Customer Cohorts

Tracking the LTV within cohorts of your marketing channels is vital for making informed ad spend decisions for your company.

Triple Whale avatar
Written by Triple Whale
Updated over a week ago

Intro

Cohorts analysis is vital in helping you track and understand your customer behavior patterns, make informed decisions, and identify valuable customer groups for retention.

Analyzing cohorts helps you understand how much revenue a new customer will generate over time, and determine the length of time it takes until customer acquisition cost (CAC) is returned.

What is Cohort Analysis?

What is a cohort?

A group of people with shared characteristics.

Cohort analysis is a method of analyzing your customers by first separating them into groups, and then tracking how each group (or “cohort”) engages with your business over time. The standard grouping in cohort analysis is by the timeframe a customer made his first purchase. Therefore, once a customer belongs to a cohort he/she is there for life.

The Cohort analysis also provides tools to refine the customer base which helps you better understand the LTV and retention of different subsets of your customers, and which will help with making important decisions such as:

  • How much should you spend to acquire new customers?

  • Which of your marketing efforts are contributing most to customer retention?

  • Which product categories are contributing least to long-term profits?

  • Which subset of customers are engaging with the store?

  • When do customers usually churn?

  • How much of your revenue comes from new vs long-time customers?

  • When is the best time to re-engage your customers?

How to use Cohort Analysis?

The report will track the behavior of customers who made their first purchase in the selected time period. These customers will be divided into cohorts based on the selected timeframe (ie: Week, Month, Year) of their first purchase. The report will calculate the selected metric per time frame subsequent to the cohort's first order.

By default, the report calculates accumulated LTV per customer over the preceding 365 days, with customers grouped by the month of their first order.

Analyzing a single Cohort:

In this example we have chosen to analyze the cumulative LTV of our customers in the last 12 months, with a month time frame:

  • Cohort: the first cell of every row identifies the cohort that you’re looking at. In this case, the cohort consists of customers who made their first order in May 2022.

  • Customers: the total number of unique customers in the cohort. These are new customers that made their first order in May 2022.

  • NCPA (New Customer Cost Per Acquisition): total ad spend divided by number of new customers. This is an estimate of the average cost of acquiring each customer in the cohort.

  • RPR: the percentage of unique customers in the cohort who made at least one additional order within the selected time period of the report

  • 1st order: the metric value for the 1st order, ie:

    • LTV: Total Sales of 1st order / number of unique customers in cohort - this is an estimate of the AOV of 1st order for the cohort

    • Total sales: The total sales of first order across all customers of the cohort

    • Number of customers: The total number of unique customers for 1st order

    • Retention rate: the % of returning customers - for 1st order this is always 100%

  • Month 0: will show the cumulative LTV for any additional orders done by the cohort in May

  • The remaining cells in the row show us how the estimated LTV of a customer in the cohort has increased month by month since first order, impacted by members of the cohort that have returned to your store to make additional orders.

As we track these customers after their first purchase we can get a sense of their repurchase value - we are interested in seeing are the values growing in time - in this case it is - we see that on average after 12 months from their first purchase, customers in this cohort are generating $69.73.

Color coding:

To keep the data visualization simple and to help immediately spot troublesome areas of churn or positive areas of growth , the cohort table uses color coding. We use various shades of blue to denote how the metric values fluctuate from the maximum to the minimum.

Once an interesting time period has been identified - you can further investigate the data by customizing the cohorts table.

Analyzing multiple Cohorts:

Another interesting analysis is how cohorts are compared to one another - comparing 2 cohorts can give insights on behaviors of your customers.

For example if we compare the May and June cohorts we immediately see that there are many more new customers that came in June than in May and that their repurchase rate is higher. Its interesting to analyze and understand whats impacting this difference:

  • How are we targeting customers in each cohort?

  • Were customers in the June cohort promoted with a specific campaign?

  • Are customers in the June cohort using a specific discount code?

It is also efficient to analyze the cohorts diagonally - especially for seasonal sales these values will often pop out. For example: customers that bought in the holiday months Nov-Dec we will often see higher numbers as customers will come back and repurchase.

How to customize the Cohort table?

You can customize your cohort table in order to dig deeper into you customer data and derive insights on problematic time periods for specific segments of customers or types of orders:

Time Period:

You can decide to focus on specific time periods for cohort analysis.

We will look at all customers who made their first purchase in the selected period.

(The minimum time period supported is “Last 7 days”)

Note on historical data:

The Cohort analysis report should show your store’s data from the beginning of time. In the case you see missing data please contact your CSM to reimport the store data.

