ArticlePDF Available

How can we reduce information asymmetries and enhance trust in ‘The Market for Lemons’?


Abstract and Figures

The used car market is characterized by information asymmetries and mistrust. Blockchain technology promises to resolve these problems using a system which stores data over the life cycle of a vehicle. However, while blockchain technology is strong in preserving the stored information, sense-making of this information is still essential to bring value to end consumers of the system. In this paper, we take an exploratory approach and create a prototype, which is then evaluated in realistic car sale conversations between buyers and sellers. We demonstrate and discuss how the interplay of different design elements of an application, built on top of a blockchain-based platform, helps to reduce information asymmetries and enhance trust. Our findings suggest that though providing more information about a used product (a car) leads to fewer information asymmetries in general, a reputation mechanism and data analysis are both beneficial in improving the situation further. As for trust, such a system enhances trust between buyers and sellers and, in general, makes the overall purchase process more trustworthy. However, to achieve these positive effects, the quality of the stored information should be guaranteed and properly communicated to the end-user.
Content may be subject to copyright.
The used car market is characterized by information asymmetries and mistrust. Blockchain technology promises
to resolve these problems using a system which stores data over the life cycle of a vehicle. However, while
blockchain technology is strong in preserving the stored information, sense-making of this information is still
essential to bring value to end consumers of the system. In this paper, we take an exploratory approach and create
a prototype, which is then evaluated in realistic car sale conversations between buyers and sellers. We demonstrate
and discuss how the interplay of different design elements of an application, built on top of a blockchain-based
platform, helps to reduce information asymmetries and enhance trust. Our findings suggest that though providing
more information about a used product (a car) leads to fewer information asymmetries in general, a reputation
mechanism and data analysis are both beneficial in improving the situation further. As for trust, such a system
enhances trust between buyers and sellers and, in general, makes the overall purchase process more trustworthy.
However, to achieve these positive effects, the quality of the stored information should be guaranteed and properly
communicated to the end-user.
1 Introduction
Mistrust and uncertainty rule automobile markets all over the world. Notably, the market for used cars is deeply
affected by these issues. The lack of trust in this market has a long history: In the 1970s, Nobel Prize winner G. A.
Akerlof offered insight into why the used car market is characterized by massive distrust, in his noted work “The
Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism” (Akerlof 1970). He found that the main
reason lies in the fact that while the sellers have good knowledge about the quality of the car being sold, the buyers
are left in the dark - hence the uncertainty. Since sellers are aware of this effect, they will try and sell their car for
a price that is too high in relation to the quality of the specific car. This information asymmetry creates
opportunistic behavior, which results in mistrust between buyers and sellers and can even lead to a complete market
failure. These characteristics still exist in the modern used car market; it is still haunted by information
asymmetries, distrust and uncertainty.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
In a novel approach to tackle these issues, a Swiss blockchain-based electronic platform project named
“Cardossier” has commenced. The project was initiated by a several organizations from the car-related ecosystem
(an insurance company, a car sharing company, an importer and car dealer, and a road traffic agency), a software
company, and two universities
. This platform aims to store and exchange the complete information about the life
cycle of a car, from production to disposal. The idea is that by storing as much information as possible about the
car’s history, the information asymmetries can be reduced, as well as potentially ensuring the trust between buyers
and sellers. Instead of a conventional database for data storage, the participants of the project opted for usage of
permissioned blockchain system (participants know each other), which due to its characteristics (such as
immutability of transactions, transparency, integrity of data, and decentralized character (Seebacher and Schüritz
2017)) claims to bring more trust in relationship of transacting parties (Fleischmann and Ivens 2019). The
participating organizations leverage the technology to benefit in three ways: Shared operational efficiency,
controlled customer intimacy, and joint product innovation (more on this aspect can be found in Bauer et al. 2019).
Why blockchain? This choice was determined by several reasons. First, the participants of the project (public and
private organizations from the car-related ecosystem) did not want to rely on one single provider (be it one from
the ecosystem, e.g. an insurance company, or an external like Amazon or Google) to run the system for them and,
thus, give away the steering powers. Second, a distributed platform allows to comply to strict EU data privacy
regulations. In this study, we propose a design for an interface application to the cardossier platform to reduce
information asymmetries and improve trust between used car buyers and sellers.
