In this article, Evangelo Zervoulias, Senior Pricing Specialist of TX Markets' central strategy team shares some insight into how pricing segmentation works and how it has developed and adapted over time.
Efficiently charging different prices to different customers has been known as Pricing Segmentation for a long time. There are different degrees, to which this segmentation can be done. Third-degree Price Segmentation is the mildest form and includes retailers offering the same service or product at a discount to a predetermined customer (think student or elderly discounts). With the boom of the internet and consumers having more agency over where to buy from and finding the right price, a second-degree segmentation was introduced. Since customers were able to efficiently and quickly compare prices of different suppliers online, suppliers responded by offering different prices based on quantity or quality of the product and price and demand. However, the attributes that determined the final price were mainly lying on the side of the supplier. This is also something that we refer to as dynamic pricing. An excellent example of this is airline travel, where prices largely depend on demand, seasonality and the amount of people. Another example would also be the plain bulk discounts at the supermarket. However, there is one more degree of segmentation that has been able to develop thanks to the abundance of data at hand and the sophistication of automation online: Personalised (or algorithmic) Pricing.
Identifying Attribute Groups
1st-degree pricing segmentation provides a good level of accuracy and consistency for the sellers. In order for this segmentation to work, one would need a set of attributes to determine the final price that is shown to a potential customer.
- Product/Service: all of the attributes linked to what is actually being sold
- Customer: attributes linked to whom the product/service is being sold to
- Transaction: attributes linked to the (history of) transactions
Update intervals for Dynamic Pricing (2nd degree) are pre-set; Algorithmic Pricing would allow for real-time price adjustments taking into account a myriad of digitized data variables. However, very few laws and barriers exist to curb the extent to which this practice could be used, which is why companies are hesitant to up the degree to which segmentation would be technically possible. The goal of companies that work with this type of pricing - and any company really - would be to capture the Willingness to Pay (WTP) for as many segments as possible.
Changing Decision Making
Pricing segmentation seems to be accepted by the general public in a variety of areas, but not in others. A great example - again - is airline travel; it is generally known that not everyone has paid the same price who's sitting on a plane and this is just what we are used to. However, it would be frowned upon if you went to a bookstore and somebody paid less for the exact same book. In some areas, the increased amount of specification of private choice changes decision making. And the customer needs to be aware of this, as not to lose his or her trust in the company. This is why a proactive approach to disclosing pricing is key.
Personalised Pricing on Marketplaces
For marketplaces and classifieds, personalised pricing works the same but also differently. The actual 'traded' products that are offered on marketplaces are listed by (private) people who want to sell their goods. They have control over the price and don't have the amount of data or infrastructure at their hands to even consider a dynamic pricing approach. They can however benefit from -for instance- visibility features for their listings. The prices of those visibility features might vary based on the demand of the object, location of the object and surely vary based on how much the feature boosts visibility and the eventual success of the listing. However, the price of the actual product will remain unaltered since it is somebody else deciding on the final price. Thus, pricing decisions for those companies are restricted to 'middleman' type of offerings such as listing fee, commission and visibility features, that need to be accurately placed and, when possible, personalised.
Ultimately, nothing can compensate for building trust and reliability with consumers, which is proven to reduce privacy concerns and increase acceptance. This requires transparency from the company and thorough communication. Since further degrees of pricing segmentation are yet to be explored, businesses will probably steer towards a new pricing future and will make sure to proactively inform and react to customer's concerns and needs.
Thank you to Evangelo for providing us with information on this topic.