The Banana Stall ‹ Market Research ‹ Portfolio ‹ Meridian ‹ meridian.net.au
As part of my Certificate IV in Marketing at the North Coast Institute of TAFE, I needed to do a major assessment task for the Market Research module.
As a group, we chose to study the characteristics of customers that shop at The Banana Stall – a small fruit and vegetable store in Port Macquarie, on the Mid-North Coast of New South Wales.
The research project set out to establish the following:
Typical customer profile
What are the customers' demographics?
Purchase motivations
What do customers consider important when shopping for fruits and vegetables?
Strengths and weaknesses
How well does the store meet the needs as determined above.
New products and services
What new products and services would customers like to see introduced?
With the objectives defined, we designed a questionnaire to collect the information.
The questionnaire utilised mainly closed ended questions where responses were selected from a list of possibilities. We did this because the results were easier to code and analyse.
The 'Considerations' and 'Strengths and Weaknesses' sections used a scale on which respondents could rate their feelings towards the questions. Again, this was easier to code.
Once we had collected the desired number of questionnaires, my IT skills kicked in. I replicated the questionnaire as a web form and used PHP to submit the results to a MySQL database. Using a database to store the results made it easy to develop queries to analyse the information.
Getting meaningful results is one thing; communicating those results is another. This is where my graphics design and information design knowledge played a big part.
Data tables – not bar charts
There are no bar graphs or pie charts as part of this report, and that was quite deliberate. Such charts are usually visual overkill for such small datasets. For example, each question could be summarized with about half a dozen numbers. Tables are much better at conveying this information quickly.
Colour was used on one table to indicate those values that were above and below the average by a significant amount.
Use colour to hilight and maps to personalise the information
More interesting are the maps showing where customers came from within Port Macquarie. This information could have been presented as a summary of the number of respondents living within each section. However, by hilighting the streets where respondents lived on a map of the town, it instantly shows the clustering during week days, and the dispersion during the weekend.
Increase data density
Matrix for quality
Matrix for organic
The responses from the two rating-scale questions were used to form a matrix of small multiples. The x-axis on the matrix represents performance (left to right: very good, good, poor, very poor, unsure). The y-axis on the matrix represents importance (top to bottom: very important, somewhat important, unimportant, unsure).
Shading represents the percentage of responses falling into each interval. A matrix was produced for each of the criteria over all respondents, and also broken into demographics.
This display clearly shows the tight grouping of responses for quality (top-left or very important/very good), and the bimodal distribution of responses for organics (top-left or very important/very good, and bottom-right or unsure/unimportant).
Page 18 of the report, therefore, summarizes some 2,000 data points on one page &ndash 5 performance × 4 importance × 10 criteria × 10 demographic segments.
Download a ZIP archive of all the images [661 KB]
Below are thumbnails of each page of the document. I have not included an original Microsoft Word or Adobe PDF version as I do not want copies circulating around the Internet. You can, however, download a ZIP archive of all the images.
Click on each image to get a larger view of the page.
Title page, Table of Contents, List of Tables, List of Maps, List of Figures
Pages 1 through 22
The questionnaire