Logistic Growth Modeling

In this lesson, you will learn how to use SQL to project when a product’s adoption or usage growth will reach its "ceiling," moving beyond linear trends into the reality of market saturation.

Why Linear Trends Fail

While simple linear regression can tell you how many users you added last month, it fails to account for the finite nature of markets. Eventually, every product experiences a "diminishing return" phase as it nears its total addressable market. Logistic growth modeling acknowledges this by introducing a carrying capacity ($L$)—the maximum limit that your data points will eventually approach.

The Logic of the S-Curve

A logistic growth model creates an "S-curve." It starts with exponential growth, hits an inflection point, and then slows down as it approaches $L$. Mathematically, we define the expected value at time $t$ as: $f(t) = \frac{L}{1 + e^{-k(t-t_0)}}$

Where:

• $L$: The saturation point (e.g., total potential users).
• $k$: The steepness of the growth.
• $t_0$: The midpoint of the growth curve.

Implementation in SQL

To implement this, you typically calculate the parameters ($L, k, t_0$) using historical data and then store them in a configuration table. You then project future growth by joining your time-series dimension against this model:

SELECT 
    time_period,
    -- The Logistic Growth Formula
    L / (1 + EXP(-k * (time_period - t0))) AS projected_adoption
FROM model_parameters
CROSS JOIN time_dimension
WHERE time_period <= 24; -- Projecting out 24 months

By applying this model, you can identify if your current usage growth is healthy or if it is signaling that you are approaching the saturation limit of your existing user base. If your actual cumulative_usage starts to deviate downward from this calculated curve, it may indicate a churn issue or a sudden drop in market interest, providing a much more accurate "Data Quality Gate" than a simple linear forecast.

Key Takeaway: Logistic growth modeling allows us to replace unrealistic straight-line projections with "S-curve" saturation models, providing a more accurate representation of long-term product adoption and market limits.