> For the complete documentation index, see [llms.txt](https://docs.roivenue.com/roivenue-resources/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.roivenue.com/roivenue-resources/roivenue-training/roivenue-app/reports/cohort-analysis.md).

# Cohorts

Cohorts shows what the customers you acquire are worth over time: how many come back, how fast they buy again and how much revenue they bring, by acquisition period or channel.

Most reports tell you how a channel performed in a period. Cohorts tells you what the customers it brought are worth later. It groups customers by when they first bought, or by the channel that brought them, and follows each group over time. Instead of "which channel is cheapest", you can ask "which channel brings the most valuable customers".

Use Cohorts when you want to know:

* whether newer customers are more or less valuable than customers you acquired six months ago,
* which campaign brings one-off buyers and which brings repeat customers,
* whether a launch, a price change or a promo brought a better or worse cohort than usual,
* whether customers from one country, device or product line come back more often than others.

The page has six parts:

1. **Header:** Share, Save, currency selector, attribution model and date range.
2. **Cohort KPIs:** six cards with the value of the customers you acquired.
3. **Cohort Settings:** metric, dimension, granularity and number of periods. The settings apply to the whole page.
4. **Customer Retention Cohort Table:** each cohort in one row, its development over time in columns.
5. **Cohort Performance Overview:** the cohorts as lines in one chart.
6. **Filter Panel:** filters for the whole page. When you close it, open it again with the filter icon in the header.

To open the page, go to **Reports → Cohorts** in the left navigation. The date range decides which customers you analyze. The attribution model matters only for channel dimensions. See [Date Picker](https://docs.roivenue.com/roivenue-resources/roivenue-training/roivenue-app/date-picker) and [Filter Panel](https://docs.roivenue.com/roivenue-resources/roivenue-training/roivenue-app/filter-panel).

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2FPc8eY9fI6edUQ5euIvoG%2F01-cohorts-overview.png?alt=media&amp;token=789ff480-fcd7-4bd9-8fd0-5285f2691597" alt=""><figcaption></figcaption></figure>

Cohorts need a transaction feed. See [What Cohorts needs](#what-cohorts-needs).

***

### How to set up the page

**Cohort Settings** apply to the table and the chart.

| Setting               | What it does                                                                                                                                                                               |
| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| **Metric**            | What the table and the chart measure, for example **Conversions**, **Net Revenue** or **Gross Profit**. The list contains metrics from your transaction feed.                              |
| **Dimension**         | How customers are grouped. **Date** groups them by when they first bought. A channel dimension, for example **Platform Name** or **Campaign (auto)**, groups them by where they came from. |
| **Granularity**       | The length of one period: **Daily (1 day)**, **Weekly (7 days)**, **Monthly (30 days)**, **Quarterly (90 days)** or **Yearly (365 days)**.                                                 |
| **Number of periods** | How far into each cohort's life you follow it, so how many columns you see after **Base**.                                                                                                 |

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2FrpLuzcEXqXOgH0BfnNnz%2F02-cohorts-settings.png?alt=media&amp;token=cf6a82af-fe33-4cc3-9a54-cc377d44695d" alt=""><figcaption></figcaption></figure>

The date range at the top decides who counts. Cohorts include every customer whose first purchase falls into the selected period.

{% hint style="info" %}
**Tip:** Align the date range with the granularity, for example whole weeks for **Weekly (7 days)** or whole months for **Monthly (30 days)**. The first and last cohorts then cover full periods.
{% endhint %}

***

### How to read the KPI cards

**Cohort KPIs** show the headline numbers for the customers acquired in the selected period. Roivenue follows these customers up to today, not only to the end of the period.

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2FjRoI0r0ErpMKSy815wfw%2F03-cohorts-kpis.png?alt=media&amp;token=8129991e-3876-449a-af12-298682078649" alt=""><figcaption></figcaption></figure>

| KPI                                    | What it tells you                                               | How it is calculated                            |
| -------------------------------------- | --------------------------------------------------------------- | ----------------------------------------------- |
| **Customer Acquisition Cost**          | What you paid on average to win one new customer.               | Total cost ÷ new customers                      |
| **Avg. Order Value**                   | How much a typical order from these customers is worth.         | Total revenue ÷ total orders                    |
| **Avg. Time to Second Purchase**       | How quickly new customers come back for their second order.     | Average days between the 1st and the 2nd order  |
| **Returning Customer Rate**            | The share of new customers who bought again at least once.      | Customers with 2 or more orders ÷ all customers |
| **Avg. Number of Orders per Customer** | How often an average customer buys.                             | Total orders ÷ customers                        |
| **Avg. Lifetime Value (LTV)**          | The total revenue an average customer brings across all orders. | Total revenue ÷ customers                       |

The cards update when you filter the page, for example to one channel or campaign. To add a card to a dashboard, click the dashboard icon on the card.

