Cohort Analysis

Cohort Analysis is a powerful behavioral analytics tool that slices data into groups of people, or 'cohorts,' who share common characteristics over a specific time frame. Think of it as a class photo for your customers. Instead of looking at one giant, messy group picture of all your users, cohort analysis groups them by their “graduation year”—the month or year they first signed up or made a purchase. By tracking each of these classes over time, investors can uncover deep insights into business trends that a simple look at overall revenue or user numbers would completely miss. For a value investing practitioner, this isn't just a fancy chart; it's an X-ray into the health and durability of a company's customer base, especially for businesses with recurring revenue models like SaaS (Software as a Service) companies. It helps answer the most critical question: are customers sticking around, and is their loyalty growing?

Imagine a company proudly announcing it added 10,000 new users last month. Sounds great, right? But what if it lost 12,000 existing users during the same period? Top-line numbers can be deceptive. They are like looking at the total volume of water in a bucket without noticing it's full of holes. Cohort analysis is the tool that lets you inspect those holes, measure how big they are, and see if the company is getting better at plugging them over time. For an investor, understanding customer retention is paramount. It is a direct indicator of a company's product-market fit, customer satisfaction, and, ultimately, its competitive moat. A business that consistently retains its customers has a sticky product and pricing power. It spends less on acquiring new customers to replace the ones who leave (a lower churn rate), leading to higher profitability and a more predictable stream of cash flow. Cohort analysis lays this all bare, transforming abstract concepts like “brand loyalty” into cold, hard data.

A cohort chart might look intimidating at first, but it's quite simple once you know the layout. It’s typically presented as a triangle-shaped table.

• The Rows: Each row represents a distinct cohort, usually labeled by the month or quarter they joined (e.g., Jan 2023, Feb 2023, etc.). The oldest cohorts are at the top, and the newest are at the bottom.
• The Columns: The columns represent the passage of time from the cohort's starting point. This is usually labeled as Month 0, Month 1, Month 2, and so on. Month 0 is the month of acquisition, representing 100% of the initial group.
• The Cells: The data points within the table show the percentage of the original cohort that is still active or paying at that specific point in time. For example, if the cell at the intersection of the “Jan 2023” row and “Month 6” column shows 45%, it means 45% of the customers who signed up in January 2023 were still active six months later.

As an investor, you're a detective looking for clues about the business's health. Here’s what to focus on:

1. The “Smiling” Curve or Upward Trend: Look down the columns. For any given month (e.g., Month 6), are the retention percentages for newer cohorts (lower rows) higher than for older cohorts? If the numbers in a column are improving as you go down, it’s a fantastic sign. It shows the company is improving its product, onboarding, or support, making the service “stickier” for new users.
2. The Stabilization Point: Look across the rows. It's normal for retention to drop in the first few months. The key is to see where it flattens out. If a company’s retention for a cohort drops to 40% by Month 3 but then hovers around 40% for the next two years, it means they have a strong, loyal core user base. This stable percentage is crucial for forecasting future revenue.
3. The Holy Grail: Net Revenue Retention > 100%: The most sophisticated cohort charts track revenue, not just users. If a company shows a cohort's revenue retention is over 100%, it's an exceptional sign. This means that the revenue gained from remaining customers in that cohort (through upgrades or buying more) is greater than the revenue lost from the customers who left. This phenomenon, often called “negative churn,” is a powerful engine for growth and a hallmark of elite SaaS businesses.

Let's say you're analyzing two fictional streaming companies.

• StreamFlix: Its cohort chart shows that for every new group of sign-ups, only 20% are still subscribed after three months. Worse, the retention for newer cohorts is the same as for older ones. This signals a “leaky bucket” business. They might be spending a fortune on marketing to attract users who don't see long-term value. This is a red flag.
• BingeMore: Its cohort chart shows that its first-year cohorts retained 40% of users after 12 months. But its most recent cohorts are retaining 55% of users after 12 months. The retention rate is improving with each new class of customers! This tells you the company is strengthening its moat and building a more sustainable business. It's a clear green light for a value investor.

While incredibly useful, cohort analysis isn't a silver bullet. Keep these points in mind:

• Data Availability: Most private companies and even many public ones don't share detailed cohort charts. Investors often have to rely on management commentary or piece together clues from earnings calls and investor presentations.
• Context is Key: A cohort's behavior can be influenced by one-off events. For instance, a cohort acquired during a promotional “90% off” sale might have a much lower retention rate than one acquired through normal channels. Always consider the context behind the numbers.
• It's a Rear-View Mirror: Cohort analysis shows past performance. While it's a great predictor of future trends, it's not a guarantee. The competitive landscape can change, and a company's ability to retain customers could deteriorate. It's a tool for analysis, not a crystal ball.