Ecommerce cohort analysis: understand retention and maximize LTV

Introduction: cohort analysis, a time-based view of your customers

Average metrics lie. An overall retention rate of 30% can hide a very different reality: a January cohort retaining 45% of its customers and a March cohort retaining only 15%. Without cohort analysis, you will never see this disparity.

Cohort analysis consists of grouping your customers by a common criterion, typically their first purchase date, then observing their behavior over time. This temporal approach transforms static data into a dynamic view of your ecommerce business.

Unlike traditional dashboards that display global averages, cohort analysis answers precise strategic questions: do customers acquired during sales return as much as others? Is retention improving quarter over quarter? What is the real return on investment for each acquisition channel?

This guide explains how to read a cohort table, which metrics to track, how to leverage this data for your marketing decisions and how Fullmetrix automates the entire process for your online store.

What is cohort analysis?

Cohort analysis is a statistical method that divides a population into homogeneous groups (cohorts) sharing a common characteristic, then tracks the evolution of these groups over time. In ecommerce, a cohort typically represents all customers who made their first purchase during a given period.

The principle is simple: instead of viewing all your customers as a single block, you separate them by acquisition wave. You can then compare each wave's behavior: does the January cohort behave differently from February's? Is retention improving over time?

Types of cohorts in ecommerce

• Acquisition cohort: groups customers by first purchase date (most common). Example: all customers acquired in January 2026.
• Behavioral cohort: groups customers by a specific action. Example: all customers who used a promo code on their first order.
• Channel cohort: segments by acquisition source. Example: customers from Facebook Ads vs Google Ads vs organic traffic.
• Product cohort: groups customers by first product purchased. Example: customers whose first order contained the flagship product vs an entry-level product.

How to read a cohort table

The cohort table is the central tool for this analysis. It presents cohorts in rows and tracking periods in columns. Each cell indicates the percentage of customers from the cohort who made a repeat purchase during the corresponding period.

Interpreting the data

Reading works in two directions. Horizontally, you follow a cohort's evolution over time. The January cohort drops from 100% to 35% at month 1, then 22% at month 2. This means that out of 520 customers acquired in January, 182 repurchased at month 1 and 114 at month 2.

Vertically, you compare cohorts at the same maturity stage. At month 1, the May cohort shows 41% retention versus only 30% for April. This 11-point gap deserves investigation: what changed between April and May? A different welcome campaign? A catalog change? A promotion?

The table's diagonal is also revealing. It shows how your base behaves at a given point in time: in June, you simultaneously have the January cohort at month 5 (13%), the February cohort at month 4 (14%) and so on. This cross-sectional view helps understand the composition of your recurring revenue.

Key metrics for cohort analysis

The retention table is the starting point, but a complete cohort analysis relies on several complementary metrics. Each illuminates a different aspect of your cohort performance.

Retention rate per cohort

The retention rate measures the percentage of customers in a cohort who return to purchase after a given period. It is the most direct metric from the cohort table. In ecommerce, a month 1 retention rate between 20% and 40% is considered acceptable depending on the sector. Above 40%, your product generates strong adhesion. Below 20%, there is a satisfaction or offer relevance problem.

Cumulative revenue per cohort

Beyond customer count retention, cumulative revenue per cohort shows the financial contribution of each acquisition wave over time. A cohort may have average retention but high cumulative revenue if returning customers increase their basket size. Conversely, a cohort with good retention but declining baskets signals an upselling problem.

LTV per cohort

Lifetime Value per cohort is the ultimate strategic metric. It represents the total revenue generated by an average customer in the cohort over their entire lifespan. By comparing LTV across cohorts, you identify the most profitable acquisition periods. If the LTV of the cohort acquired via a Facebook campaign in January is 180 EUR versus 95 EUR for the one acquired via Google Ads in March, you know where to focus your budget.

LTV/CAC ratio per cohort

By crossing cohort LTV with the customer acquisition cost (CAC) for the same period, you get the LTV/CAC ratio per cohort. This ratio tells you whether customers acquired at a given time are profitable. A ratio above 3 is generally considered healthy in ecommerce. A ratio below 1 means you are losing money on that cohort.

How to use cohort analysis in ecommerce

Cohort analysis is not an academic exercise. It is a decision-making tool that directly impacts your marketing investments, product strategy and financial forecasts. Here are the most powerful practical applications.

Identify the best acquisition sources

By creating cohorts per acquisition channel (SEO, Facebook Ads, Google Ads, email, influencers), you discover which sources generate the most loyal customers. It is common to find that a channel with a high CAC produces high-LTV cohorts, making the investment profitable long-term. Conversely, a low-CAC channel may attract customers who never return.

Measure the impact of marketing actions

When you launch a new welcome email sequence, a loyalty program or a packaging redesign, cohort analysis measures the real impact. Compare cohorts before and after the change: if month 2 retention goes from 20% to 28%, you have quantified proof of your action's effectiveness. Without cohort analysis, this improvement would be hidden in global averages.

Refine revenue forecasts

Cohort retention curves enable precise future revenue modeling. If you know your cohorts reach a retention plateau at 12% after 6 months, you can forecast the recurring revenue generated by each new wave of customers. This modeling is far more reliable than projections based on historical averages.

Fullmetrix: automatic cohorts for your store

Building a cohort table manually in a spreadsheet is tedious and error-prone. Formulas quickly become complex, data must be updated monthly and the slightest methodology change requires recalculating everything. Fullmetrix fully automates this process.

Fullmetrix connects directly to your WooCommerce, PrestaShop or Shopify store and automatically generates cohort tables from your real transactional data. Cohorts are calculated in real time, with no CSV exports or data manipulation required.

• Native cohort visualization by month, quarter or year with instantly readable retention heatmaps.
• Automatic cross-referencing with acquisition channels to identify sources that generate the most loyal customers.
• Automatic LTV calculation per cohort with projections at 12 and 24 months.
• Unified multi-store tracking: compare cohorts across your different stores from a single dashboard.
• Automatic alerts when a recent cohort shows abnormally low retention compared to previous ones.

Instead of spending hours building and maintaining manual reports, Fullmetrix gives you access to professional cohort analyses in a few clicks. You focus on interpretation and decisions, not data collection and formatting.

Frequently asked questions about cohort analysis

What is the difference between cohort analysis and retention analysis?

Retention analysis measures how many customers return over a given period, across all customers. Cohort analysis adds a temporal dimension by separating customers by acquisition wave. Global retention is an average; cohort analysis shows the reality behind that average. Two complementary metrics, but cohort analysis always provides richer insights.

How many customers do you need for reliable cohort analysis?

For statistically significant trends, aim for at least 100 customers per cohort. Below that, individual variations skew percentages. If your monthly cohorts are too small, switch to quarterly cohorts to increase group size. The key is that each cohort is sufficiently represented for trends to be reliable.

How often should you analyze your cohorts?

A monthly review is ideal for most ecommerce stores. Each month, examine the previous cohort's month 1 retention, check trends for older cohorts and compare performance across cohorts. For high-volume stores (more than 1,000 orders per month), weekly analysis of recent cohorts allows faster problem detection.

Can you do cohort analysis with Google Analytics?

Google Analytics offers a basic cohort report based on sessions and users, but it does not natively track repeat purchases or revenue per cohort. For complete ecommerce cohort analysis including LTV, cumulative revenue and channel cross-referencing, a specialized tool like Fullmetrix is necessary. GA4 remains useful for engagement cohorts (visits, session duration) but falls short for transactional cohorts.