***

### How to read the Customer Retention Cohort Table

Each row is one cohort: a group of customers acquired in the same period or through the same dimension value. Each column is one step later in the cohort's life.

* **Date** or **Dimension** is the cohort. Use the filter icon in the column header to find a cohort by name.
* **New Customers** is the number of customers in the cohort.
* **Base** is the period in which the customers first bought.
* **+7d**, **+14d** and further columns are fixed windows after the base period. Their length follows the granularity. With **Monthly (30 days)**, each column is 30 days, not a calendar month.
* Each cell shows the value of the metric and its percentage of **Base**.
* The **Total** row summarizes all cohorts.
* Colour shading shows which cohorts grow and which fade.
* Hover over a cell to see the cohort, the exact dates of the period and the value.

When the table has more columns than fit the screen, scroll it to the side. **How It's Calculated** in the table header opens this page.

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2FEZRxFeleob8mPo6jd6cS%2F04-cohorts-table-cumulative.png?alt=media&amp;token=e108888b-65cc-4993-8a54-84967d79c508" alt=""><figcaption></figcaption></figure>

Columns that lie in the future stay empty. A cohort acquired last month cannot have a value for +90d yet.

{% hint style="info" %}
**Example:** Granularity **Monthly (30 days)**, metric **Net Revenue**, dimension **Date**. The January row shows EUR 50,000 in **Base** and 124% in +30d with **Cumulative**. Customers acquired in January bought for EUR 50,000 in January and for another EUR 12,000 in the next 30 days.
{% endhint %}

#### Cumulative and Absolute

The **Data view** selector in the table header changes what each column means:

* **Cumulative** shows a running total. Use it to see how much a cohort has brought so far.
* **Absolute** shows only the single period, marked with **+**. Use it to see when customers buy and when activity drops.

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2FoerKLdXlT6dCrQ6Rtano%2F05-cohorts-table-absolute.png?alt=media&amp;token=c56d993e-369e-41d0-90f0-ee027680715e" alt=""><figcaption></figcaption></figure>

#### Date and channel dimensions

With the **Date** dimension, cohorts go from the oldest to the newest. With other dimensions, the cohorts with the most new customers come first.

With a channel dimension, **New Customers** can be a decimal number. Roivenue splits each customer between the channels in the journey, see [How customers are counted](#how-customers-are-counted).

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2Fj38WKTOgVG4eXxA8Jtzi%2F06-cohorts-table-platform.png?alt=media&amp;token=78f70d16-4099-4889-ab82-745c2cc4b936" alt=""><figcaption></figcaption></figure>

***

### How to read the Cohort Performance Overview

The line chart puts every cohort on the same starting line, so you can compare their shape no matter when they started. Each line starts at 100% in **Base** and shows how the cohort develops against its own start.

* The chart shows the first 5 cohorts by default.
* In **Dimension Value**, choose which cohorts to show. You can compare up to 10 at a time. Use the search to find a value.
* The dotted **Avg of all Cohorts** line shows the average for comparison.
* Hover over the chart to see the percentage and the value of every cohort in one period.

<figure><img src="https://4279879158-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-MGvvrlVUvq4TNL0oYP6%2Fuploads%2FwRza2NDakaNWrRVxXVpo%2F07-cohorts-performance-chart.png?alt=media&amp;token=8978f7ba-d33f-458a-92bc-20ccf5f90940" alt=""><figcaption></figcaption></figure>

***

### How customers are counted

A customer is a unique person identified by a customer ID, with all their orders linked over time. The first order makes them a new customer in a cohort. Every later order counts as a repeat order. A customer is not the same as an order: someone with three orders in January counts as one customer in the January cohort.

How a customer is assigned to a cohort depends on the dimension:

* **Date:** one customer, one cohort. The date of the first order decides the cohort. All later revenue and orders stay in that cohort.
* **Other dimensions**, for example **Platform Name:** one customer can be split across cohorts. A customer journey often touches several channels, so Roivenue splits the customer between them by the credit of the selected attribution model. The channel totals then match the rest of your reporting.

***

### How filters work in Cohorts

Filters follow where a customer started. A customer stays credited to the cohort and channel of the first purchase. When you filter to one channel, you see the customers that channel acquired, including revenue they later brought through other channels.

