Navigating the Pricing Conundrum - Azcoitia, Iordanou, and Laoutaris (2021)

Comment: Recommended. An independent analysis of pricing schema for data products and services.

(2021) S. A. Azcoitia, C. Iordanou, and N. Laoutaris, "Measuring the Price of Data in Commercial Data Marketplaces," in DE '22: Proceedings of the 1st International Workshop on Data Economy, December 2022, 1–7, https://doi.org/10.1145/3565011.3569053 (Accessed Q12024).

The work became a pre-print for: S. A. Azcoitia, C. Iordanou and N. Laoutaris, "Understanding the Price of Data in Commercial Data Marketplaces," 2023 IEEE 39th International Conference on Data Engineering (ICDE), Anaheim, CA, USA, 2023, pp. 3718-3728, doi: 10.1109/ICDE55515.2023.00300.

The Azcoitia, Iordanou, and Laoutaris study (2023) reported earlier is a detailed exploration into the complexities of figuring out how much data is worth in the ever-changing world of commercial data marketplaces. The research delves into the journey of data from creation to trade. This study examines the pricing structures of data products, services and marketplaces.

• Part (1) introduces the data marketplace ("DM").
• Part (2) & (3) reports the mechanisms of pricing of data products.
• Part (4) deep dives into the marketplace for AWS data.
• Part (5) compares data products across marketplaces.
• Part (6) dives into the features driving pricing.
• Part (7) & (8) presents related works, conclusions and future work plans.

Part (1) introduces the problem of developing "Data-driven decision making powered by ML algorithms;" in this study using two data marketplaces to develop a transfer pricing study mechanism.

Part (2) reports the mechanisms of pricing of 10,772 data products, ranging from one-off purchases to telecoms, manufacturing, and gaming data. The researchers interestingly note that a majority of data products are pricing from "direct negotiation between the seller and interested buyers." 

It is very much the picture of an age-old story. Bartering to determine value on the spot.

Part (3) & (4) reports the details of the market for data products and services; using the AWS ecosystem to explore a marketplace. Part 2.1 dives into the current vendor trifecta: Data Providers, Data marketplaces, and Personal Information Management Systems. The section discusses the entities uncovered by the team; and drills down into the characteristics of a sample of each vendor class. The section charts market share of each class. The geographic spread demonstrates US dominance of available data. Of these, "4,162 products from 443 distinct providers provided clear information about their prices" which led to the assessment that the median price is US$1,417 per month. Pricing ranges from free to $500,000; with "one-third of all data products, including targeted market data and reports for example,...sold for US$2,000-5,000 per month."

Part (5) & (6) continues the exploration of data marketplaces; by developing a pricing concordance and methodology "to build different classifiers to help us compare data products between the two DMs including more price references, namely DataRade (destination DM) and AWS (source DM)." 

The authors used the classification schema to compare pricing distribution of two categories-‘Financial’ and ‘Retail, Location and Marketing’ data products; from this work concluding that "it is mostly ‘what´, as captured in product description and categories, and ‘how much´ data is being traded that determine the prices of data products." 

Therefore, it is data quality evaluation and price--the value determined by the user determining what they want--that determines the value--therefore price--of data products.

Part (7) & (8) concludes with the note that this analysis "is, to the best of our knowledge, the first empirical measurement study that deals with the prices of data products sold in commercial data marketplaces." Further, that "the lack of empirical data around dataset prices is considered as a key challenge in data pricing research."

References section: Includes discussion of the methodology created to construct the analysis. 

Notes and analysis re-written with the assistance of a paid ChatGPT account. Image: Pxhere. Public Domain.

Labels: Costs;Biological system modeling;Ecosystems;Pricing;Data engineering;Data models;Telecommunications;Data economy;data marketplaces;measurement;data pricing.