From Complex Data to Profitable Growth: Using Cohort Analysis to Improve Course-Fill Forecasting

What We Found

For an online education provider, traditional monthly reporting was masking a much longer customer journey than expected. By analyzing customer cohorts and the time between key actions, we found that prospective students could first visit the website nearly a year before starting a course and register as much as four months before the course began.

These insights changed how the organization could approach marketing timing, course planning, and demand forecasting, creating a clearer path toward more consistent and profitable course fill.

The Challenge

An online education provider with a strong reputation, affordable pricing, and highly specialized course offerings had a compelling product but struggled to achieve consistent profitability.

The challenge wasn't a lack of market opportunity. It was a lack of reliable data to understand when demand was being generated, how customers moved through the funnel, and how marketing activity ultimately translated into course registrations.

Historically, performance was primarily evaluated on a calendar-month basis. That approach created challenges for a business with a long customer journey. Registrations could occur months before a course began, meaning the impact of marketing activity often wasn't visible within the month in which the marketing occurred.

Inconsistent and incomplete historical data also made it difficult to forecast course demand or determine how much and when to invest in marketing.

The business needed a better way to connect marketing activity to eventual course demand.

Our Approach

We started by cleaning up and analyzing the data to create a more reliable foundation for decision-making.

Rather than looking exclusively at monthly registrations, we analyzed customer behavior through cohorts and latency curves. This allowed us to follow groups of prospective students through their customer journey and understand the time between key actions.

We focused on two critical relationships:

• First website visit → course start

• Course registration → course start

We also incorporated historical course-fill rates, registration trends, course start dates, website activity, account creation, seasonality, search impression share, and other leading and lagging indicators.

This shifted the question from "How did marketing perform this month?" to "What is the eventual impact of the customers generated by this marketing activity?"

What We Discovered

1. The customer journey was nearly a year long.

Prospective students were entering the organization's ecosystem much earlier than expected. Customers were first landing on the website as much as 336 days before starting a course.

Implication: Marketing courses only in the weeks immediately preceding their start dates risked missing a significant portion of potential demand.

2. Students registered months before their courses began.

Customers could register for a course as much as 113 days before the course start date, with an average registration-to-course-start latency of 61 days.

This demonstrated that registration was not an immediate response to marketing exposure. Students often needed significant time to research, evaluate, plan, and ultimately commit to a course.

Implication: Marketing and registration strategies needed to account for a substantially longer consideration period than previously assumed.

Why Cohort Analysis Mattered

These findings demonstrated why traditional in-month reporting could be misleading.

A campaign launched in May might generate a prospective student who registers in July and starts a course in September. Looking only at May performance would fail to capture the full value of that customer.

Similarly, a strong month of registrations doesn't necessarily translate into immediate course starts or revenue.

By tracking cohorts over time, we could better understand the eventual value of marketing activity and distinguish between short-term performance and the longer-term development of demand. This provided a more accurate framework for evaluating marketing investment and forecasting future course fill rates.

Strategic Recommendations

The analysis led to several changes in how the organization could approach marketing, forecasting, and course planning:

• Market courses well in advance

• Open registration approximately four months in advance

• Evaluate marketing by cohort

• Incorporate latency into forecasting

• Maintain consistent course frequency

• Improve the underlying data infrastructure

The Impact

The analysis changed how the organization could evaluate marketing performance and anticipate future demand.

Rather than treating registrations as isolated monthly events, the organization gained a framework for understanding the full customer journey and the delayed impact of marketing activity.

This graph illustrates latency between a student’s initial landing, registration date, and course start.

The recommendations informed changes to marketing timing, registration windows, course planning, and forecasting. These changes have already contributed to record-setting performance, with the affected cohorts continuing to mature.

Just as important, the analysis created a foundation for ongoing optimization. As additional cohorts mature, the organization can refine its latency curves, improve forecasting accuracy, and better understand the relationship between marketing investment and eventual course demand.

The Takeaway

The biggest lesson wasn't simply that the customer journey was longer than expected. It was that the way performance is measured can shape the decisions made from that performance.

When customer behavior unfolds over months, evaluating marketing one calendar month at a time can obscure where demand is actually coming from and when marketing is influencing it.

Cohort analysis provided a way to see that longer-term picture, turning historical customer behavior into a more useful tool for planning future marketing investment and course demand.

The result: a clearer understanding of how customers move from initial interest to enrollment, and a more informed approach to filling courses profitably.