Step 1

Preparation of Data. For price analytics, we will have to use historical pricing data which consists of demand at the day or week, or month level, the price at which item was sold, actual price, discount, quantities sold, etc. We have to be sure of the quality of this data before we use it. After this, we will have to remove outliers from it. It will be great to also have stock level info such as opening and closing stock

Step 2

Estimating the elasticity of price or demand elasticity of price is an important step in the pricing process. We will have to evaluate the right function to capture the customer demand. In a textbook-based ideal market, the price elasticity function will be linear and a straight line. If a seller prices the price below market price, the demand for the price will be very high and vice versa. In reality, the price elasticity is not linear for many products. The analysts can try to fit functions in various types of elasticity curves such as linear, exponential, Weibull, etc. Whichever fits best the data can be considered.

Step 3

Forecast the demand using the function found in step 2 using additive or multiplicative forecasting techniques based on the input data.

Step 4

Once step 2 is determined, we can go ahead and set objective functions and constraints and optimize the price. The objective function can be revenue maximization, profit maximization, volume maximization, etc. Whether the seller needs to clear out his inventory or needs to increase the price at which he is currently selling is a business call or input that has to come from the customer. This can be ascertained with the help of the data as well. But we need to have access to the entire historical data. I will be writing a separate chapter on these steps, hence not going deep into each topic.