Customer lifetime value number came from revenue minus cost of each transaction. Align with the
research model that we develop, we use another multiple regression by using SPSS 22 with transaction
frequency and initial monetary value as independent variable as well as customer lifetime value as
dependent variable. We also use descriptive statistics method to show deeper insight from this
phenomenon.
Result
Multiple regression result shows that the research model for study 2 is significant with p < 0.01 (p =
0.0001) and multiple coefficient determination 0.656. It shows that the model explain 65.6% variability of
the response data around its mean. Transaction frequency (p = 0.0001) and initial monetary value (p =
0.0001) are significant (p < 0.01) as independent variable. The positive coefficient explains that both
transaction frequency and initial monetary value positively affect customer lifetime value in the 4 months
window. Through correlation test, initial recency also has significant effect (p < 0.01) to customer
lifetime value as well as transaction frequency. However, when we run multiple regression with
transaction frequency, initial recency, and initial monetary value together as independent variable and
customer lifetime value as dependent variable, initial recency become non-significant (p = 0.274). This
results suggest that transaction frequency acts as mediator for initial recency that affects customer lifetime
value indirectly. Sobel test also shows the significance of transaction of transaction frequency as mediator
with p < 0.01 [40]. Therefore, the research model in study 2 supports hypothesis 3, hypothesis 4, and
hypothesis 5. The model can be used for predict the customer lifetime value with formula:
CLV = -166028,533 + 181477,001*(Transaction Frequency) + 1,261*(Initial Monetary Value) (1)
There are some more interesting insights from study 2. The top 1% customer worth almost 7 times of the
average customer lifetime value. The indirect effect of initial recency to customer lifetime value can be
seen clearly. Top 25% rank of customer lifetime value has 30.72 average days gap between initial
purchase and second purchase. On the other hand, bottom 25% rank of customer lifetime value has 94.57
days of initial interpurchase time.
4. General Discussion and Implication
E-commerce is used to facilitate people to buy and own products that are able to fulfill their needs or
wants with lower searching cost and transaction cost. In contrast, babyloania.com unlock new business
model by adapting product-service system in e-commerce form to provide temporary access, not
ownership, for products to fulfill customers’ needs. For a new concept with new access and new scale,
product-service system needs initial trust to be accepted and to create loyal customers. Within this system,
our findings show that the initial trust was created from the experience in the first purchase. The
customers feel a unique experience compare to other e-commerce portals through access-based price
offered by product-service system based e-commerce. They experience perceived value from the saving
benefit between access-based price and retail price. On their website, Babyloania.com shows the
comparable price of each product lowest usage rental price to retail price. The shown rental price is 5-
10% of average retail price [41].
The results of Study 1 shows that perceived value has a significant effect to transaction frequency.
Product-service system based e-commerce play a different game than conventional e-commerce or store
and does not compete directly with pricing or product or service quality. By using the advantage of
perceived saving benefit, product-service system based e-commerce are able to offer high quality