However, just storing the information life cycle of the car is not enough. There must be business logic applied in
order to present the information in such a way that involved parties can make educated decisions about the quality
of a car, or whether a party is trustworthy or not. To solve the problem of information asymmetry, sellers - and
especially buyers - have to be informed about the actual quality and condition of the car. Though underlying
blockchain technology is strong in preserving the stored information (by immutability of stored records or data),
sense-making of this information is still required to bring value to the end consumers in the system. For this,
research must be conducted to establish the kind of information buyers and sellers are looking for and how it
In March 2019, a non-profit association was founded to prepare market entry and expand the ecosystem. Ten new
organizations joined the project by July 2019.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
should be presented to them to solve the problems of existing information asymmetries and mistrust in the used
car market. Thus, we state the research questions as follows:
RQ1: How can the provision of car usage data in a blockchain platform reduce information asymmetries and
enhance trust in the used car market?
RQ2: How can further relational data be included to reduce information asymmetries and enhance trust in the used
car market?
Consequently, the two main research objectives are to propose a design for reduction of the information
asymmetries between buyers and sellers of used cars as well as for enhancement of their trust to each other by
provision of car usage information as well as additional relational data. We recognize that this study takes an
exploratory approach and aims to propose initial design ideas and explore their potential, rather than proposing a
mature design theory. Furthermore, though there are other perspectives that play an important role in the design
of such a platform (for example, its governance (Ziolkowski et al. 2019), business value for participating
companies (Bauer, Zavolokina, Leisibach, et al. 2019), data privacy (Zavolokina et al. 2019)), in this study we
focus on the perspective of an end user, who interacts with the frontend application and does not deal with the
blockchain-based backend.
On the example of the used car market, this study contributes confirmation of positive influence of blockchain
application on reduction of information asymmetries and trust enhancement in the used car market, design ideas
and their evaluation to the question of how information asymmetries between buyer and seller can be decreased,
and their trust enhanced. For researchers, this study provides a foundation for further research on the role of trust
in IT artifacts in the context of used car market, reputation mechanisms and the design of applications on top of
blockchain technology. For practitioners, this work offers a practical view on the implementation of design
components suggested in this study. Insights concerning the importance of data quality over data quantity are
given, which is something a blockchain-based car data platform should incorporate.
In this paper, we first introduce the research design and then describe the activities of design science research
methodology (DSRM) by Peffers et al. (2007), performed in this study. These activities inspire the structure of
this paper. In the section Literature Review, we initially identify problems & objectives. Then, we define the
solution. Afterwards, the design & development of the IT artifact is presented, followed by demonstration &
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
evaluation of its use in an experimental setting. We then discuss our results and draw conclusions. The paper closes
by outlining limitations of our study.
2 Research Design
In this study, we follow a Design Science Research (DSR) approach, which aims to develop new knowledge about
artifacts that contribute to the body of academic knowledge or prove to be useful in practical applications (Gregor
and Hevner 2013; Hevner et al. 2004; Nunamaker et al. 2015). We create an IT artifact for a proof-of-concept
research stage. One of the key goals of proof-of-concept stage is ‘to demonstrate the functional feasibility for a
potential solution to an important class of unsolved problems in the field. (Nunamaker et al. 2015). Proof-of-
concept prototypes are often simplistic but have enough functionality for testing their functional feasibility
(Nunamaker et al. 2015). This goal inspires our research. Proof-of-concept prototypes have higher risk of failure,
but are useful for laying a foundation for next stages in research (i.e., proof-of-value followed by proof-of-use)
and quicker provision of deeper and richer insights (Nunamaker et al. 2015). The created IT artifact in our study
is an instantiation (March and Smith 1995), based on research on ‘The Market for Lemons’, trust and reputation
mechanism. As a research paradigm, DSR is widely accepted among information systems researchers and can be
used to derive generalizable knowledge from specific problems and solutions. The research approach is based on
the methodology of Peffers et al. (2007). Peffers et al. (2007) established the DSRM, a robust framework for
conducting research in the area of Design Science in Information Systems. The methodology offers principles,
practices and procedures which are helpful in carrying out such research. DSRM was chosen as the main
methodology of this work as it is a well-grounded framework for successfully conducting DSR (Peffers et al.