A few order-level filters apply to all orders instead, for example **Conversion Type**, **Delivery Type** or **Order Status**.

<details>

<summary>Full list of order-level filters</summary>

Conversion Type, Coupon, Customer Birth At, Customer City, Customer Country Code, Customer Email Domain, Customer Email Hash, Customer Gender, Customer Is Company, Customer Latitude, Customer Longitude, Customer Phone Hash, Customer Postal Code, Customer Segment, Customer Street, Delivery Type, Destination, Item Names, Loan Attribution, Order Status, Payment Type, Processing Type, Product Type, Sales Channel, Source System.

</details>

***

### How to save and share the view

* **Save** in the header stores the settings, the date range and the filters as a bookmark. See [Bookmarks](https://docs.roivenue.com/roivenue-resources/roivenue-training/roivenue-app/bookmarks).
* **Share** copies a link to the current view.
* **Export to XLSX** (download icon in the table header) downloads the cohort table.
* **Download PNG** (download icon in the chart header) downloads the chart.
* **Add to Dashboard** (dashboard icon) adds the table, the chart or a KPI card to a dashboard. See [Dashboard](https://docs.roivenue.com/roivenue-resources/roivenue-training/roivenue-app/dashboard).

***

### What Cohorts needs

Cohorts needs at least one site with a matched **transaction feed**, the order data from your order management system. Without it, the page shows "This feature is not included in your plan." Contact your Customer Success Manager or submit a request in the [Roivenue Help Center](https://docs.roivenue.com/roivenue-resources/roivenue-help-center.md) to set it up.

***

#### FAQ

<details>

<summary>1. Why can't I open Cohorts in my account?</summary>

Cohorts needs a transaction feed connected to at least one site. Without it, the page shows that the feature is not included in your plan. Contact your Customer Success Manager or submit a request in the [Roivenue Help Center](https://docs.roivenue.com/roivenue-resources/roivenue-help-center.md) to set up the transaction feed.

</details>

<details>

<summary>2. Why doesn't +30d match the next calendar month?</summary>

Columns after **Base** are fixed windows: 7 days for Weekly, 30 days for Monthly, 90 days for Quarterly and 365 days for Yearly. They count from the end of the base period, not by calendar months. Every cohort gets windows of the same length, so the cohorts stay comparable.

</details>

<details>

<summary>3. What does "Date range doesn't match granularity" mean?</summary>

Cohorts follow your date range. When the range does not start and end with the granularity, the first and last cohorts cover only part of a period. Roivenue flags these cohorts but does not adjust them. Hover over a cohort to see its exact date range, or align the date range with the granularity, for example whole months for Monthly.

</details>

<details>

<summary>4. What does "Data aggregated" mean?</summary>

The selected period is shorter than the granularity, for example two weeks with Monthly. Roivenue then aggregates the data to the whole containing period, so the total can include data outside your date range. Choose a smaller granularity or a longer date range.

</details>

<details>

<summary>5. Why are the last columns of recent cohorts empty?</summary>

These periods have not happened yet. A cohort acquired last month cannot have data for +90d. The cells fill in as time passes.

</details>

<details>

<summary>6. Why do the KPI cards include orders after the end of my date range?</summary>

The date range decides which customers count, not which orders. Roivenue follows the selected customers up to today, so their later orders are part of **Avg. Lifetime Value (LTV)**, **Returning Customer Rate** and the other KPIs.

</details>

<details>

<summary>7. Why do the numbers by channel change when I switch the attribution model?</summary>

With a channel dimension, Roivenue splits each customer across the channels in the journey by the credit of the selected model. A different model splits the credit differently, so each channel gets a different share of customers and revenue. With the **Date** dimension, the model does not matter, because each customer belongs to one cohort.

</details>

<details>

<summary>8. Why does a channel filter show revenue from other channels?</summary>

Filters follow where a customer started. When you filter to one channel, you see the customers that channel acquired and everything they bought later, also through other channels. That is the value the channel brought.

</details>

<details>

<summary>9. Why can't I find a metric or a dimension in Cohorts?</summary>

Cohorts offers only metrics from the transaction feed. Some metrics and dimensions can also be hidden on certain pages. Go to **Data → Metrics** or **Data → Dimensions**, open the settings of the item and check on which pages it is hidden. Uncheck **Cohort Analysis** and save.

</details>

<details>

<summary>10. How is Cohorts different from Performance?</summary>

Performance shows results in the period when they happened. Cohorts shows results by when the customers were acquired and follows them over time. Use Cohorts to judge the long-term value of customers from a channel, not only the conversions in one month.

</details>
