Cohort reports group your customers by the month of their first purchase and track how each group performs over the following months. Instead of a single average across all customers, you see how value actually accumulates for each generation of customers.
This is the clearest way to answer questions like: how much is a customer worth after 6 months? Are the customers we acquired this quarter better or worse than last quarter's? How long does it take for a cohort to pay back its acquisition cost?
What you can do with cohort reports
Track LTV over time for each monthly cohort, from month 0 (the first purchase) onward.
Compare cohorts against each other to see whether newer customers are more or less valuable than older ones.
Spot your payback point, the month when a cohort's accumulated value covers its acquisition cost.
Switch metrics to analyze cohorts through different lenses beyond LTV.
Add cohort widgets to dashboards to keep this view alongside your other metrics.
Before you start
You need enough history. Cohort analysis gets meaningful with at least a few months of order data. Very recent cohorts will naturally show fewer filled months.
Connect your ad platforms if you want CAC and payback insights. Cifra needs your acquisition spend to calculate when a cohort breaks even.
Reading the cohort table
The cohort table is a grid:
Each row is a cohort: all customers whose first purchase happened in that month.
Each column is a period: Month 0 is the month of first purchase, Month 1 is the following month, and so on.
Each cell shows the metric's value for that cohort at that point in its life. Color intensity makes patterns easy to spot: the stronger the color, the stronger the value.
Reading a row left to right shows how one cohort matures. Reading a column top to bottom compares all cohorts at the same age.
Step 1 — Open the cohort report
Go to Reports and select the cohort report. The table loads with your default date range and metric.
Step 2 — Choose your metric
Use the metric selector to choose what each cell displays. LTV is the most common choice for cohort analysis, but you can switch metrics depending on the question you're answering.
Step 3 — Set the date range
Pick the range of cohorts you want to analyze. Cohorts are grouped in monthly buckets, so your selection adjusts to whole cohort periods to keep every row complete and comparable.
Step 4 — Explore the details
Hover over any cell to see the details behind the number: the cohort, the period, and the underlying values.
When you're analyzing LTV with the payback point enabled, the tooltip also includes the cohort's acquisition economics, so you can see how the accumulated value compares against CAC at that point in the cohort's life.
Step 5 — Mark the payback point
Enable the payback point option to highlight, for each cohort, the month where accumulated value covers acquisition cost. This turns the table into an immediate answer to "how long until a customer becomes profitable?"
If a cohort never reaches the marker within the visible range, it hasn't paid back yet, which is a signal worth investigating: either CAC is too high or early repeat purchasing is too low.
Add a cohort widget to a dashboard
You can add the cohort view as a widget on any dashboard:
Open the dashboard and add a new widget.
Select the cohort widget type.
Configure the metric and options.
The widget follows the dashboard's date range. Because cohorts work in monthly buckets, the widget adjusts the range to the nearest complete cohort period, so the numbers in the widget always represent full cohorts.
How to interpret common patterns
Rows getting weaker over time (newer cohorts underperforming older ones at the same age): acquisition quality may be declining, or you're reaching a colder audience.
A strong Month 0 but flat months after: customers buy once and don't return. Retention and post-purchase marketing are the levers.
Payback point moving later in recent cohorts: CAC is rising faster than early customer value. Review channel mix and offers.
One standout cohort: check what happened that month, like a promotion, a product launch, or a press feature. Standouts often reveal repeatable plays.
Tips
Judge cohorts at the same age. Comparing a 3-month-old cohort's total against a 12-month-old cohort's total is misleading. Compare Month 3 to Month 3.
Recent cohorts are incomplete by definition. Empty cells on the right side of recent rows are expected, not missing data.
Averages hide the story. A stable blended LTV can hide improving old cohorts and deteriorating new ones. The table shows both.