It isn't very easy to imagine modern pricing and profitability management without using advanced pricing analytics. There are many examples of large corporations that acquire and store vast amounts of data but don't have any idea how to examine trends and find patterns that drive valuable insight. Implementing pricing analytics techniques can be broken down into three main areas: descriptive, predictive, and prescriptive analytics.

 

Descriptive Analytics

 

The use of pricing analytics typically starts with a historical assessment of data that can help deliver bottom-or or top-line improvements over a short time. This involves uncovering price alignment opportunities and profit leakages and tracking margin drivers.

  Pros of Descriptive Analytics

  • Cost-effective and fast
  • Helpful for decision making

  Cons of Descriptive Analytics

  •  Possibility for error and subjectivity
  •  Only reflect past performance

In addition to better understanding their past performance, descriptive analytics can benefit companies in the following ways.

 

Price Waterfalls

 

What are Price Waterfalls?

Price Waterfalls illustrate where the margin is being lost at the most granular level. Most companies only focus on setting invoice prices and tend to lose sight of the fact that many discounts, allowances, and rebates are not included on the invoice.

Example of Price Waterfall:

A potential price waterfall illustrates how a manufacturer is giving up a substantial portion of its margin by providing a $25 trade allowance to a retailer. In this situation, the manufacturer may need to renegotiate the terms of the trade allowance to realize more margin.

How do you take action on this?

  • Identify any profit leakages you did not know about
  • Determine what channels or products are least profitable
  • Set expectations for ROI

Learn more about price waterfalls here.

 

Price Dispersion Charts

 

What are Price Dispersion Charts?

Price dispersion charts can show the sales associated with every customer in your business and their relative price position for each specific product. If your company has a wide range of customers, from small retailers to big distributors, there is an opportunity to see how prices for the same products are spread across each. Price dispersion gives you a clear picture of whether there are price misalignments in terms of customer size and provided discounts.

 

Example of Price Dispersion Chart:

By building price dispersion charts, a business can quickly identify customers that are mispositioned and develop effective and uniform pricing and discount strategies based on different types of customers to avoid profit leakages (e.g., if a small online customer gets higher discounts compared to large distributors).

How do you take action on this?

  • Determine which discounts or rebates have to be adjusted or removed
  • Develop a uniform pricing strategy for each group of customers
  • Prioritize investing in high potential customers

 

Customer Segmentation

 

The main question that any business asks is: "How can I reap maximum profit from my customers?" It is here that customer segmentation models are vital, as they allow us to examine which segments are the most and least profitable. Segmentation is the process of dividing customers into groups based on common attributes. Based on this, the marketing team can better target high-value customers and drive higher customer segmentation revenue.

Example of Customer Segmentation:

Descriptive analytics allows us to create segments in many different ways, from simple customer groupings (e.g., based on descriptive attributes such as customer location, age, or spending habits) to more advanced statistical analysis (e.g., k-mean clustering or neural networks).

The airline industry is the best example of this pricing policy because it accounts for individual purchase behaviour and offers prices based on multiple criteria like time of purchase, location, and volume.

How do you take action on this?

  • Identify what customer groups are the most profitable
  • Create effective marketing communications
  • Establish ways to improve customer service and customer retention rates

 

Predictive Analytics

 

Predictive analytic tools attempt to make predictions about future financial outcomes based on historical data. Generally speaking, there are several core techniques for forecasting future performance, such as decision trees or regression analysis.

Pros of Predictive Analytics

  • Improve the efficiency of business processes
  • Identify revenue opportunities

Cons of Predictive Analytics

  • Accuracy heavily depends on the input parameters
  • No visibility on a course of action

 

Decision Trees

 

What are Decision Trees?

Decision trees are among the most common methods of visualizing a chain of possible outcomes that companies use to quantify the effect of different pricing actions based on statistical probability. The main advantage of decision trees in machine learning is that you can see relationships among multiple activities simultaneously.

Example

In the airline industry, decision trees can play an essential role in setting the right ticket price. Discounting tickets too early can lead to profit losses. On the other hand, there is always the possibility that tickets will not sell before a scheduled flight resulting in significantly higher losses. By drawing a decision tree, an airline can estimate the expected revenue based on the probability of selling tickets.

How do you take action on this?

  • Identify the potential risks and who is bearing them
  • Estimate the impact of promotions and discounts

 

Regression Analysis

 

What is Regression Analysis?

Regression analysis is used to determine the relationship between a dependent variable (the parameter you are trying to predict or estimate) and independent variables (the parameters that impact the dependent variable). Regression models may include a wide range of independent variables to increase the accuracy of the simulation results like competitor data, product differentiators, or macroeconomic factors.

Example

Regression analysis can be a powerful forecasting tool to predict sales volumes using price changes and other factors like competitor prices or seasonal trends. Businesses need to define the market response to their pricing strategies by measuring price elasticity, which is one of the essential concepts in the price analysis. Price elasticity shows whether raising or lowering prices will allow you to achieve your pricing objectives. Applying a regression model helps to measure sensitivity to change of the quantity demanded of a good relative to its price change. The slope gives you an idea of how much volume your business would lose for every dollar increase.

How do you take action on this?

  • Define the effect of changes in price on sales
  • Improve forecasting and planning models

Predictive models go far beyond calculating price elasticity and can be implemented to forecast market share, revenue, and inventory requirements.

 

Prescriptive Analytics

 

Prescriptive analytics help determine which aspects of your business can be improved now or in the future. Prescriptive analytics can help you find the best course of action, by not merely forecasting future sales or profits, but by providing recommendations on improving current or future performance. It begins with collecting data from different descriptive and predictive sources for its models and applies them to quantify the effects of future decisions.

 Pros of Prescriptive Analytics

  •   Machine learning algorithms reduce the possibility of human error
  •   Significantly optimize business performance

Cons of Prescriptive Analytics

  • Require large amounts of data and resources
  • Expensive to implement

For example, prescriptive analytics is widely used to predict crude oil prices, which change almost every second and are affected by multiple factors, from political changes to demand parameters. Essentially, prescriptive analytics is a combination of neural networks, business rules, and statistical analysis. Neural networks represent the most sophisticated part of prescriptive analytics and consist of multi-layer networks of neurons used to classify data and make predictions. These types of models can find patterns in historical data and provide optimal solutions without human intervention.

The implementation of this approach requires significant effort to define the objective of the analysis and to design a pricing engine that can collect data from IT, sales, and finance departments and provide business insights in real-time. Eventually, this effort will pay off, boosting sales performance and optimizing business processes.

 

Summary

 

For a company to be successful, more business decisions need to be driven by advanced pricing analytics. This includes using automation tools to track the company's current and past financial performance, predictive models to build robust forecasts, and segmentation analysis to see how customers value your offerings. Companies have an abundance of data at their disposal, learning to leverage this data with advanced analytics will drive higher returns and ultimately benefit your success.

ABOUT THE AUTHOR  Evgeny Kubashevskiy is a Senior Consultant at Revenue Management Labs. Revenue Management Labs help companies develop and execute practical solutions to maximize long-term revenue and profitability. Connect with Evgeny at [email protected]