This is the second part of "Making Sense of Data", the first part can be found here
The final phase of the project, where one has to present findings of the primary research, is more an exercise on “data-story-telling”. It is not limited to looking at outputs of individual questions in your questionnaire. It is rather an exercise on how well you can draw linkages from the results of different questions that you have asked in the questionnaire, validate your hypotheses using these linkages and finally, seamlessly communicate a story.
Let us take an example of a term project done by a group of students a year back. Two new brands of shampoo were launched in the market, targeting SEC A/B. The group wanted to understand the brand health (in terms of Awareness, Trial, Repeat Purchase, Regular Usage, Disposition, Product Perception and Brand Imagery) of the two brands among their target consumers. There were some hypotheses that they had developed, regarding the performance and consumer perception of the two brands.
The questionnaire, in this case, would have separate questions on Brand Awareness, Trial, etc and finally a battery of statements on product perception and brand imagery. Making question-by-question charts in your PPT would mean a dull reporting of the good-to-know facts. However, if you just try do draw a simple linkage between the parameters – Awareness, Trial, Repeat Purchase and Regular Usage of the two brands – you can obtain two “Brand Funnels” of the following nature**:
From the above “Brand Funnels” (which are merely summaries of data from 4 different questions and conversion ratios from one step to the other), you can draw several conclusions that will help you in your data-story-telling. Some of the conclusions can be:
1. Brand A enjoys higher awareness levels than Brand B among target consumers. However, conversion from awareness to trial has been significantly lower for Brand A. This could be for several reasons like poor availability of the brand, higher price point (or higher price perception), communication failing to give enough reasons to buy, etc
2. Brand B has higher trial levels, but conversion to repeat purchase is happening at a much slower rate than Brand A. This could be a result of consumers not being satisfied with the quality of the product after trying Brand B once.
Looking at responses to other questions in the questionnaire and linking them with the above parameters can further build stories from the above conclusions. For example, if you look at the Product Perception and Brand Imagery scores among “Aware Non-triers” vs “Triers” vs “Trier Non-repeated” vs “Repeaters”, the gaps may give you an idea of why consumers are not converting from one level to the next one for Brand A and Brand B.
Another common kind of hypotheses that students form is of the nature of “Brand X is more appealing to younger people” or “Brand X is more male-centric”. In such cases too, instead of asking direct questions and looking at the data of their responses, simple linkages can help. You must have collected data on “age group” or “gender” of the respondent in the questionnaire. So, divide the sample size of the study into sub-groups (“Young” / “Middle-aged” or “Male” / “Female”) and compare the results of all relevant parameters (Awareness, Trial, Brand Imagery, etc) by these sub-groups. Are you getting significant differences? You may validate your hypotheses in this manner. However, a point to note here – While designing the study, you should have kept enough sample sizes (at least 30!) for each sub-group to analyze the data in this manner.
We are still left with one very important aspect of data-story-telling in the final phase of the project. When do we use what type of chart to communicate the story? The “Brand Funnels” drawn above were kind of an exception! In most other cases you may have to decide whether to use a Bar Chart or a Pie Chart or a Line Graph… and most importantly, in many cases, whether to make a Chart / Graph at all or not!
[To be continued...]
- By Prof. PD Purkayastha
** - References can be drawn from Philip Kotler’s textbook (15th Edition), Chapter 6: Analyzing Consumer Markets
(The 3rd Part of this article on “Making Sense of Data” will be uploaded)
The final phase of the project, where one has to present findings of the primary research, is more an exercise on “data-story-telling”. It is not limited to looking at outputs of individual questions in your questionnaire. It is rather an exercise on how well you can draw linkages from the results of different questions that you have asked in the questionnaire, validate your hypotheses using these linkages and finally, seamlessly communicate a story.
Let us take an example of a term project done by a group of students a year back. Two new brands of shampoo were launched in the market, targeting SEC A/B. The group wanted to understand the brand health (in terms of Awareness, Trial, Repeat Purchase, Regular Usage, Disposition, Product Perception and Brand Imagery) of the two brands among their target consumers. There were some hypotheses that they had developed, regarding the performance and consumer perception of the two brands.
The questionnaire, in this case, would have separate questions on Brand Awareness, Trial, etc and finally a battery of statements on product perception and brand imagery. Making question-by-question charts in your PPT would mean a dull reporting of the good-to-know facts. However, if you just try do draw a simple linkage between the parameters – Awareness, Trial, Repeat Purchase and Regular Usage of the two brands – you can obtain two “Brand Funnels” of the following nature**:
1. Brand A enjoys higher awareness levels than Brand B among target consumers. However, conversion from awareness to trial has been significantly lower for Brand A. This could be for several reasons like poor availability of the brand, higher price point (or higher price perception), communication failing to give enough reasons to buy, etc
2. Brand B has higher trial levels, but conversion to repeat purchase is happening at a much slower rate than Brand A. This could be a result of consumers not being satisfied with the quality of the product after trying Brand B once.
Looking at responses to other questions in the questionnaire and linking them with the above parameters can further build stories from the above conclusions. For example, if you look at the Product Perception and Brand Imagery scores among “Aware Non-triers” vs “Triers” vs “Trier Non-repeated” vs “Repeaters”, the gaps may give you an idea of why consumers are not converting from one level to the next one for Brand A and Brand B.
Another common kind of hypotheses that students form is of the nature of “Brand X is more appealing to younger people” or “Brand X is more male-centric”. In such cases too, instead of asking direct questions and looking at the data of their responses, simple linkages can help. You must have collected data on “age group” or “gender” of the respondent in the questionnaire. So, divide the sample size of the study into sub-groups (“Young” / “Middle-aged” or “Male” / “Female”) and compare the results of all relevant parameters (Awareness, Trial, Brand Imagery, etc) by these sub-groups. Are you getting significant differences? You may validate your hypotheses in this manner. However, a point to note here – While designing the study, you should have kept enough sample sizes (at least 30!) for each sub-group to analyze the data in this manner.
We are still left with one very important aspect of data-story-telling in the final phase of the project. When do we use what type of chart to communicate the story? The “Brand Funnels” drawn above were kind of an exception! In most other cases you may have to decide whether to use a Bar Chart or a Pie Chart or a Line Graph… and most importantly, in many cases, whether to make a Chart / Graph at all or not!
[To be continued...]
- By Prof. PD Purkayastha
** - References can be drawn from Philip Kotler’s textbook (15th Edition), Chapter 6: Analyzing Consumer Markets
(The 3rd Part of this article on “Making Sense of Data” will be uploaded)
No comments:
Post a Comment