Cohort Analysis (Google Analytics) Use Cases for ECommerce & Customer LTV

Cohort Analysis is an analytical model which is used to study the behavior of a group of users experiencing the same common event over a time period. For example, five different groups of users are exposed to the same event and then their behavior is studied. These different groups of users could be young professionals, smokers, professors, movie stars, and runners. These groups will get exposed to Yoga for a period of 9 weeks and then the effect of Yoga is measured by measuring their blood pressure.

The above example was from a general medical field. In the digital marketing world, we often measure a performance metric for different audience type based on a date range. In this article, we will talk about the Cohort Analysis in Google Analytics.

Cohort Analysis Report (Google Analytics)

If you login to your GA account, you will see the option of Cohort Analysis in the left panel under the Audience section. If you are measuring all the conversions and goals correctly in Google Analytics, you will see a working Cohort analysis in Google Analytics. It has four fundamental dimensions as highlighted in the screenshot above

Cohort Type

The dimension that’s the basis of the cohorts. You can only select one dimension at a time. Currently, the acquisition date is available in GA as the Cohort type. Based on the basis of acquisition date or the date on which the user was first acquired, the Cohort analysis will be done.

Cohort Size

The time frame that determines the size of each cohort.

Metric

The Metric that you want to use to measure the performance of your marketing effort through Google Analytics. Date Range is self-explanatory. You can also see that I have highlighted the audience segment at the top. You can choose the segment to see the performance of different audience types such as Organic, Paid, Social, Direct traffic, etc.

If you do not have your Cohort Analysis fundamentals cleared, I would recommend watching this video:

Before we begin Cohort analysis, you should note that there are multiple factors that affect the metric performance and users behavior. You should not totally rely on the data and make data-based decisions on that. Rather, you should also consider other factors and try to relate to your Cohort analysis such as:

• Do you run weekly promotions?
• You email schedule
• Seasonality of products
• Long weekends
• Weather
• Offers and Discounts
• Competitor’s initiatives etc.

The Cohort analysis and use cases that you will follow are just to give a general idea.

Cohort Analysis Use 1 – All Users

Here are the two screenshots that I am going to use for the Cohort Analysis purpose. These screenshots are from an Ecommerce client that I handle

Cohort Analysis

• Users initiating sessions on Tuesday (You can find the date from date) have the longest number of following transactions (shopping spree). So, offers should be released on Tuesdays, if possible, to have the highest number of possible followed transactions.
• Users arriving on Thursday like to continue to shop on Friday as well. So, maybe, the offer should last until Friday.
• Users arriving on Sunday do not like to continue shopping on Mondays
• Overall user retention is pretty good for a week
• Users making a transaction on Tuesday have the highest possibility to buy on Wednesday as well. As a marketer, you can offer special next day coupon code to users making a purchase on Tuesday.
• Keep more marketing budget for the first half of the week
• Reduce your overall marketing effort from Sunday to Monday

Cohort Analysis Use Case 2 – ECommerce Checkout Users

Here are the two screenshots that I am going to use for the Cohort Analysis purpose. In this case, Ecommerce Checkout Users are users who must visit the checkout page on the website along with the other pages on the site.

Cohort Analysis

• Users with Checkout page visit are more likely to perform another transaction the next day as well. However, their session continues until Day 4 to Day 7 heavily which can be attributed to product shipping and arrival. Maybe these sessions are more to look at product shipping status.
• On average, the ratio of a transaction to a session is 0.18 overall. However, this ratio is higher on certain days and lower on certain days. I would recommend setting up a custom metric in Google Analytics and setup that metric as a goal and measure the metric value over a longer period of time to see if there is a weekly trend that follows

In this case, I am just analyzing the traffic by default acquisition channel that comes with Google Analytics. You can create your won segments in Google Analytics or even create custom channel grouping to analyze the data better. As far as the Google Analytics Audience Segments are concerned, I rely heavily on my RLSA in Google Ads at different funnel levels as I set my bids and budget as per their position in the funnel.

Cohort Analysis Use Case 3 – Organic Traffic

Here are the two screenshots that I am going to use for the Cohort Analysis purpose of the Organic traffic.

Cohort Analysis

• Organic traffic is the highest engaging traffic with repeated high engaging sessions until day 4. It can also be said that the highest user retention is until day 4. However, most of the transactions take place within 2 days for organic traffic.
• There is a weekly trend of higher transactions to sessions ratio on Tuesday which mimics the trend of overall Users

Cohort Analysis Use Case 4 – Paid Traffic

Here are the two screenshots that I am going to use for the Cohort Analysis purpose of the paid traffic.

• There is a weekly trend of higher user retention on Every Saturday. This also converts in higher sessions on these days but looking at the transactions data, these users do not necessarily convert.
• Most of the transactions go until the second day of acquisition.

Cohort Analysis Use Case 4 – Social Traffic

Here are the two screenshots that I am going to use for the Cohort Analysis purpose of Social Traffic.

• Social traffic likes to perform shopping on the same day. The chances of returning and continuing shopping is very minimal.
• Social Media can be used for flashy offers especially during valentine day or holiday season such as Boxing day, Cyber Monday, Christmas, etc.
• The user retention is also pretty low

Cohort Analysis Customer LTV

You can do Cohort analysis on different audiences and accordingly tweak your marketing plan. Here is a screenshot of customers LTV based on different acquisition channel for one of the clients that I handle

Some of the analytical things are pretty straight forward:

• LTV of the direct users is the highest followed by referral traffic, paid search, and organic search.
• Least customer LTV comes from the display networks such as social traffic and Google display network.

However, based on the above LTV result, you should not decide the marketing initiatives. In this case, there are other factors that need to be considered such as

• Have we analyzed branded paid search vs unbranded paid search LTV?
• The display networks bring cheaper ToFu traffic. If we are going to reduce the Budget then from where ToFu traffic comes.
• Have we analyzed the LTV of display remarketing audience?
• From where such vast direct traffic is coming and how can we more capitalize on that?
• Since social’s transactions cohort is very high on day 1, can we try to capitalize on that? Is there any way to introduce daily promotions and offers?

For a better analysis using the acquisition channels, you can create custom channels in Google Analytics