2007). DSRM consists of six activities in a nominal sequence (Peffers et al. 2007): (1) problem identification and
motivation, (2) definition of the objectives for a solution, (3) design and development, (4) demonstration, (5)
evaluation and (6) communication. These activities are used to structure this study. In this work, a problem-
centered approach is taken because the lack of trust and uncertainty plague the used car market. Thus, there is a
need for an efficient solution that solves these issues.
The framework was applied in this work as follows: Starting with a problem-centered approach, activities 1 to 5
were executed. DSRM is an iterative process and may result in several iterations between activities 2 and 6. In this
study, we made a second iteration through activities 3 to 5, integrating outcomes of the design and demonstration
from the first iteration. All these activities are described in more detail in the subsequent sections of the paper.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
3 Literature review
3.1 Identify Problems and Objectives
The specific research problem should be defined, and the value of a solution should be justified. To identify the
problems and the motivation for the study, a standard backward and forward search (Levy and Ellis 2006) of
literature about the used car market, information asymmetries, trust, reputation and blockchain was conducted.
Based on the results of the literature search, a precise problem statement was defined, and a solution to the problem
was both justified and clear in its motivation. In our study, we examine two closely connected problems that are
present in the lemonmarket: (1) The presence of information asymmetries, and (2) The missing trust between
buyers and sellers of used cars. Based on the problem definition, objectives for a solution should be inferred
(Peffers et al. 2007).
Most buyer-seller relationships, such as those in the used car market, are characterized by information asymmetry
(Ba and Pavlou 2002). This is based on the fact that the seller usually has more information than the buyer about
the quality of the product or the service (Ba and Pavlou 2002; Mishra et al. 1998). Only after owning a specific
car for a while can the buyer form a good idea of its actual quality, which is more accurate than the original estimate
(Akerlof 1970). Furthermore, the information asymmetry may give rise to opportunism (Akerlof 1970). Ba and
Pavlou (2002) define opportunistic behavior as “self-interest seeking with guile”. In other words, the informed
seller tries to cheat the uninformed buyer in the hope of securing a higher sale price. This can happen through
misrepresentation of the true condition of a car, incomplete disclosure of information, actual quality cheating,
contract default, failure to acknowledge warranties, or other ways (Ba and Pavlou 2002; Mishra et al. 1998).
Akerlof (1970) explains that this opportunism can potentially lead to a full market failure: As the buyer cannot
differentiate “good cars” from “bad cars”, most traded cars will be lemons’, as these have the potential to earn
sellers the greatest profit. Consequently, the “bad cars” tend to drive out the “good cars”, as they will sell at the
same price.
If both uncertainty and incomplete product information (information asymmetry) are present in a transaction, trust
becomes critical (Ba and Pavlou 2002; Swan and Nolan 1985). In the used car market, both of these factors are
present. The buyer does not have full information about the quality of a car and often does not know the seller,
which creates uncertainty. Additionally, the buyer is aware of these relationships, which makes him wary of the
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
seller. Finally, both the buyer and the seller, have a motivation to complete the sale with a low and respectively
high price, which creates an underlying major conflict of interest. Together, all these factors and effects lead to
significant distrust between the two parties.
