Research: Business strategies adopted across darknet marketplaces
Over the past few years, many researchers have attempted to explore darknet marketplaces to analyze this novel form of illegal e-commerce. From an economic point of view, a darknet marketplace, or a crypto market, can be defined as a virtual tripartite market comprised of anonymous vendors, merchandise, and customers. In this context, a crypto market represents a good example of an economic model of totally rational and anonymous participants with a limited number of contacts and established relationships based solely on economic exchange. With such a unique framework, it is logical to question if it is possible to develop an economy of participants who do not directly trust each other. Furthermore, what are the reasons that make some marketplace vendors develop a competitive advantage despite their anonymity?
A recently published master’s thesis attempted to answer these questions via an innovative methodology that combines economic complexity theories with the literature of strategic management. Barney’s Sustained Competitive Advantage Model entails that different vendors within the same market can exhibit different levels of performance which are fueled by the unique resources of each vendor. The thesis illustrates that these ideas, with respect to vendors, are compliant with the Theory of Capabilities proposed by Hidalgo and Haussman, which postulates that the group of different products of various vendors represent the main source of competitiveness. Accordingly, by restructuring Barney’s model into the Hidalgo and Haussman Theory of Capabilities, authors of the paper present an analysis of the studied darknet dataset.
Starting from the aforementioned arguments, in order to present a quantitative analysis, the researchers formulated a generalization model of the Fitness-Complexity Algorithm developed by Pietronero and colleagues in 2017. Mainly, a new metric was proposed over a tripartite network of vendors, customers, and products, which focuses on the ability to diversify over an extended group of customers as the main source of vendors’ competitiveness. Results illustrate a prominent relationship between vendors’ fitness and their generated revenues, which supports the significance of this new parameter.
This thesis made two important contributions. Firstly, it has successfully analyzed vendors’ dynamics in crypto markets. Secondly, it has developed a novel approach that increases the applicability of theories of economic complexity to an individual level and, potentially, could successfully illustrate this fitter-gets-richer model across other tripartite competitive ecosystems.
Lessons learned from studying the business model of Silk Road:
Three different phenomena could be identified via analyzing Silk Road’s business model:
1. Vendors generating low income will work on boosting their income in an attempt to reach the market’s average income. This phenomenon can be explained by the fact that customers were always willing to buy from new vendors due to the market’s high levels of trustworthiness.
2. Vendors with higher income levels and a lower fitness were always willing to boost their fitness. This can be explained by the following:
– Vendors tend to reinvest their capital in their activity on the marketplace, boosting their products’ diversification and, hence, their fitness.
– Customers prefer vendors with high feedback scores and those who have been already praised on the forum – namely, those who have completed a substantial number of successful transactions and, hence, have high income. This customers’ behavior triggers new transactions within new product categories. In other words, customers are capable of increasing the diversification of vendors and, thus, their fitness levels via purchasing new products from their inventories.
3. Vendors with incomes considerably higher than the average income are considered overextended with respect to the market’s liquidity and, hence, tend to lower their incomes.
In the context of the global economy, if a country is capable of exporting a rare product, such as spaceships, the fitness algorithm will entail that this country is capable of producing almost everything else, and, consequently, is one of the fittest. In the global economy model, this is true. Nevertheless, in the context of a crypto market, if a vendor is capable of selling the rarest product, the algorithm entails that they can also provide less complex ones, but in reality, this does not actually happen.
It is interesting how just a small percentage of vendors could really diversify their product listings, thus, increasing their fitness levels, but this behavior it is not reflected by an increase in their revenues. In the darknet marketplace model, diversification doesn’t have an immediate positive influence on the market, opposite to the global economy model. An explanation may be provided via examining the different demands of the two market models; in the global economy model, countries end up exporting products only if there is a sustained global demand, while in crypto markets, most of the offered products have to cope with customers’ adoption, hence, they don’t affect vendors’ revenues. How can the fact that incomes of vendors are closely related to few products be correlated with the fact that most vendors are able to generate almost the same amount of money, despite their fitness levels?
