THESIS
2020
xiii, 82 pages : color illustrations ; 30 cm
Abstract
Retailers selling perishable products for which the demand is random aim at yielding the maximum possible
profit before the sales horizon ends. Therefore, practical questions arise: How much should they order? How
should they price their products? Indeed, production lead time is very long in many industries, preventing
replenishment during the season. Moreover, items unsold at the end of horizon create losses for the firm since
they can only be salvaged at a negligible value. As a result, a good supply decision is essential. In addition,
deciding how much to charge for the product at any moment is also fundamental: the firm wants to avoid
losing clients or income by pricing too high or too low.
This thesis focuses on the pricing and supply decisions that can most benefit the firm...[
Read more ]
Retailers selling perishable products for which the demand is random aim at yielding the maximum possible
profit before the sales horizon ends. Therefore, practical questions arise: How much should they order? How
should they price their products? Indeed, production lead time is very long in many industries, preventing
replenishment during the season. Moreover, items unsold at the end of horizon create losses for the firm since
they can only be salvaged at a negligible value. As a result, a good supply decision is essential. In addition,
deciding how much to charge for the product at any moment is also fundamental: the firm wants to avoid
losing clients or income by pricing too high or too low.
This thesis focuses on the pricing and supply decisions that can most benefit the firm and the consumers.
We are particularly interested in uncovering whether a solution based on supply analytics (such as the
Newsvendor) is better or worse than a solution based on demand analytics.
This work is divided into two parts. In the first part, time is modelled as continuous so that pricing
decisions can be made at any moment. In the second part, the sales horizon is represented as a sequence of
stages, only allowing price changes at the beginning of each stage. We make computations of the supply and
pricing decisions using Python to calculate the associated expected profits. Then, we resort to simulations
to derive more information about the variability of this profit and the consumer’s surplus. We find that, in
most cases (when the cost is not too low), it is more beneficial for the firm to focus on demand analytics.
Moreover, it appears that this choice might often benefit the consumers. However, our study focuses on the
single-product case and further research is needed for multi-product application.
Post a Comment