Linear performance pricing

from Wikipedia, the free encyclopedia

Linear Performance Pricing (LPP) or Linear Price Performance Measurement (LPPM) or linear price analysis or price-performance analysis or performance pricing is a simplifying aid based on regression analysis that is used in purchasing and development to identify different products or Compare services from different suppliers in terms of price and performance. For this purpose, the products are mapped as coordinate points in a linear price representation, the so-called value graph.

Basic process and structure of an analysis

An analysis requires properties and prices of products as input data. Each product is specified / described through properties such as weight, service life, etc. - defined as value drivers in VDI guideline 2817 - and the performance / value (hence the word "performance") is defined.

The price for each product is also given. This is necessary because the method wants to find a relationship between the product properties and the price.

On this basis, a formula is calculated with the help of regression analysis , which depicts the relationship between the properties and the price. Various so-called "estimators" can be used. One of the best known is the least squares estimator .

In order to determine whether a given price corresponds to the model, the known product properties are taken, inserted into the formula and the so-called technical value is calculated. The difference between the real price and the technical value gives an indication of possible price distortions.

Basic structure of a value graph

The price is plotted on the ordinate of a linear price representation and the performance of the products to be compared is plotted on the abscissa. A regression line is drawn through the point cloud thus created in the representation , which depicts the assumed linear relationship between price and performance.

Quantifying performance

While the price is already available as a numerical variable, the performance must still be quantified for the illustration. To do this, you choose one or more typical, measurable properties of the products that determine their performance from the buyer's point of view. For example, when comparing engines, values ​​such as mechanical power, torque or fuel consumption could be included in the calculation of the COP.

However, when calculating a performance figure, the problem arises that the individual characteristics are combined. However, it is unknown how best to do this. The use of a performance figure therefore leads to an unknown error size in the result. The more clever approach is to forego a performance figure and several dimensions (= characteristics of products ). analyze at the same time. This ensures that the information content of each feature is retained and incorporated into the calculated result.

Derivation from a value graph

If a product in the value graph is above the regression line, with a suitable quantification of the performance it can be concluded that this product is too expensive compared to the other products. By marking the points in the value graph, you can also see whether certain suppliers or required product properties are too expensive in comparison.

Use in practice

The linear price analysis was first introduced into practice in 1997 by the management consultancy McKinsey . Today it is a tool for a quick price comparison that can give an initial overview of prices.

The advantage of the linear price analysis is that different products with different performance characteristics can be compared in price with one another in a very simple way, whereby the quantification of the dimension "performance" can be completely defined from the perspective of the buyer. Linear price analysis is often used for price negotiations with suppliers and for comparing purchase prices within the company.

The quantification of the performance component is disadvantageous. It is aggregated to a single value, which can lead to distorting simplifications, especially with complex products. Products whose prices deviate from the linear price line can therefore have hidden non-linear performance features, the non-consideration of which makes the prices appear too high or too low. In practice, for example, the marginal price increases depending on the performance. There are several ways to counter this.

Non-linear performance pricing

In order to better represent economic effects (e.g. economies of scale and marginal utility ), non-linear models are used. Non-linear analyzes can better recognize these effects and depict them in a model.

The information as to whether a characteristic correlates linearly or non-linearly with the price is already available in the data and must be recognized. It is therefore necessary that performance pricing solutions can automatically recognize from the data what kind of correlation exists. This is possible by evaluating various correlation models. However, this also assumes an accuracy of the given price that is hardly true in reality. Furthermore, a higher non-linear approximation and lower deviations from the price are not associated with a higher gain in knowledge, since the deviations identify potential for price negotiations with suppliers.

Example of a linear price analysis

Example of a linear price analysis

In the illustration on the right, 2 products from supplier A and 3 products from supplier B are compared. After the performance has been quantified and the products have been entered in the linear price display, a regression line is drawn through the point cloud. In the example it becomes clear that product 2 from supplier A is too expensive, namely by the amount that the price of this product in the Y-direction is away from the regression line.

literature