Analysis of Qualitative Data using Affinity Diagram and Pareto Principle(With working example)
Why Affinity Diagram? Is it relevant to even business manager to make business decisions, identify pain points of customers etc.? If yes, how?
Let’s start with what affinity diagrams help us to do :
1. They help make sense of variety of data (user opinion/feedback, ethnographic research, facts, issues, etc.)
2. Group information such that insights and business decisions can be taken basis them
It is used excessively by User Experience designers to take understand multiple things like user’s pain points, their needs, preferences and helps them identify opportunities also. It is a physical exercise which is performed using sticky notes, markers and board.
Below is how a typical scene looks like in during the exercise :
In this article we are focusing on Affinity Diagram from the perspective of a business manager and how the manager can get insights from qualitative data.
Let’s take up a small but complete hypothetical case to understand the importance and utility of Affinity diagram technique in understanding core problem areas from qualitative data.
Case : You are owner of Hind Business Hotel and have been running this hotel successfully since last 20 years. Usually the occupancy rate of your hotel remains at around 90% but lately this has dropped to 75% over the last one year. Your guests are mainly repeat guests who travel and stay your hotel at regular intervals. You want to understand what are the causes for this drop. As the owner of the hotel you want to find the core problems which the guests are facing during their stay.
Action Plan : You recalled that you read about the 80/20 rule (Pareto principle) and how affinity diagram can help in categorising the problems post which Pareto principle can be applied.
You ask the manager of your hotel to prepare the list of all the guest reviews over the last 60 days. The manager helps you to filter out all the negative reviews and shares this is the link of all negative the customer reviews your hotel got over the last 60 days. (P.S. Data has extracted from data.world’s hotel review database.)
Now, categorise the reviews under different headers/groups. These groups are basically collection of the broader scope of the review. Ideally not more than 8–10 groups should be created. Below is how the sheet should look like.
In sheet 2 (refer to above screenshot), we have categorised all the complaints under these 8 categories. Now, put 1 for the category if the review falls into it otherwise 0. Calculate the sum of every category, i.e. the count of 1’s in every category.
Next, create the a table as shown in below table. This would show category-wise count and cumulative values.
In the above Screen 3, create the table and generate the Pareto map. If you are using Microsoft Excel 2016, you can find Pareto chart in the histogram section.
Insights from the analysis
About 80% of the issues can be resolved by focusing only on 4 things. Rooms and its facilities have the highest contribution in this 80% figure. Even if we fix only Rooms and its facilities we can bring down the negative comments significantly.
Next Action : As the owner you now want to understand what went wrong in room and its facilities and if there is some common persistent problem in the rooms because of which so many people are commenting negatively about them. I might also want to analyse the positive reviews similarly to see whether the results I am getting are right or not.
Hope this analysis helped you gain insight into the qualitative data and understand the core problem.
If you found this article useful or it helped you in some way, please hit clap, and share. Incase of queries, feel free to ask!
To access the sample excel sheet with all calculations click here.
