The importance of purpose in research: Always start with a WHY!

Research is more difficult that it seems. It needs strong methodology and a solid mathematical knowledge. In my personal opinion, the most important aspect (and where most of marketers fail) is the main question that triggers all the research, the main WHY. Within the following article, we will try to focus in this important question in order show some ways on how to accomplish a healthy market resarch study.

By fail I don’t say that marketers don’t take this question into account, I just state that they don’t put enough time and effort on it. Within the following article I will go through some basics of research and some categories that will help the researcher to classify their problems/questions and which will trigger the different ways and methods that are commonly use to get to the famous ‘insights’. 

Always start with a WHY.

To begin every single research I always try to figure out what I am expecting of it. In other words, how would I use the results. For example, it is very different a research to solve some pricing issues that product development issues. It important to understand what we want to do before jumping into the methods of doing research. This is what is called ‘diagnostic’. There are lots of types of definitions, but I always go for simplicity so this is one once I was given within some research course

Type I: Exploratory Research

This type of research is important in order to solve managerial questions that are not very well defined. Usually are given by hunches or vague insights. They often provide the context and data for further investigation. Some question we often solve are:
  • Why are sales increasing/declining/flat? 
  • What people think of our product?
  • Do users understand our value proposal?
Some examples of Exploratory Research are Focus Groups, Internet Communities, among others.

Type II: Descriptive Research

Secondly, the descriptive type of research is recommended to understand defined problems in which we need to expand our knowledge. 
For example we may need to understand: 

  • Who are our consumers?
  • How to aggregate them?
  • How much they like our products?
  • How much better are we than a certain product of the competitors?
  • What is our Share of wallet?
  • What is our NPS?
This can be done by active data collection or passive observing behavior. The first group is based on surveys (for example with Qualtrix or Surveymonkey) and the second mainly scanner data (for example of Nielsen or Ipsos), mobile data or media planning such as radio, tv or social media.

Type III: Causal Research

Finally, the Causal Research is based primary in testing the problems in order to get to the answer of the WHY.

For example, we may want to address some of the following problems:
  • How would I increase my website conversion rate if change the order of two elements?
  • What if I offer some discount based on the location of the user? What about the segment of the user?
  • Which promotion is more efficient?
  • Are our price discounts increasing the consumers’ average ticket?  
  • Which is the best way to monetize my app?
Some important fact we need to understand Causal Research. In order to simplify we can name the following:
  • Correlation: relationship between two variables.
  • Causation: one variable producing an effect in another variable.
  • Correlation is different from Causation.
To sum up, we must follow some important rules for this type of research in order to assume causation of X and Y:
  1. Correlation: Evidence of association between X and Y
  2. Temporal antecedence: X must occur before Y
  3. No third factor driving both: Control of other possible factors
Most of this type is solved by testing. The most famous test is the A/B Test because its simplicity and actionable results. Lots of companies can help marketers to do it, in order to name some you can refer to Google, Optimezely or in case of mobile data set Leanimplum or Mixpanel.

As always, we really hope you enjoy our article. Thanks for taking the time to read!

1 comentario:

  1. After putting Scottish witch law into European context, he gives examples of the full range of victims executed there: innocent political pawns, beneficial cunning folk, the probably mentally ill Electrical Installation