Research has shown that creating a design for trust is not a trivial task and involves thorough understanding of
trust concepts. Even though the concept of trust is easy to recognize, as we experience it daily, it manifests itself
in various forms which makes it difficult to describe (Jøsang et al. 2007; McKnight and Chervany 1996). The term
trustcan be defined as the willingness of a trustor to be vulnerable to the actions of a trustee based on the
expectation that the trustee will perform a particular action important to the trustor, irrespective of the ability to
monitor or control the trustee” (Mayer et al. 1995; Söllner and Leimeister 2013). Per the definition, trust is a
relationship between a trustor and a trustee. The trustor chooses to make himself vulnerable by trusting the actions
of a trustee. They cannot control what actions the trustee will take: The trustee can either honor the trust and act
in the trustor’s interests, or they can abandon the trust and abuse it. While the trustor is generally understood to be
a human, different categories of trustee exist: human beings, organizations, institutions and IT artifacts (Söllner
and Leimeister 2013). IT artifacts usually take the mediator role (supporting the relationship between trustor and
trustee), but they can also take the role of a trustee. In this study, we look how an IT artifact may be designed to
influence the trust of humans in other humans (IT artifact in the role of a mediator in trust relationships between
buyers and sellers). For this, an understanding of the formation of trust is crucial. Most researchers adopted the
model of Mayer et al. (1995) and its three dimensions of trustworthiness. In the model, trust between humans is
assessed based on the trustee’s ability, benevolence and integrity (Söllner and Leimeister 2013). Ability refers to
the trustor’s perception that the trustee has the skills, competencies and characteristics that enable him to have
influence in a specific domain. Benevolence reflects the trustor’s perception that the trustee wants to “do good”
for the trustor and does not follow an egocentric profit move. Integrity reflects the trustor’s perception that the
trustee has a set of principles and rules that they adhere to, which will be acceptable to the trustor. We adopted
these constituents of trust when constructing the surveys used to evaluate the designed IT artifact in the role of a
mediator for the inter-personal trust relationship.
Since IT may help in the reduction of information asymmetries and the improvement of trust, our initial objective
is to design a system which positively influence these aspects. In order to formalize the designed solution, a
rigorous literature review has been conducted. This literature review takes a closer look at the creation of market
transparency as a means of: reducing information asymmetries, enhancing trust, and designing such systems on
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
top of blockchain platforms. Furthermore, we examine how provision of relational data (product-related as well
as person-related) may be designed to attain the goals. Table 1 gives an overview of the objectives of the proposed
solution discussed above, and components of the proposed solution discussed below.
3.2 Define Solution
Increasing market transparency could help in reduction of information asymmetries (Rezabakhsh et al. 2006).
Market transparency can be defined as the level of availability and accessibility of information about products
and market prices” (Granados et al. 2008). Companies use IT to improve and automate their selling mechanisms
with informational features. Thus, IT increases the potential for complete, accurate, and unbiased market
information (Granados et al. 2006). In this context, market transparency refers to two types of information: product
information and price information; both are important drivers of a consumer’s purchase decision. Therefore, it’s
necessary to provide transparency over both types of information. Product transparency refers to disclosure of
product attributes and quality information, while price transparency refers to more and better information about
market prices and market dynamics. Therefore, both of these dimensions lead to a more transparent market for
consumers (Granados et al. 2006, 2008).
To reduce the information asymmetries, relevant information must be provided to both parties. The lack of trust,
on the other hand, is harder to mitigate, as it is rooted in a complex mechanism and is different for every party
(Söllner and Leimeister 2013). Reputation mechanisms provide one way to improve trust and such mechanisms
have been successfully applied in the past (Fuller et al. 2007). Thus, transparency over one’s reputation should be
provided to enhance trust. We use reputation mechanism to reflect person-related data. Using reputation, sellers
of low-quality cars achieve lower prices, which in turn results in a healthier market with a variety of quality and
prices (Resnick et al. 2000). Additionally, it allows for sellers with favorable reputations to achieve an
improvement in revenue, as buyers may be willing to pay extra for the security and comfort of their services
(Resnick et al. 2000). Reputation transferability (which allows for transfer of a reputation score from one platform
to another) between different platforms can further reduce uncertainty and develop the trust of customers
(Kokkodis and Ipeirotis 2015). In their extensive study, Fuller et al. (2007) investigated whether reputation, in the
form of external information, would positively affect a user’s trusting beliefs (which in turn results in trust). Fuller
et al. (2007) conducted a large experiment (N = 369) and examined how reputation information influences
consumer decision making in an e-vendor environment. They found that reputation information that focuses on
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
ability (called competence in their study), benevolence and integrity directly affects a buyer’s trusting beliefs
regarding the ability, benevolence and integrity of the seller, with significant results for each of the different
criteria. Therefore, in our study, we investigate if and how reputation mechanisms may foster trust between buyers
and sellers and make the sales process more trustworthy.