An interpretation can be that, due to competition, there is indeed an equilibrium between optimal fitness and sales. It may be that the similarity with the countries’ economy model should not necessarily apply in the case of crypto markets, as the space of solutions and the potential for creativity and growth are very different from those of the countries’ economy model. Actually, in the case of crypto markets, the only products that can considerably move the market are the ubiquitous ones which are offered by almost all vendors. Moreover, it should be emphasized that the common average rating between vendors and the large trustworthiness of customers in the escrow system, render the customers willing to buy those few products from whoever is capable of selling them, increasing the revenues of all vendors, despite their fitness levels.
It can be clearly observed that the most important products are usually those categorized in a broad manner, this evident error could explain why the generated incomes are so disproportionally distributed over the products’ category. It is not easy to distinguish the typology of the most appreciated products, such as “Weed, Books or Pills,” while on the other hand, it is easily distinguishable, with an extreme degree of specificity, that some psychedelic products or some forms of edible magic mushrooms represent a rarer and consequentially a fitter product category. It is comparable to if you are in a mall and trying to make a comparison between the ubiquity of the shoes’ category and the ubiquity of the watches’ category.
Vendors’ business strategies on Agora’s darknet marketplace:
The Agora crypto market offers the chance of analyzing a network that doesn’t suffer the evolutional features of Silk Road. The number of vendors and products is almost identical in each time window, between 700 to 900 vendors, with an average number of offered products between 63 and 77. Moreover, the Agora dataset doesn’t exhibit the sloppy categorization of products on the Silk Road. The number of products is actually lower, and they all share the same degree of specificity. Opposite of what one might presume, the results are relatively similar, denoting that the fitness algorithm has yielded a general feature due to the anonymous nature of interactions between participants in crypto markets.
The evolution of vendors’ fitness rank indicates how, similarly to the Silk Road case, vendors that have generated bigger revenues are not those who have had the best fitness rank. It is evident how, even in this more stable market ecosystem, the push-down regime is even more powerful, denoting how hard it is for vendors to keep increasing their revenues over the average value, as well as how ineffective it is to maintain an over diversified product inventory in order to be competitive.
What does determine the success of vendors in Darknet marketplaces?
In this master’s thesis, it has been shown and justified how to cast this question into the scientific grounding of a new economic model, or an economic complexity approach, that yielded an opportunity to compare monetary data such as generated incomes with the competitive attributes of each vendor. A rich strategy literature indicates that vendor fitness is catalyzed by organization-specific features to reliably offer specific groups of products, which suggests a potential link with the Hildago & Haussman Model of Capabilities. However, the group of products offered by a vendor are typically conceptualized within the tradition of Hidalgo & Haussmann as being indicative reflective of their innate capabilities. Surprisingly enough, the study found no evidence for this model in crypto markets. This represents an essential reason why the thesis’ analysis of the bipartite network failed.
Both the tripartite generalization of the fitness complexity algorithm and the parallel analysis of vendors’ diffusion over the client space as well as the product space helped authors of the paper understand what determines vendors’ competitiveness and what sets it apart from country’s competitiveness in the global economy model. This yielded two main findings:
– The first explanation focuses not merely on capabilities to produce specific products, but on dynamic capabilities to update and adjust the inventories of offered products. No such evidence was detected that the fittest vendors continuously updated their product inventories, while it was proven that the fittest and richest vendors all adopted the same focus strategy, as they focused on a limited number of products (around 3-5) while tending to transact with a large number of clients.
– Moreover, it was shown that these vendors gradually moved from the client space’s periphery to its core. In other words, the most successful vendors were capable of navigating profitably throughout the client space.
This thesis analyzes some of the most common business strategies that could be observed across darknet marketplaces. The study attempted to illustrate a comparison between the global economy model and the darknet marketplace model. It has been proven that the most successful, i.e. fittest, vendors tend to sell a limited number of products, yet they tend to try to transact with a large number of clients. On the other hand, customers are always willing to purchase from new vendors, thanks to the trustworthiness offered by crypto markets’ escrow and vendors’ feedback systems.