Customer Cohort Analysis
Analyze customer retention, lifetime value, repurchase rates, and acquisition ROI by cohort in MerchantFlow. Group customers by first purchase date.
Customer Cohort Analysis
Cohort analysis in MerchantFlow groups your customers by the date of their first purchase, then tracks how each group performs over subsequent periods. This reveals customer retention trends, lifetime value (LTV) trajectories, repurchase behavior, and which acquisition periods produce your most valuable customers.
What Is Cohort Analysis?
A cohort is a group of customers who share a common characteristic -- in MerchantFlow, they are grouped by the week, month, or year of their first order. By comparing cohorts side by side, you can answer questions like:
- Are newer customers spending more over time than older ones?
- Is my retention rate improving or declining?
- Which marketing campaigns acquired the highest-LTV customers?
- How quickly do I recover my customer acquisition cost?
How Cohort Analysis Works
MerchantFlow automatically groups customers into cohorts based on their first online order date. You can then track metrics like revenue, gross profit, repurchase rate, and ROAS across subsequent periods without any manual setup.
Only online (D2C) orders are loaded, so B2B and wholesale orders never seed a cohort or contribute to one -- whether you entered them by hand, reclassified them with Adjust order, or automatic wholesale detection classified them. A buyer whose first purchase was wholesale lands in the cohort of their first storefront order instead. A wholesale-only buyer appears in no cohort at all.
How to Access Cohort Analysis
Navigate to Customers > Cohorts (/dashboard/cohorts). The page is titled Cohort Intelligence.
Four tiles summarise the view: Total Cohorts, Total Customers, Avg Cohort Size, and Cohorts with Payback. The header also carries Ask AI, which opens Flow AI, and Export CSV, which downloads the summary table.
How to Configure Your Cohort View
Grouping Options
- Weekly -- group customers by the week of their first purchase (best for short-term campaigns)
- Monthly -- group by month (default and recommended for most analysis)
- Yearly -- group by year for long-term trend analysis
Metric Options
- Revenue -- total revenue generated by each cohort over time
- Gross Profit -- revenue minus COGS for each cohort (requires COGS setup)
- Repurchase Rate -- percentage of customers in the cohort who ordered again
- ROAS -- return on ad spend for each cohort (requires ad platform connections)
Display Mode
- Accumulated -- running totals across periods (recommended for LTV analysis)
- Period Only -- values for each specific period only (useful for spotting period-specific spikes)
ROAS always uses accumulated mode, regardless of this setting, because a period-only ROAS would divide a single period's revenue by the whole cohort's acquisition spend.
Calculation Mode
- Per Customer -- average per customer in the cohort (normalizes for different cohort sizes)
- Whole Cohort -- total for the entire cohort (shows absolute impact)
Periods
Choose how many periods after acquisition the matrix shows: 6, 8, 10, 13, 20, 26, or 52. The default is 13.
Metric, Display and Calculation are applied in the browser, so changing them redraws the matrix instantly. Grouping and Periods change the query and go back to the server.
Time Range
Cohorts respect the global timeframe selector, and the window bounds everything on the page, not just which cohorts appear. Only orders placed inside the window are loaded, so:
- A customer's cohort is the period of their first online order inside the window. Someone who bought from you two years ago and again last month lands in last month's cohort if the window starts after their original purchase.
- A cohort's later-period revenue, profit and repurchase figures stop at the end of the window. They do not continue accruing from orders placed after it.
Widen the range before reading long-run retention. If you want lifetime figures that ignore the window after acquisition, use Customer LTV instead -- there the timeframe filters acquisition only.
How to Read the Cohort Matrix
The cohort matrix, headed Performance Matrix, is a color-coded grid where:
- Rows represent cohorts, newest at the top (e.g., "2026-08" = customers who first purchased in August 2026)
- A Size column after the cohort label gives that cohort's new-customer count
- Columns represent periods after acquisition (P0 = first period, P1 = second period, P2 = third period, and so on)
- Cell values show the selected metric for that cohort in that period
- Color intensity is scaled per column, so each period is shaded against the best and worst cohort in that same period rather than against the whole grid. A grey cell means no data for that cohort and period.
- A small green dot marks the cell where that cohort reached payback
Cohort labels follow the grouping: 2026-W07 for weekly, 2026-08 for monthly, 2026 for yearly.
