Skip to content
Home » The Ultimate Guide to Sampling for Lean Market Research

The Ultimate Guide to Sampling for Lean Market Research

In Brief

Before conducting any survey in the framework of lean market research, it’s key to perform a relevant sampling. This post is about The Ultimate Guide to Sampling for Lean Market Research. This aims at clarifying practically why and how to conduct a consistent sampling for an insightful market intelligence.

Sampling stands for how many respondents belong to a given research. Sampling also focuses on how to split those samples into relevant groups.

The big picture about sampling

A usual question that exist in market research is: For conducting an accurate and representative quantitative research, how many samples should I take into account.

In other words, how many people should I ask when conducted a survey so that I can be confident enough on the relevance of the collected data ?

This is exactly the question that sampling is targeting.

In this framework, let’s start first by defining sampling:

Sampling stands for statistical and mathematical rules that can be used for assessing how much data should be gathered in order to be confident enough about the deduced results.

These rules can be sometimes counter intuitive hence it’s important to know them without using gut feeling and intuition.

A typical sampling is to decide to consider 2000 random samples to be representative of a population of 50M people.

Why considering sampling rules

As lean market researcher, there are numerous gains that you can win out of mastering sampling rules like for instance:

  • Reducing the cost of quantitative research
  • Quantifying the accuracy and reliability of the gathered statistics
  • Taking decision whether or not more quantitative data will be beneficial to the study
  • Providing more credibility to the conducted study knowing that credibility is the most valuable asset to a market research

Let’s take a look then at these precious sampling rules:

  • There are diminishing returns in having a large sample size: going from 500 to 1000 improves the accuracy from 4.5 to 3.2
  • Patterns and causation should be investigated among the collected data
  • 2000 random sample size in a population of 300M gives an accuracy of 2.2% of the true result with a confidence

For instance, this is a representative graph exhibiting how the error decreases considerably through the 100 first samples but than starting from 500 to 1000 samples the gain in adding new samples, i.e. decreasing the overall error, is diminishing exponentially.

Pros tips – The Ultimate Guide to Sampling for Lean Market Research

  • The questionnaire and research accuracy should be defined clearly from Step1 (Define) of the analysis
  • Indeed accuracy has a cost and there is no need to add cost to the survey if it can be avoided
  • On the other hand it’s vital to provide a good accuracy for a higher credibility to the conducted study
  • Consumer sampling focusses on covering a specific group of consumers or future ones based on a specific set of features like age, gender, behavior, professional background, nationality, culture etc
  • Whereas Business to business sampling aims at sampling buyers from companies of different sizes and different specialties
  • Statistics are key for extracting the essence out of the gathered quantitative research
  • Great statistical tools like Excel, Tableau, Google sheets, R, Python etc provide great features for conducting highly relevant statistical studies even for non statisticians
Print Friendly, PDF & Email

Leave a Reply

Your email address will not be published. Required fields are marked *