Previous research has shown that blockchain technology can be an appropriate basis for a solution to resolve
information asymmetries and trust issues in the used car market by creating a blockchain-based vehicle history
(Bauer, Zavolokina, Leisibach, et al. 2019; Brousmiche et al. 2018; Notheisen et al. 2017; Zavolokina et al. 2019).
The main benefits that blockchain technology brings are authentication of data and transactions stored in a
blockchain ledger (Miscione et al. 2018), and resulting out of it trust (Fleischmann and Ivens 2019). A blockchain-
based vehicle history is an alternative solution to traditional vehicle history reports, like ones offered by Carfax or
car manufacturers. Today, traditional vehicle history reports have several shortcomings which can be resolved by
blockchain technology. These shortcomings include lack of trust of end users in the single data source, strong
control over the data from one single data provider, low data quality (due to its incompleteness and incorrectness),
no proof of authenticity of the provided data, and finally high costs. In blockchain-based solutions, car-related data
is collected from and validated by various organizations involved in the life cycle of a car. Often, such solutions
are based on permissioned blockchains, where participants are known, which makes the overall system more
transparent and trustworthy. Storing data in a blockchain-based system makes this data immutable, data providers
transparent, and overall control over the system distributed between its participants. Blockchain technology cannot
be the holy grail to resolve all of the mentioned shortcomings. However, it provides a possibility to create and
access a more secure, reliable, and thus, more trusted data in a vehicle report due to the technology’s capabilities
of authenticating data (Miscione et al. 2018) and distributed data collection. The value of this data was discussed
in several studies. Simulating an online used car marketplace, we could demonstrate a positive impact of the trusted
car data both for buyers and sellers in the used car market from the market perspective (Bauer, Zavolokina, and
Schwabe 2019). Furthermore, business value for involved organizations was identified as such: organizations
benefit in the areas of shared operational efficiency, joint product innovation, and controlled customer intimacy
(Bauer, Zavolokina, Leisibach, et al. 2019). All these values are enabled by access to trusted, blockchain-based
car data. Furthermore, usage of blockchain technology for such an application enables to manage digital rights for
data ownership, storage, usage, access provision (Zavolokina et al. 2020). In the study at hand, we examine the
user perspective on a blockchain-based history report being used in a face-to-face sales situation.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
While most recent studies focused on the technical aspects of blockchain systems, only a few focused on the user’s
perspective and actual application design. Zavolokina et al. (2019) formulated generic requirements for such a
system to support used-car buyers experiencing uncertainty when purchasing a car and searching for information
about its quality. These requirements include: data analysis (the analysis of stored raw data to make sense of it),
provision of a timeline of a vehicle’s life cycle with event records, independent parties’/experts’ evaluation. We
use data analysis and evaluation from independent parties as the product-related data. However, the value of these
requirements has not yet been proven. In our study, we adopt these requirements, show how they may be
incorporated in the design of the system, and reflect on their usefulness in reducing information asymmetries and
their influence on trust relationships. We do not fully focus on the technical capabilities offered by the underlying
architecture of blockchain technology, but rather on an application enabled by the technology. Therefore, the
discussion of blockchain technology here, and further in the paper, is limited to core concepts.
Objectives of a solution
- Reduction of information asymmetries and trust
enhancement through provision of car usage data
- Basic car data
- Timeline of a car’s life cycle
(Zavolokina et al. 2019)
- Reduction of information asymmetries and trust
enhancement through provision of further relational
- Product-related data (independent
parties’/experts’ evaluation, data
analysis (Zavolokina et al. 2019))
- Person-related data (reputation
mechanism (Fuller et al. 2007))
Table 1 Objectives and the proposed solution
4 Design and Development
The designed and developed artifact is a prototype for the interface application, developed to utilize and
demonstrate data in the dossier about a specific car, available on the cardossier platform, to the end user (a car
buyer or a car seller). The artifact is a prototype that accompanies development activities in the overall cardossier
project. Figure 1 illustrates the architecture of the cardossier platform. Cardossier Core is a blockchain-based
storage for data exchange (in our case, based on Corda blockchain). The cardossier Dapp store allows to develop
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
decentralized applications (so-called Dapps), which utilize the stored car-related data to execute business logic
(for example, to create an insurance police, for fleet management, or other use cases). These Dapps, in their turn,
are then connected to external systems, such as web portals (for example, an online portal for used cars), web
applications, or core systems of the consortium partners. In this study, we focus (marked as purple on Figure 1) on
the design of the interface application (and its business logic provided by the Dapp) that encompasses the dossier
of a car to support the end users, used car buyers and sellers. We further refer to this prototype for the interface
application as cardossier.
