How to forecast revenue per customer?
Forecasting Revenue per Customer entails estimating the average revenue generated from each customer over a specified period.
This process considers various input factors, including purchase frequency, average order value (AOV), and customer segmentation.
Example:
Consider an Indian e-commerce platform in Q1 2024.
The platform's returning customers have an AOV of ₹9,000 and a purchase frequency of twice a month, while new customers have an AOV of ₹6,000 with a purchase frequency of once a month.
If there are 200 returning and 100 new customers, the revenue forecast would be:
Revenue from Returning Customers = 200 × ₹9,000 × 2 = ₹36,00,000
Revenue from New Customers = 100 × ₹6,000 × 1 = ₹6,00,000
Total Revenue = ₹36,00,000 + ₹6,00,000 = ₹42,00,000
Average Revenue per Customer = ₹42,00,000 / (200 + 100) = ₹14,000
Step-by-step Explanation:
Forecasting Revenue per Customer thus involves forecasting retention rates by cohort, as well as average order value and purchase frequency by cohort.
As in the case of projecting retention rates, the projection of average order value and purchase frequency follows same 3-step process:
1. Breaking down existing customers into more granular cohorts, since different cohorts have different behavior.
2. Projecting baseline for retention rate for each cohort, using historical trend for that cohort.
3. Applying any seasonal or ad-hoc corrections (impact of new price changes, promotions) to the baseline, that have not been factored into the trend.