Growblocks uses predictive cohort analysis to understand the behavior and lifecycle of your customer base over time. By segmenting customers into cohorts based on their start date and other characteristic dimension (e.g. region, company size, product plan, etc.), we can model and predict how groups of customers will expand or contract, helping you model out the evolution of your customer base.
What is a Cohort?
A cohort is a group of customers who signed up for your service during a specific time frame, such as January 2023, that belongs to a certain pre-defined customer segment, such as SMBs in EMEA.
Monthly Snapshots
At the start of each month, we take a 'snapshot' of your customer base. This snapshot includes the age of each customer in months since they first signed up. This means that at the start of February 2023, all customers signed in January 2023 will have an age of 1 month.
Predictive Modeling
Step 1: Select Historical Data Length
This step involves choosing the duration of historical data to calculate average growth rates. The default is 12 months, but it can be adjusted to fit your needs.
Example 1: If your company has recently pivoted its services or significantly altered its product in the last 6 months, you might want to use only this recent data to calculate your growth rates, as it is more representative of your current business model.
Example 2: Conversely, if your company has been steadily growing and your business model has remained consistent, you might want to use a longer time frame, such as 24 or 36 months, to calculate your average growth rate. This can help smooth out any short-term fluctuations and provide a more accurate long-term growth rate.
Step 2: Fine-tune the Terminal Rate
This step requires you to choose which customer cohorts should be used to calculate the terminal rate. The default is to use customers older than 12 months, but can adjust this by specifying the starting and ending cohort.
Example 1: If you have found that customers who started using your service 9-15 months ago demonstrate a stable growth rate, you might want to use this cohort to calculate your terminal rate.
Example 2: If your product is for enterprise and has a long onboarding phase with implementation taking several months, or multi-year contracts, then you may want to use a terminal starting and ending month that is more reflective of this, such as 24-48 months.
Step 3: Apply the Terminal Rate
This step involves deciding when to start applying the terminal rate to your predictions. This will affect your projections for customer expansion or contraction. The default applies this from the 13th month of a customer’s lifetime.
Example 1: If your company has been around for 5 years, but signed most of your customers over the past 24 months, then you might want to apply the terminal rate for all customers older than 24 months to avoid a very long tail.