visitors with high-quality product photos.
Posted: Tue Jan 21, 2025 9:07 am
Example of a Cohort Analysis Report in Google AnalyticsExample of a Cohort Analysis Report in Google Analytics
This type of analysis is not popular in RuNet, although in practice it can bring additional clarity to the comparison of the effectiveness of advertising sources. And also help to predict the success of new advertising that is launched for the first time.
Alexey Kulichevsky wrote about this type of analysis in detail in his OnMyStats blog . Frankly, we were even upset after the service closed last year. And now, Google Analytics finally has the ability to provide full-fledged reports for cohort analysis! First, let's figure out what the unfamiliar term "Cohort" means.
A cohort is a group of users who visited your site on the same day. For example, you placed a banner on a forum on your topic and on February 8, 100 people visited yours for the first time. On February 9, 100 new users, and on February 10, another 100. These are the 3 cohorts that we can monitor. Users visit your site but buy almost nothing, and you change the banner. After the change, some time passes and sales increase. However, the question arises: what ultimately led to sales: the new banner design, or did your customers simply “mature”? Cohort analysis will give us the answer to this question.
Example of transactions in cohorts
Example of transactions in cohorts
The screenshot shows an example of what cohort analysis looks like for the tested source. Having collected statistics and analyzed 4 cohorts, it is clearly visible that the banner change did not affect the purchases made. The second cohort followed the first banner, the fourth – the updated one. In each case, the transaction was made on the 7-8th day.
Let's say, having tested different banner formats over a longer period of time, we have confirmed that the average transaction cycle is up to one week. Now we have new useful knowledge! By placing the same banner on other forums, we know when we will receive an answer about the effectiveness of the site. This means that there is no point in buying a whole month of placement for the test. One week is enough, and in 1 month with approximately the same budget you can test 4 different sources.
Cohort analysis is also very useful for comparing advertising campaigns with each other. Starting from individual campaigns in contextual advertising, ending with retargeting and teaser systems. In Google Analytics, for this type of report, as in all brazil whatsapp number database others, there is support for segments. This means that you can get a lot of information not only on channels and sources, but also on technologies, demographics, a group of specific products, etc.
Segments in Google AnalyticsSegments in Google Analytics
Another example is given by PracticalEcommerce on their blog .
Let's imagine that your online store is launching an email signup campaign. You're creating three campaigns: a Facebook contest, a pop-up on your website, and a static link on a partner's website that invites people to sign up for your newsletter.
Over time, the marketing team might consider email open rates from these three sources:
This type of analysis is not popular in RuNet, although in practice it can bring additional clarity to the comparison of the effectiveness of advertising sources. And also help to predict the success of new advertising that is launched for the first time.
Alexey Kulichevsky wrote about this type of analysis in detail in his OnMyStats blog . Frankly, we were even upset after the service closed last year. And now, Google Analytics finally has the ability to provide full-fledged reports for cohort analysis! First, let's figure out what the unfamiliar term "Cohort" means.
A cohort is a group of users who visited your site on the same day. For example, you placed a banner on a forum on your topic and on February 8, 100 people visited yours for the first time. On February 9, 100 new users, and on February 10, another 100. These are the 3 cohorts that we can monitor. Users visit your site but buy almost nothing, and you change the banner. After the change, some time passes and sales increase. However, the question arises: what ultimately led to sales: the new banner design, or did your customers simply “mature”? Cohort analysis will give us the answer to this question.
Example of transactions in cohorts
Example of transactions in cohorts
The screenshot shows an example of what cohort analysis looks like for the tested source. Having collected statistics and analyzed 4 cohorts, it is clearly visible that the banner change did not affect the purchases made. The second cohort followed the first banner, the fourth – the updated one. In each case, the transaction was made on the 7-8th day.
Let's say, having tested different banner formats over a longer period of time, we have confirmed that the average transaction cycle is up to one week. Now we have new useful knowledge! By placing the same banner on other forums, we know when we will receive an answer about the effectiveness of the site. This means that there is no point in buying a whole month of placement for the test. One week is enough, and in 1 month with approximately the same budget you can test 4 different sources.
Cohort analysis is also very useful for comparing advertising campaigns with each other. Starting from individual campaigns in contextual advertising, ending with retargeting and teaser systems. In Google Analytics, for this type of report, as in all brazil whatsapp number database others, there is support for segments. This means that you can get a lot of information not only on channels and sources, but also on technologies, demographics, a group of specific products, etc.
Segments in Google AnalyticsSegments in Google Analytics
Another example is given by PracticalEcommerce on their blog .
Let's imagine that your online store is launching an email signup campaign. You're creating three campaigns: a Facebook contest, a pop-up on your website, and a static link on a partner's website that invites people to sign up for your newsletter.
Over time, the marketing team might consider email open rates from these three sources: