Price differentiation – that is, setting different prices for different customers – is typically positioned as one of the silver bullets in the arsenal of the pricing manager. In this blog, we will outline the justification for price differentiation. Moreover, we will show that there is a catch; if price differentiation is not done correctly, it will immediately hurt the bottom line. It turns out that price differentiation is like blowfish sushi – great if prepared correctly, but deadly otherwise – and most of the time.

Why differentiate prices?

Price differentiation is generally justified by some version of the following example. You sell a single product with unit costs of $50. Demand is linear with a maximum demand of 3,000 (at a price of $0), a price elasticity of -2 at a price of $100, and a maximum willingness-to-pay of $150.

If you sell to all your customers at a single price point (Graph A), then the profit optimum is at a price of $100 with a margin of 50% (on price elasticity and pricing see here). Then, your profit is 1,000*($100-$50) = $50,000.

Benefits and Perils of Price Differentiation - Graph A

Graph A

What if you don’t know your customers’ WTP perfectly – and you still differentiate prices?

To investigate this question, assume you still know the overall demand function shown in Graph B, but you have no knowledge of each individual customer’s willingness-to-pay. Specifically, the individual price you set for a customer and that customer’s willingness-to-pay are uncorrelated. If you then still fully differentiate prices between $50 and $150 as shown in Graph B, we find via simulation using buynomics’ Virtual Customers, that the total profit you make is only around $22,500.

Benefits and Perils of Price Differentiation - Graph B


Naïvely trying to aim for each customers’ individual willingness-to-pay will not make price differentiation profitable. To make it work, two ingredients are required:

  • A precise knowledge of each of your customer’s willingness-to-pay that goes beyond rough concepts used for value pricing such as buyer personas and lets you actually target each customer individually.

  • The right set of methods to derive optimal individual customer prices from their expected willingness-to-pay. For example, to avoid over-pricing on individual willingness-to-pay estimates, it makes sense to systematically charge less than you expect a customer to be willing to pay. Graph D shows how profits increase if you charge each customer $10 less than you expect each customer to be willing to pay (pink) versus the curve (blue) from Graph C. Another method to avoid overcharging individual customers is via self-selection – e.g., with a portfolio offer.

Benefits and Perils of Price Differentiation - Graph C

Benefits and Perils of Price Differentiation - Graph D

We showed that price differentiation is much more difficult than is generally asserted, and that it requires both a precise understanding of individual customers’ willingness-to-pay and a method to set an individual’s price based on that understanding.

Price differentiation is like blowfish sushi: It needs to be prepared by a proper chef with the right toolkit.


buynomics Virtual Customers technology allows you to estimate individual customers’ preferences from your data and our a priori model of customer behavior – and it identifies the right customer-specific prices based on identified customer preferences and your objectives.

buynomics improves customer-specific pricing

To determine the profit effects of price differentiation, we simulated a market with 3,000 potential customers using buynomics’ Virtual Customers, each with an actual individual willingness-to-pay (1) drawn from a uniform distribution between $0 and $150. Further, an estimate of the willingness-to-pay (2) was also drawn for each customer from a uniform distribution between $0 and $150 and a correlation between (1) and (2) ranging from 0 to 1. If (2) was above the cost of $50, then this was the price asked from the customer – and if (1) was above (2), then the customer bought the product. The results shown in Graph C are the average over 100 simulation runs, where total profit was summed over the 3,000 potential customers.

Note, that if you don’t know the total market demand well, then the difference between the single-price profit and the profit under price differentiation becomes even greater. This we go into these details in a follow-up.

Note: $10 is not the optimal deduction. We will outline how to determine the optimal deduction using Virtual Customers in a follow-up blog.

Want to learn more?

Get Your Demo