Example interpretation: If the "Jan 2026" row shows $50 in P0 and $15 in P1 (accumulated revenue per customer), it means the average January customer spent $50 in their first month and an additional $15 in their second month.
Understanding the Summary Table
Above the matrix, a Cohort Summary table shows key metrics for each cohort:
| Column | Description |
|---|---|
| Cohort | The acquisition period (e.g., "2026-08") |
| New Customers | Number of first-time buyers in that period |
| CAC | Ad spend in that period divided by new customers |
| Revenue LTV | Accumulated revenue divided by new customers |
| Margin LTV | Accumulated gross profit divided by new customers |
| Payback | First period index at which Revenue LTV reaches CAC |
| Profit Payback | First period index at which Margin LTV reaches CAC |
| ROAS | Accumulated revenue divided by the cohort's ad spend |
| Net Margin | Accumulated gross profit as a percentage of accumulated revenue |
Every column is sortable. Payback is expressed as a period index, not days - with monthly grouping, a payback of 2 means the cohort recovered its CAC in the third month.
A cohort's CAC uses all ad spend recorded in the acquisition period, divided by the customers acquired in it. It is not a per-customer attributed cost - for that, see Customer LTV.
What Insights to Look For
- Improving LTV -- are newer cohorts accumulating more revenue over time than older ones?
- Repurchase trends -- is your repeat purchase rate increasing across recent cohorts?
- Payback period -- how quickly do you recover your customer acquisition cost? Shorter is better.
- Seasonal patterns -- do holiday cohorts have different LTV or retention characteristics?
- Marketing effectiveness -- which campaign periods produced customers with the highest LTV and best retention?
Tips for Effective Cohort Analysis
- Start with Monthly grouping and Revenue metric for a broad overview of customer behavior
- Use Per Customer mode to normalize across different-sized cohorts and compare fairly
- Compare Accumulated Revenue across cohorts to identify your best acquisition periods
- Look at P1 through P3 repurchase rates to evaluate whether your retention strategy is working
- Cross-reference cohort data with your Attribution reports to connect marketing campaigns to long-term customer value
Frequently Asked Questions
How far back does cohort analysis go?
As far back as your plan's history window allows: 90 days on Starter, 12 months on Pro, and unlimited on Plus. Within that window it covers all data synced from your Shopify or WooCommerce store. If you select a range that reaches further back than your plan allows, the start date is clamped.
What exactly does the repurchase rate measure?
It is the share of the cohort that has placed two or more orders. In accumulated mode it counts everyone who has reached two orders by that period; in period-only mode it counts just those who crossed the two-order threshold during that period.
Rates vary widely by industry. A P1 (second-period) repurchase rate of 20-30% is often cited for consumer goods e-commerce, and subscription businesses usually run higher -- but that is an external rule of thumb offered for orientation only. MerchantFlow does not compute or publish a repurchase-rate benchmark, and nothing on this page compares your cohorts against other stores. Judge your cohorts against your own earlier cohorts.
How does cohort analysis help with marketing budget decisions?
By comparing CAC, payback period, and LTV across cohorts, you can identify which acquisition periods (and associated campaigns) deliver the best long-term ROI -- then allocate future budget accordingly. See Ads and Channels for campaign-level data.
Can I see cohort data for specific products or channels?
The current cohort view groups all customers by first order date. Product-level and channel-level cohort filtering is on the roadmap.
What is the difference between Per Customer and Whole Cohort mode?
Per Customer divides the total by the number of customers in the cohort, giving you an average. Whole Cohort shows the raw total. Per Customer is better for comparing cohorts of different sizes.
Related Guides
- Attribution -- connect campaigns to cohort performance
- AI Insights -- ask questions about cohort trends
- P&L Overview -- financial context for cohort profitability
- Dashboard Overview -- navigate the full MerchantFlow experience
Last updated: September 16, 2026
Last updated on
Merchant Center Diagnostics & Feed Health
Monitor Google Merchant Center product approvals and fix disapprovals from MerchantFlow. Track critical errors, warnings, and pending reviews across your feed.
Customer LTV - Per-Customer Lifetime Value and Margin Tracking
See every acquired customer with their lifetime revenue, lifetime gross margin, CAC, LTV:CAC ratio, and acquisition channel in MerchantFlow.