luvv to helpDiscover the Best Free Online Tools
Topic 6 of 8

Channel Based Cohorts

Learn Channel Based Cohorts for free with explanations, exercises, and a quick test (for Product Analyst).

Published: December 22, 2025 | Updated: December 22, 2025

Why this matters

Channel-based cohorts group users by how they were acquired (e.g., Paid Search, Organic, Social, Referral) and then track outcomes like activation, retention, revenue, and LTV over time. Product Analysts use them to answer questions such as:

  • Which acquisition channels bring users who stick around and pay?
  • How fast does each channel pay back its cost?
  • Where should we shift budget this month to hit goals?
  • Are we attracting different user segments by channel that need different onboarding?

Concept explained simply

Think of each channel as a different door into your product. Channel-based cohorts compare what happens after users enter through each door.

Mental model

Picture a set of rows—one per channel—and columns for weeks since signup. Each cell holds a metric (e.g., percent retained). Scan across a row to see the quality of a channel over time. Scan down a column to compare channels at the same age (e.g., Week 4 retention).

Key definitions
  • Cohort: Users grouped by their acquisition channel (optionally also by acquisition week/month).
  • Retention: Share of the cohort active at a later time point (e.g., Week 4).
  • Activation: First key action (e.g., first project created) within a set window.
  • LTV (lifetime value): Cumulative gross profit contributed by a user over time.
  • CAC (customer acquisition cost): Cost to acquire one user/customer from a channel.
  • Payback: Time until cumulative gross profit per acquired user ≥ CAC.

Data you need

  • Acquisition data: user_id, acquisition_date, channel (consistent naming; split brand vs non-brand where relevant).
  • Behavioral events: timestamps for activation, usage, purchases/subscriptions, churn/cancellations.
  • Financials: CAC by channel and period, revenue and gross margin (or margin rate), refunds.
  • Attribution rule: e.g., first-touch within 7 days; document the rule and apply consistently.

How to build channel cohorts (step-by-step)

Step 1: Define channels precisely. Example: "Paid Search — Brand", "Paid Search — Non-Brand", "Organic", "Referral", "Email", "Social".
Step 2: Assign each new user a primary channel by your attribution rule (e.g., first-touch).
Step 3: Pick a cohort time grain (weekly or monthly acquisition).
Step 4: Compute metrics by channel x cohort-age (e.g., Week 1, Week 4, Week 12).
Step 5: Add economics: CAC, gross margin rate, LTV curve, and payback time.
Step 6: Compare channels at the same age; make budget and product recommendations.

Worked examples

Example 1 — Retention by channel

Suppose Week 0 new users: Paid Search 600, Organic 500, Referral 300. Active users at Week 4: Paid 120, Organic 160, Referral 150.

  • Paid Search W4 retention = 120/600 = 20%
  • Organic W4 retention = 160/500 = 32%
  • Referral W4 retention = 150/300 = 50%

Referral has the strongest retention. Weighted overall W4 retention = (120+160+150)/(600+500+300) = 430/1400 ≈ 30.7%.

Example 2 — Payback by channel

Assume CAC and cumulative revenue per user (ARPU cum) with 80% gross margin:

  • CAC: Paid $18, Organic $4, Referral $6
  • 3-month ARPU cum: Paid $13.5, Organic $9, Referral $11
  • 6-month ARPU cum: Paid $22, Organic $16, Referral $20

Gross profit (GP) = ARPU cum × 0.8.

  • Paid: GP3 = $10.8, GP6 = $17.6 → No payback by 6 months
  • Organic: GP3 = $7.2 ≥ $4 → Payback by month 3
  • Referral: GP3 = $8.8 ≥ $6 → Payback by month 3

6-month ROI (GP/CAC): Paid 0.98, Organic 3.2, Referral 2.67. Organic wins on ROI.

Example 3 — Activation rate by channel

Within 7 days of signup, % of users who complete first key action:

  • Paid Search: 42%
  • Organic: 55%
  • Referral: 63%

Higher early activation often predicts better retention and LTV; prioritize onboarding improvements for channels with low activation.

Example 4 — Budget shift logic

If Referral shows strong retention and fast payback but limited volume, and Paid has high volume but poor payback, consider shifting spend from Paid non-brand keywords to Referral incentives and Organic content—while running experiments to improve Paid onboarding.

