Cohort analysis allows us to understand consumer behaviour, which is key to improving the customer experience. This is one of the indicators that is most disputed by companies as a determinant of competitiveness because it drives customer loyalty and this can contribute to the success of the company, according to Zendesk's study on customer experience trends in 2020
Customer experience is also the first focus of organisational strategies for successfully managing companies, according to KPMG's study 2020 Outlook Innovation, Trust and Growth.
Learn how to carry out a cohort analysis and interpret its findings. Understand your customers' behaviour and use the data in your company's strategic marketing plan to improve your customers' experience and your company's competitiveness.
What is a cohort analysis?
Before defining what a cohort analysis is, let's start by knowing that a cohort is the classification of a group of individuals with common demographic characteristics during a given period of time. These characteristics are generally associated with generational data, e.g. age and historical events that marked an era.
At the end of the 19th century, the statisticians Karl Becker (1874) and Wilhelm Lexis (1875) laid the foundations of this study, which later became internationally famous, with the help of the demographer Pascal Whelpton (1949), who analysed the increase in the birth rate in the United States after the Second World War.
The importance of the cohort analysis
The British physicist and mathematician William Thomson Kelvin said: "What is not defined cannot be measured. What is not measured cannot be improved. What cannot be improved will always deteriorate".
The implementation and correct interpretation of cohort findings enables us to:
- understand consumer behaviour based on specific criteria;
- identify consumer trends;
- provide statistics on sociology and demography;
- analyse behavioural changes in defined groups;
- know how a cohort engages with the brand;
- compare behaviours with other groups;
- measure the level of customer acceptance or the impact of changes to the brand;
- study phenomena in areas such as medicine, politics and economics;
- create forecasts or future scenarios;
- validate or support the results of A/B split tests;
- make better decisions and generate strategies to solve problems;
- reformulate the customer experience (CX).
If you're on a mission to improve the customer experience in your business, learn how Zendesk is transforming the customer experience in 2020.
Types of cohort analysis
According to the methodology
There are two types of cohort analysis according to the methodology:
- Prospective: The analysis is carried out with a look to the future, in other words, identifying a cohort with its current conditions and monitoring for a long time to analysis the changes in behaviour.
- Retrospective: This studies a current behaviour, based on data from the past.
According to the scope
There are two types of analysis depending on the scope:
- Intracohort: This allows for a deep analysis and knowledge of the cohorts and their behaviour regarding the events and variables defined for the study.
- Intercohort: This is a comparative with at least one other group of people, with the aim of identifying differences.
How to carry out a cohort analysis in 5 steps
To carry out any of the types of cohort analysis, it is important to complete the following 5 steps:
1. Establish the objective of the study
Defining the topic of the study gives the research a focus. In addition, it helps to define variables and find data without going beyond what is necessary.
2. Define the events that will frame the study
These events, or variables, will be the ones that answer the objective of the cohort analysis.
3. Select the relevant cohorts
This stage is about defining the type and number of cohorts that will participate in the analysis. It's extremely important to have clear segmentation.
4. Carry out the study
In order to carry out the study, it is necessary to select the type of cohort analysis, i.e. whether it will be prospective or retrospective and whether it will be intracohort- or intercohort-type analysis. Once these definitions are in place, we can proceed with the study.
5. Interpreting the results
This refers to the stage at which the cohort analysis report is analysed according to the behavioural variations shown in the findings.
How to interpret cohort findings
The interpretation of the cohort findings involves evaluating and weighing the 3 effects or factors that influence the identified behavioural changes.
It explains the differences between the cohorts based on the existence of various social and environmental influences.
The changes are associated with age, the life span.
Behavioural changes are associated with changes in the environment, regardless of the generation or socio-demographic factors.
It is important to separate the three effects, which can occur simultaneously for each event.
An example of the application of a cohort analysis
In an e-commerce shop, the findings of the cohort analysis provide insights into consumer behaviour. It works quite well, especially in seasonal campaigns such as Black Friday.
In that case, it is possible to understand specific behaviours during that period across the different stages of the customer journey. Some of these data may include:
- the cart abandonment rate;
- the advertisement through which they were redirected to the website;
- the average ticket, among other things.
But it doesn't just apply to browsing an e-commerce shop – depending on the objective of the study, it can be a highly relevant input when it comes to measuring the impact of strategic marketing actions.
Implement cohort analysis in your company and understand consumer behaviour so that you can make decisions that improve your customers' experience and your company's competitiveness.
Zendesk is a CRM company with products designed to improve customer relationships. You can rely on Zendesk for Service. to offer your customers an end-to-end support experience.