Figure 1 Architecture of the cardossier platform
This section introduces the artifact’s desired functionality, its architecture and its actual creation. The design
process has undergone several iterations, and this chapter focuses on the final design prototype and explains its
various components. An important design decision was made at an early stage: The targeted devices for the
prototype were smartphones of all kinds. The reason for this was that the prototype should be able to support both
the buyer and the seller at the point of sale. A laptop or a desktop computer was not considered suitable for this
purpose, so the logical option was to design either a tablet or smartphone-based solution. As smartphones are
considerably more popular than tablets and also more practical they are small and do not rely on wi-fi
connectivity, the decision was made to design a smartphone-based solution. In order to increase accessibility, it
was decided that a web-based application would be developed.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
Further, we provide a description of the frontend component of the prototype. In total, we propose six design
components, which we then explore in relation to the objectives of the solution and the stated problems. We relate
design components 1 5 to the provision of market transparency (by targeting both product and price
transparency), enabling the reduction of information asymmetries. Design component 6, in turn, relates to the
provision of transparency over one’s reputation, leading to enhanced trust. We recognize that these components
should complement one another to maximize the benefits. We further describe the proposed design components in
a solution scenario. As illustrated in more detail below, there were two versions of the prototype which included
different design components. In the scenario, we describe all integrated design components to give a complete
overview (corresponding to the Complete Prototype). Screenshots of the six design components integrated into the
prototype (in Figures 2 – 7) provide visual support for the description. We describe only one scenario: when the
seller and the buyer meet, and the car purchase and examination take place offline. However, alternative scenarios
(e.g. the buyer examining the cardossier online before or after the meeting) are possible.
Protagonists: Max (Buyer), Cassandra (Seller)
Max, a 26-year-old Swiss male from Zurich, recently graduated from the university. He was then offered a job
at a Swiss software company located 100 kilometers outside of Zurich. To travel to his job on a daily basis,
he decided to buy a car which he planned to use for commuting. As his budget was small, he opted to buy a
used car. Max searched on an online platform for a 6-year-old VW Golf. All results showed prices and different
information about the cars, and result showed varying amounts of data. He found a car which was reasonably
priced and felt that the car would perfectly suit his needs. He used a form on the website to ask the car seller,
Cassandra, whether this information was true and asked to meet for a conversation about the purchase. A
couple of days later, they met in the city.
While meeting Cassandra, Max went to the website and checked the cardossier for the
Cassandra’s car by providing its VIN2. Max explored the cardossier:
(1) Basic information (see screenshots in Figure 2 and Figure 3) First, Max saw the basic information about
the car: The car’s key figures, which included its date of first registration, mileage, number of previous owners
A vehicle identification number (VIN) is a unique code, including a serial number, used by the automotive industry to identify individual motor vehicles, towed vehicles,
motorcycles, scooters and mopeds.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
and vehicle class. He carefully examined the mileage information and compared it to the information provided
by Cassandra. He knew that mileage is quite often manipulated. In the cardossier, he saw the graph showing
how the mileage status was developing over time compared to the average mileage for this specific car model.
The next section he scrolled to was titled “care”. This section gave more information about how the car was
cared for. This included warranty, smoking, car examination and service information. Next, was a safety
section that listed any accidents and recalls the car might have experienced.
