Getting the price right is one of the fast ways to gain profit. But setting a price is tricky, as companies want to offer the best price to their customers, not the highest price. Apart from the conventional ways like measuring the price elasticity of demand, using data can be crucial to make the best pricing decisions.
What is price sensitivity
Price sensitivity is basically how price change would influence the customer’s decision. It is usually measured with the price elasticity of demand or the change in demand as a function of its price change.
what factors affect price sensitivity
There is a long list of those factors, here is just a few important ones.
- quality
- Unique value: outstanding product
- Sense of urgency: customer need it right away
- Ease of comparison: compare to other brands
- Brand perception
Price sensitivity and price strategy
In order to price a product, many marketing activities involve such as understanding market demand, economic patterns, and level of competition. Based on those factors, there are different pricing strategies
- cost based: price need to cover the cost and desired profit
- demand based: price elasticity of demand
- competitor based: set price to improve market share
- value based: estimated value of a product to a consumer rather than according to the cost
Sensitivity analysis with data and machine learning
Traditionally, many questionnaires are delivered to understand people’s willingness to pay in different ranges. Then further analysis would be conducted using the price elasticity of demand. Price elasticity of demand is a measure used to show the responsiveness (elasticity) of quantity demanded of a product or service to a change in its price given all other factors are constant or remain unchanged. The concept is when the price of a product rises, the demand for that product will fall.However, it is limited to solely include price and quantity in consideration.
Nowadays, increasing studies focus on learning customers and pricing through data that is already available in-house. Data-driven models are built integrating more considerations.
Specific industries like retail have already employed data-driven models to set up smart prices. At first, factors that could affect the market situation, demand, and customer characteristics are identified. Then a tailored algorithm would be created to take those factors as input and output the prediction: the suggested price.
Here is some related use cases.
Dynamic price of Uber:
We all noticed that the price would be 1.8x or more when the number of the driver is less than the requests, which is a classical demand and supply adjustment for price. Moreover, experts would build models to predict the demand in extreme cases like Christmas. Those time series data prediction can be done using models like LSTM.
Hotel room price:
Large hotel chains like Marriot, Hilton are all adopted dynamic pricing with advanced data analytics. Model is trained based on internal data, including historical occupancy information, room type, and daily rates. Influential factors like reservation behavior and customer type are also taken into account. External data is also used, such as weather, events. For the past two years, Covid has been another crucial external dependency.
In a word, a machine learning system can make suggestions based on the information of what is happening in the market as well as referring to historical data. This would benefit the managers to make better price decisions over their competitors.
Reference:
- Vegar Paulsen, Using a three-step-approach to explore the benefits of predicting price sensitivity, https://www.bearingpoint.com/en-fi/our-success/insights/the-future-of-price-sensitivity-analysis-smarter-predictive-pricing-for-an-agile-marketplace/
2. altexsoft, Dynamic Pricing Explained: Machine Learning in Revenue Management and Pricing Optimization, https://www.altexsoft.com/blog/datascience/dynamic-pricing-explained-use-in-revenue-management-and-pricing-optimization/
