liveBook · Manning

Revenue management

As explained in Applied Reinforcement Learning

Revenue management is the practice of optimizing prices and inventory availability to maximize revenue over time.

Revenue management balances price and demand over a limited selling window. Price too high and inventory may remain unsold; price too low and it may sell out before customers willing to pay more arrive. section 3.3

For a perishable product, the decision can be modeled as a sequential pricing problem: choose a price each period while tracking time and inventory. Demand responds to price, may rise as the deadline approaches, and includes random variation. Unsold items can incur a penalty. section 3.3

Table 3.2 Elements of Markov decision process for perishable dynamic pricing
Markov decision process elements
Definitions
State
How many sales periods are left.
How many items are left to sell.
Action
The agent chooses a price from a predefined list, e.g., [8, 10, 12, 15, 18].
Reward
Calculated as the revenue from sales in that period.
Penalty if items remain unsold after the last period.
Transition Dynamics
Demand drops as the price increases (price sensitivity).
Demand slightly increases over time due to urgency (people rush to buy near the end).
Random noise is added to simulate uncertainty.

The MDP table maps the business decision to states, price actions, sales rewards, and uncertain demand. section 3.3

A fixed-inventory model assumes no restocking and treats deadline urgency as natural. It does not account for customers learning to wait for late discounts, a possible strategic response. section 3.3

How Applied Reinforcement Learning covers it

How deep it goes
Works through an RL environment with states, price actions, stochastic demand, rewards, and a coded step function.
Examples
Uses a fixed-stock selling window with discrete prices, price-sensitive demand, deadline urgency, random variation, and penalties for unsold items.
What you can do after
Build it from scratch by modeling inventory, time, prices, demand, sales revenue, and unsold-stock penalties in a pricing environment.
How it connects
Connects revenue management to reinforcement learning, Markov decision processes, dynamic pricing, and strategic consumer behavior.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Applied Reinforcement Learning. The section links point to the parts of the book it draws on.