(2) Car timeline (see screenshot on Figure 4) Directly afterwards, Max saw a timeline of the car’s life. He
was surprised that he could trace the entire history of the car back to its production. He could click through
the different years and the cardossier showed him the events for each selected year. The items in the timeline
contained information about events in the car’s life, such as accidents, mileage changes, car examinations,
services or owner changes. Each of the events consisted of an event title, a free text field giving detailed
information about the event, an event date, and information about the data source of this particular event. He
found out that the car had been involved in an accident, which Cassandra did not mention. Max asked
Cassandra questions about facts from the timeline and observed her reactions to his questions.
(3) Evaluation of car quality (see screenshot on Figure 5) Max had seen some facts in the data. However, he
was not sure if these facts necessarily meant the car was of good quality or bad quality. When he found the
section with reviews and ratings provided by an independent, third-party car expert and reviews by previous
potential buyers who inspected the car (but did not buy it), he was surprised by the availability of such
information about the car. A rating consisted of stars (ranging from 1 to 5, where 1 = worst and 5 = best; half
stars were possible), a title, a free text field and additionally positive and negative comments. The stars, title
and free text made up the general part, while the positive and negative items made up the specific part. He saw
that the expert provided a 5-star rating and that he confirmed and explained some facts (age, number of
(4) Articles On the next tab, Max saw some relevant articles (reviews by car magazines or tests by car experts)
as well as videos for the car’s model. These articles and videos were related to the general model, not the
specific car he was viewing. As Max didn’t have much expertise in the evaluation of cars, he found this section
helpful to see how cars of the same model performed.
(5) Cost prediction and life expectancy (see screenshots on Figure 6 and Figure 7) Max asked himself how
long he would be able to use the car, given its quality, and how much this usage would cost him. Max could
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
answer these questions looking at the cardossier: there was a usage section that showcased how the car was
used during its life. For this, the car’s life expectancy and its expected yearly costs were shown. These were
information components that were interactive (e.g., the costs widget would estimate his yearly costs if he drove
that specific car). Max could then increase or decrease the default yearly predicted mileage to see how it
affected the costs. In the life expectancy section, Max could look at the life expectancy of the VW Golf offered
by Cassandra, compared to average cars of the same model. Although the car had had a minor accident before,
he was happy to see that the predicted lifespan was longer that the average for this model.
(6) Seller’s personal data and reviews Max wondered why Cassandra hadn’t disclosed the information about
the accident she’d had. Did she intentionally keep it from him? Did she just forget to tell him? Or didn’t he
even ask her? He wasn’t certain he could trust her. In the cardossier, Max found information about the current
owner of the car (Cassandra). This information included the name, type (private or professional seller),
address, number of previous sales and average ratings of the owner. By scrolling further down, he found more
detailed ratings and reviews about her as a seller in the ratings section. He saw that these ratings came from a
related online auction platform, where Cassandra had already sold other smaller used items. Like earlier
sections in the cardossier, it had star ratings ranging from 1 to 5, a title, free text field and positive as well as
negative comments. Max was relieved to see that Cassandra had quite positive ratings and reviews.
Max felt reassured having been able to check the information on the cardossier website. He knew Cassandra
couldn’t influence the information, and he was able to make sense of it without relying on Cassandra’s words.
Therefore, he decided to buy the car.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
Figure 3 Basic information about
the car (continued)
Figure 4 Car timeline
Figure 6 Cost prediction
Figure 7 Life expectancy
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
5 Demonstration and Evaluation
5.1 Evaluation strategy
DSR suggests that the achievement of objectives and the identification and solving of problems should be
evaluated. Venable et al. (2012) discuss different purposes of artifact evaluation in DSR. Our study aims at
evaluating ‘an instantiation of a designed artifact to establish its utility and efficacy (or lack thereof) for achieving
its stated purpose’. More specifically, in our research, we aimed to reduce information asymmetries and to enhance
trust between buyers and sellers in the used car market through provision of (1) car usage data, and (2) further
relational data. For the evaluation of the artifact, we used the guidelines, proposed by Pries-Heje et al. (2008) and
Venable et al. (2012): We evaluated a design product (the cardossier) ex-post, in an artificial evaluation using a
realistic scenario of car sales conversation between a car buyer and a car seller. Venable et al. (2012) recommend
a lab experiment for this evaluation. As we tested a proof-of concept prototype (Nunamaker et al. 2015), rich
insights from a realistic setting are more important than statistical power ("Speed and exploratory rigor are
therefore useful to produce rich understandings of how and why people respond to various early approaches [....].