Metrics to compare across channels

  • Activation rate (by time window)
  • Retention curve (e.g., W1, W4, W8, W12)
  • Conversion to paid / revenue per user
  • LTV at fixed horizons (e.g., 3, 6, 12 months)
  • CAC, Gross Margin, Payback time, ROI at horizon
  • Time-to-first-value (median)

Exercises (do these now)

These mirror the exercises listed below (Exercise 1 and Exercise 2). Use a calculator or spreadsheet.

Exercise 1 — Retention by channel (id: ex1)

Given Week 0 new users and active users at Week 1 and Week 4:

  • Paid Search: new=600, W1 active=210, W4 active=120
  • Organic: new=500, W1 active=230, W4 active=160
  • Referral: new=300, W1 active=180, W4 active=150

Tasks:

  • Compute W1 and W4 retention for each channel.
  • Identify which channel has the highest W4 retention.
  • Compute weighted W4 retention across all channels.

When done, compare with the provided solution in the Exercises section below.

Exercise 2 — LTV, payback, and ROI (id: ex2)

Assume CAC and cumulative ARPU per user with 80% gross margin:

  • CAC: Paid $18, Organic $4, Referral $6
  • ARPU cum (3 mo): Paid $13.5, Organic $9, Referral $11
  • ARPU cum (6 mo): Paid $22, Organic $16, Referral $20

Tasks:

  • Compute GP after 3 and 6 months (ARPU × 0.8).
  • Find payback month for each channel (first time GP ≥ CAC).
  • Compute 6-month ROI = GP6 / CAC. Rank channels.
  • [Checklist] I used the same time horizon for all channels.
  • [Checklist] I kept CAC and margin assumptions consistent.
  • [Checklist] I compared channels at the same cohort age (not calendar date).

Common mistakes and self-check

  • Mixing acquisition date with calendar date. Self-check: Am I comparing W4 retention across channels for users acquired in different weeks but same age?
  • Blending brand and non-brand PPC. Self-check: Are brand and non-brand separated?
  • Ignoring costs. Self-check: Did I include CAC and margin for LTV and payback?
  • Attribution inconsistencies. Self-check: Is the attribution rule documented and applied uniformly?
  • Small-sample overinterpretation. Self-check: Do I flag very small cohorts and avoid strong conclusions?
  • Seasonality leaks. Self-check: Did I compare cohorts within similar seasonal periods or control for seasonality?

Practical projects

  1. Build a weekly channel cohort retention dashboard with W1/W4/W8 curves and volume per channel.
  2. Create a payback tracker by channel using CAC, margin, and LTV at 1, 3, 6 months.
  3. Run a 2-week experiment to improve activation for the weakest channel and report impact on W1 retention.

Who this is for and prerequisites

Who this is for
  • Product Analysts and Growth Analysts working with acquisition and retention.
  • Marketers collaborating with product on onboarding and activation.
Prerequisites
  • Basic cohort analysis and retention metrics.
  • Comfort with spreadsheets or SQL for aggregations.
  • Understanding of CAC and margin concepts.

Learning path

  1. Review cohort fundamentals (definitions, cohort age).
  2. Define channels, attribution, and data sources.
  3. Build retention tables and charts by channel.
  4. Add economics: CAC, margin, payback, LTV.
  5. Make recommendations and run follow-up experiments.

Next steps

  • Complete the two exercises below.
  • Take the Quick Test. It is available to everyone; only logged-in users will have their progress saved.
  • Pick one practical project and implement it this week.

Mini challenge

In one paragraph, recommend a budget shift across two channels based on hypothetical W4 retention and payback results. State the data you used and any assumptions.

Practice Exercises

2 exercises to complete

Instructions

Given Week 0 new users and active users at Week 1 and Week 4:

  • Paid Search: new=600, W1 active=210, W4 active=120
  • Organic: new=500, W1 active=230, W4 active=160
  • Referral: new=300, W1 active=180, W4 active=150

Tasks:

  • Compute W1 and W4 retention for each channel.
  • Identify which channel has the highest W4 retention.
  • Compute weighted W4 retention across all channels.
Expected Output
W1: Paid 35%, Organic 46%, Referral 60%. W4: Paid 20%, Organic 32%, Referral 50%. Best W4: Referral. Weighted W4 ≈ 30.7%.

Channel Based Cohorts — Quick Test

Test your knowledge with 8 questions. Pass with 70% or higher.

8 questions70% to pass

Have questions about Channel Based Cohorts?

AI Assistant

Ask questions about this tool