralasgamuwa. We assume for this experiment
Boralasgamuwa and other sites have similar demographies.
Anticipated Issues
The data for the experiment will be collected before and during the research in the next few
weeks. We do not anticipate encounter any problems as for this experiment; seasonality does not
impact. The only limitations that may arise would be some employees may not have the physical
infrastructure available to participate in the test, like not having a smartphone or having a limited
data package.
The above issues can affect the outcome of sales, but we believe it will be a very insignificant
number that could hardly have an impact.
Experiment 2.0
Another version of this experiment is to go public with a TV and social media campaign. Hence,
this can be a full factorial design with independent variables like sales, app downloads, and a
theme “smart options for smart people.”
Media and web analytics will tell us buying insights of targeted audiences
This version of the experiment is expensive as we are launching an above the line TV campaign as
well as a relatively low-cost social media campaign. As this campaign will reach out to a much
more significant potential, customer base, sales results can be tremendous. The TV and social
media campaigns are going to cost us $52,000. We expect the number of users to grow from
6,000 to 600,000. The average order value is around $20 per customer. The net margin per order
is about 5%.