Things to Consider Before Diving into Data Analysis
Before you start opening programs like Excel, Power BI, Tableau, or Looker Studio to analyze data, consider these 5 important points:
Objective – The Purpose of Analysis:
The first thing you should ask yourself is, why do you need to analyze this data? What is the objective of your analysis? For instance, do you want to identify market opportunities, understand customer behavior, analyze the effectiveness of a marketing campaign, or assess customer satisfaction? Clearly defining the objective will help focus your analysis and provide a clear direction.
Methodology / Sampling Method – How the Data is Collected:
The method of data collection is crucial. Are you using Focus Groups, online surveys, in-depth interviews, or data from POS systems? Additionally, consider how you select your sample groups to ensure the data is reliable and unbiased.
Respondent / Source – The Origin of the Data:
Where does your data come from, or who provides the information? Data can often be biased if you ask the wrong people or if the respondents have a hidden agenda. For example, if you ask people if they like exercising, most would say yes because they want to appear health-conscious. Data from POS scanners might be skewed if combined with data from stores without POS systems, like small neighborhood shops. Similarly, data from Facebook might not cover information about the younger generation who use TikTok more. Loyalty program data from supermarkets might show the same membership number for an entire family, leading to incorrect profiling.
Unit of Measurement – How the Data is Measured:
Different units of measurement can lead to different interpretations. Therefore, it’s crucial to specify the unit of measurement clearly. For example, are sales measured in terms of revenue (in currency), the number of items sold, or the number of customers?
Underlying Assumption – Assumptions in Data Collection:
Underlying assumptions are the hypotheses you make before or during data collection, which can influence how you interpret the data and the results. Clearly identifying these assumptions helps in evaluating the accuracy and reliability of the data. For example, how many people are in your sample group, or what comparisons are you making?
Examples:
Example 1: A toothpaste claims to remove plaque 3 times more effectively. You need to check what it is compared against.
Example 2: The top-selling product – you need to know what products it’s compared with and during which period.
Example 3: On the surface, it may seem that Shopee has a larger market share than Lazada, but the unit of measurement might be the number of millions of visits to these websites in Thailand.