Experimental Design

Marketing Experimental Design
Britannia 50/50
Introduction
This report aims to design a marketing experiment for 50/50 product which is a registered brand of
Britannia Industries Limited. This product comes under the category of salted-cookies snacks.
The campaign planned here would be a TV commercial focused on taste and experience associated
with having 50/50 snacked cookies during evening tea. Today’s generation spends so much time
stressing over so many multi-lateral issues, throughout the day, that evening tea break time actually
relieves that tension. The campaign wants to emphasize on the fact that, this evening tea break can
be more enriching with Britannia’s 50/50.
The experiment will be a before –after design experiment. Not going for full factorial, as it will be
difficult to quantify impact if several parameters are simultaneously changed.
Experiment Design
In this experiment, the independent variable is increased advertising in the form of Television
Commercials. The dependent variable observed here is sales.
The experiment will be run in Noida, Uttar Pradesh. The control market observed will be Gurgaon,
Haryana. The test and control markets are so chosen that they resemble closely in terms of
population demographics and other city attributes. Columbus and Indianapolis are similarly sized
cities with a population of around 700,000.
Since the entire cities will be exposed to the experiment, the sample size would be near to 1000
people to be statistically significant.
Since it is a before-after design experiment, sales will be recorded for both control and test markets
for a 2-month average from April-May. The experiment will be run for the next two months from
June to July. The change in sales during the experiment will be then used to calculate the lift in sales.
The experiment certainly adheres to the first three rules of causality. However, the experiment has
no control over the fourth rule, which deals with the presence of external factor. Prima facie, there
appears to be no external factor impacting the experiment, however there might be competitive
response that shall be accounted for.
Possible Issues
Both the before and during experiment results are collected during the summer months from April
to July. If the field implementation is executed anywhere between August to September, the
autumn season (as it rains heavily during these two months in India, and consumption of snacks
and tea increases) might impact the sales due to the seasonality involved.
Any other external factor during implementation such as a new competitor entering the market or
decrease in prices from competitors could adversely affect the field results.
The above-mentioned issues might have various impacts on the sales, like, seasonality might
contribute to increased sales, and thus making the calculation of ROI on TV ads complicated, or
there might be an adverse effect on sales due to external factors; however, the experiment will still
demonstrate the impact of TV campaign on sales. For seasonality, past data can be used to see the
pre-existing lift between different seasons sales. The seasonality and external factors
notwithstanding, the experiment will provide a fair guideline on whether to go ahead with the
nationwide marketing campaign. The lift in sales would also provide a concrete idea on the
appropriate amount of spending on the campaign.
Digital Experiment
Another version of the experiment can be through paid web advertising. Web advertising provides
much better control on the target audience, test markets and analytics. Hence, it can be a full
factorial design with the independent variables as both price and advertising theme.
This will also yield more informative results as
 It demonstrates the impact of more than one independent variable. The web analytics gives far
more insights on the buying preferences of the audience
This version of experiment is also generally cheaper and quicker to implement than the offline
campaigns such as TV or billboard advertising.
The only concern with the web experiment is that it does not include the chunk of audience that is
not so active on web, and it is imperative that we include as many people in web-based experiment
as we can. In the context of this particular product, that chunk of inactive users could be huge.

Experimental Design

Uploaded by

Copyright:

Available Formats

Experimental Design

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Experimental Design

Uploaded by

Copyright:

Available Formats

Marketing Experimental Design

This will also yield more informative results as

You might also like