Forecasting cohort’s data from the cohorts value

Introduction

The task of forecasting the future in business is both highly complex and critically important. Forecasting enables a company to understand what resources are required to hit target metrics, how those resources will be allocated, and how goals will be achieved. All of this reduces operational risks and allows management decisions to be made based on data rather than intuition alone. 

The ideal scenario is to build a predictive model of business process dynamics, which then serves as the foundation for forecasted Profit & Loss (P&L), Cash Flow (CF), and Balance Sheet statements. However, constructing forecasts for these financial documents is extremely challenging and time-intensive. Using unit economics solves this problem in a much simpler way, because the entire model is built on a small set of metrics that are directly tied to business processes the company already understands, controls, and plans. 

To achieve this, by applying Goldratt’s Theory of Constraints, it becomes possible to forecast the optimal configuration of unit economics metrics required to reach the planned target value — for example, gross margin, net profit, or revenue. Based on the forecasted unit economics, a product plan can be built — a month-by-month roadmap of how unit economics metrics will change over the planning horizon. From this product plan, it is then straightforward to derive the P&L, cash flow statement, and balance sheet. 

However, working with unit economics introduces a new challenge for the entrepreneur: understanding and forecasting cohorts. The reason is that unit economics operates in cohorts. All metrics reflect cohort-level values, not calendar time intervals. Yet the business lives, keeps books, and reports to regulators in calendar periods — months, quarters, and years. 

Therefore, modeling future cohorts is an extremely important task for any business, and it is the focus of this article.