Customer Lifetime Value

Author: Robin Jia，Tommy Lin

Last Update：2024/10/10

1. Model Introduction

The Customer Lifetime Value model focuses on evaluating the overall value contributed by individual customers from a long-term perspective. By analyzing users' overall purchase behavior from their first contact with the brand to churn, this model can help evaluate the total value (CLV) created by individual users for the brand. Generally speaking, brands hope that the value provided by customers (CLV) is much higher than the cost of acquiring a customer (CAC).

Through this model, the following business questions can be analyzed:

• What is the current long-term profitability level of the brand/product?
• How to measure the future purchase potential of customers in different time cycles under the brand/product?
• How long does it take to recoup the advertising cost from past customers?

2. Model Interpretation

This model mainly consists of two parts. The first part shows the overall customer lifetime situation, and the second part shows the monthly customer lifetime and other indicator situations, and provides a cost recovery locator.

In the predefined model, the only selectable option is the time range, which is fixed at a one-year data cycle.

The overview section includes overall indicator cards within the selected time and a trend chart of the overall LTV and ROAS of customers in each month after purchase.

Here, an auxiliary line for customer acquisition cost is provided. If the blue bar chart in a certain time period is higher than this line, it means that the average sales contribution per user exceeds the average customer acquisition cost.

The monthly breakdown section presents the purchase behavior-related data of customers who made purchases in different calendar months over the past year, in each subsequent month as time progresses. The further away from the present, the more data there is, and the closer to the present, the less data there is. Based on the average at the bottom, we can forecast the trend for the whole year.

Here, the cards at the top can be interacted with by clicking to select the displayed metrics. The customer acquisition cost metric is a fixed value and cannot be selected. It is used as a data reference on the left side of the table. The cumulative value switch in the upper right corner of the heatmap can determine whether the model's monthly data is cumulative.

Through the data and the trend of card colors in the heatmap, horizontally, it can be found that the performance of customers who purchased in certain months is better/worse than other months. Vertically, it can be observed that users have fixed repurchase behavior after a fixed time of purchase.

This chart tracks users who made a purchase in a given month and their subsequent purchasing behavior. The "acquisition month" column represents the month when users made their first purchase. Moving from "first purchase" to "first purchase month" and then to "2nd Month," "3rd Month," etc., on the horizontal axis, the chart shows the ongoing behavior of these users.

• "First purchase" refers to data from the users' initial transaction
• "First purchase month" reflects any additional purchases made by these users in the same month as their first order (since some users may buy multiple times in the first month).
• "2nd Month," "3rd Month," and subsequent months display the purchasing data from these users in the following months.

Clicking the filter option in the upper right corner can set the cost recovery locator. After turning on the switch, the heatmap will mark the number of calendar months after purchase in which customers from different calendar months recouped the advertising cost. Here, you can customize the profit margin of the brand/product to calculate the actual profit and further know the actual time of advertising cost recovery.

3. Model Customization

The custom model conditions of this model provide the function of filtering by ad account/ad type/ad campaign, so as to understand the data situation of a specific ad type/ad campaign, and support Amazon Purchase Data Insights to expand non-ad reach traffic data. In addition, you can also define the specific ASINs purchased by customers to only count the purchase behavior data for some products.

Since the CLV model requires a sufficiently long data cycle, the time range of the custom model needs to be selected to be greater than 180 days to be created.

4. Model Data

An important component of this model's data: LTV, compared to ROAS, this metric has the following characteristics in measuring advertising marketing effectiveness

• LTV focuses on the long-term value of customers, while ROAS measures the short-term impact of specific ad campaigns. This means that CLTV can provide brands with a more comprehensive perspective that changes over time.
• ROAS only considers the revenue and conversions directly brought by ads, while LTV considers both the conversions directly brought by ads and the conversions not brought by ads, thus comprehensively reflecting the entire purchase journey of users.
• LTV considers customers' repeat purchase behavior and the potential to purchase other products around the brand halo, while ROAS only considers the revenue generated from initial sales.
• LTV provides a more accurate profitability measurement metric. It considers the cost of acquiring customers, the cost of retaining customers, and long-term fixed costs that occur such as production costs and shipping costs. In contrast, ROAS does not consider customer acquisition costs, which may lead to inaccurate profitability calculations.
• LTV can support more targeted advertising strategies, focusing on customer retention and loyalty, thereby bringing higher profits over time.

Glossary