Time Frame:

The time frame determines the buckets of time to group the cohorts.

They will automatically change according to the Time period selected

  • Month (default):
    ie: all customers who bought in Jan, Feb etc

  • Day: (available only when time period is a month or less)
    ie: all customers who bought on Monday April 18, 2022

  • Week:
    Weeks begin on Monday
    ie: all customers who bought in week beginning on April 18, 2022

  • Year
    ie: all customers who bought in 2022

Breakdown:

You can decide to break down the cohort data by different categories.

The standard breakdown of cohorts is by the time of the customer’s first order (per selected timeframe).

Coming soon: the ability to breakdown your cohorts by different characteristics of the first order, such as: products/SKUs, Location, Discount code, Channel etc

Metrics:

You can select which metric to view in the cohorts table:

  • Total sales:

Total revenue earned from customers within each cohort (after taxes, shipping and discounts)

Gross Sales + Tax collected + Shipping collected - Discounts

  • LTV:

Total sales of cohort per unique customers in the cohort:

Total Sales / number of unique customers in cohort

  • Number of customers:

The number of unique customers

  • Retention Rate

The percentage of unique customers who made at least one additional purchase during the selected time frame.

Please note:

We count the unique customers per timeframe (month, week etc) who have returned to make an additional purchase, but there may be overlap of customers between timeframes - ie: the same customer coming back to make a 2nd, 3rd, 4th ... purchase in following months/weeks etc.

Therefore when selecting the "Cumulative" view, the number of customers appearing in the last timeframe might be larger than the unique number of customers in the cohort.

Coming soon: Additional metrics …

NCPA Payback:

One of the most important insights to be gained from the Cohort analysis is an estimate of how long it takes to break even on new customers. Your NCPA Payback gives you an idea of how much time it will take you to earn back the marketing spend that you have used to bring in new customers. This metric is based on your CPA of new customers, and your Total sales (ie: Total revenue after taxes, shipping and discounts)

The NCPA payback helps you visualize where this breakeven point is for each of your cohorts.

Please note:

We currently do not calculate the NCPA payback given a specific customer segment - so when filtering by segment the NCPA column will show "--".

Cumulative (toggle)

The cumulative toggle will be ON by default when you arrive at your cohort report. Cumulative means we will sum up the total value created by the cohort over the time period selected.

For example, if you're looking at "Total Sales", the last month displayed for each cohort will represent the total amount of sales driven by that cohort

Alternatively, toggling Cumulative OFF will display the total additional sales generated by that cohort per month that it was generated in.

2nd Order Only (toggle)

The 2nd order only toggle will be OFF by default. Toggling this setting on will change the table to only reflect the orders, sales and LTV up to the customers second order.

Why use this setting?

  1. Analyze the most common 2nd purchase point for each cohort. Are the majority of your customers placing their second order 3 months after their first order? Are any cohorts placing their second order at a faster rate than the average?

  2. A more accurate reading of Repeat Purchase Rate

    1. With this setting toggled on the repeat purchase rate in the furthest column will match the repeat purchase rate that is stated in the RPR column.

Segments:

By default the analysis will be done on all customers who made their first purchase in the selected time period. Selecting a specific segment enables you to analyze the retention of a subset of your customer base.

New segments can be created in the SCDP.

Filters:

You can add filters based on the 1st order and derive insights on how specific products/locations/attribution etc are impacting LTV and repurchase.

Custom filters:

Click on the filter icon

Click on Add/View Custom Filter

For each category (Orders, Product, Customers, Attribution) you can add multiple filters.

Between multiple filters the “OR” operator is applied.

When done click on the “Apply” button and the custom filter will be applied to the dataset and the icon filter will show “1”.

Applied filters affect only the cohort page and will be reset at the end of the user session.

Saved filters:

Custom filters can be saved. In the custom filter window click on “Save” button and enter the name of the filter.

Once the filter is saved it will appear in the filter list when clicking on the filter icon.

You can select multiple saved filters and apply them - in this case the “AND” operator is applied between them.

Within the list of saved filters - each filter can be: edited, renamed or deleted.

Filters are saved on the account level - to be shared across users.

Clearing filters: to clear applied filters click on the “Clear” button

Coming soon: we will gradually add additional filter options to each category

Notes:

0$ Orders are automatically filtered

How to save / share Cohort data?

CSV Download:

Click on the download button and automatically a CSV containing the cohort data will be downloaded.

Share:

Click on the share button and then Google Sheets:

Confirm Title name and Google Sheets Account and click on save.

Note: if your account does not appear then goto Settings/Integrations/Google Sheets and click on “Manage”.

Did this answer your question?