Quantitative experimental rigor is less useful in proof-of-concept research" (Nunamaker et al. 2015, p. 19)). Such
design research gains its power by concatenation, i.e. a series of studies that observe similar phenomena in varying
contexts. In our case, the reported study is preceded by a large scale survey (Zavolokina et al. 2019) that support
the identification of problems and requirements. And some of the insights presented here are supported by a later
study applying experimental techniques to a larger experimental sample (Bauer, Zavolokina, and Schwabe 2019).
This paper contributes the rich contextualized knowledge of a realistic experiment testing a novel artefact.
Therefore, we applied not only quantitative methods (to gain indicative numbers) but also relied on qualitative
methods (described in the next sub-section) that allowed us to gain a deeper understanding.
This type of an evaluation was chosen after careful considerations regarding evaluation goals and tradeoffs
between richness and rigor of obtained results. We took an exploratory approach to test early design ideas for the
novel cardossier application to solve the discussed problems in the used car market. Exploratory research is aimed
at describing a phenomenon of interest when there is little or no scientific knowledge about it. Given the novelty
of blockchain applications in general and their use in regards to trust and information asymmetries in the context
of the used car market, we believe our approach is appropriate.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
For the evaluation of the artifact, we used experimental techniques (Riege et al. 2009) to understand if we reached
the objectives of the study, how the design ideas contribute to this goal attainment, and what practitioners can learn
from the observations for the future design and development. Therefore, we did not strive for a large number of
subjects, but rather for a broad variety of different techniques, both qualitative (observations, taking notes, audio
recordings of sales conversations, interviews) and quantitative (survey). This approach helped us to achieve richer
insights, however, has its limitations in terms of generalizability of the results. In order to gain insight about how
the system may be used in the real life, the evaluation setting was made as close to the realistic situations of used-
car sales as possible. For this, buyers and sellers were hired to play the roles of buying and selling a car. The
evaluation compared a traditional sale scenario with a cardossier-supported one. Furthermore, as it is important
for us to differentiate between “raw datastored in the system and design components that provided aggregated
or analyzed information, we split the IT-supported conversations into two, as further described in the section
Evaluation method. To measure the goal attainment, we looked at how the perceived knowledge about the car (to
meet the first objective of reducing information asymmetries) and the perceived trust of buyers in sellers (to meet
the first objective of enhancing trust) change between settings.
5.2 Evaluation method
In our research, two within-subject experiments were conducted: one using a private sale scenario and one using
a professional sale scenario. Both experiments used a near-realistic setting that was kept as similar as possible. In
the experiments, the test persons were given scenarios and the task to either buy a car as cheaply as possible
(buyers) or sell the car as profitably as possible (sellers). For both experiments, a real car was arranged for sale.
This was done to make the experiments as realistic as possible, as the buyers could visually inspect the car and the
sellers could talk about features and the condition of the car. Apart from a test drive, the buyers and sellers could
interact with the car as they desired (a test drive was not possible due to time constraints). Two private sellers were
hired for a full day and sold cars to a total of 5 buyers each. The professional seller, who had 4 sales sessions, was
not paid for participation in the evaluation, as she works for a car dealer, which is a partner in the research project.
Thus, the evaluation was conducted with 14 buyers. Each of the sales sessions in both scenarios lasted for 90
minutes. Buyers were hired for 90 minutes each (45 minutes car sale, 15 minutes survey, 30 minutes interview).
Such an evaluation requires high effort, therefore, the researchers had to balance between keeping the evaluation
manageable and larger number of evaluation participants. Another study was conducted to investigate what overall
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
impact the cardossier has on the used car market from a macro-level perspective, where a larger number of
participants (50) were involved (Bauer, Zavolokina, and Schwabe 2019). The study at hand, however, focuses on
a micro-level perspective of the end users.
Each buyer was taken through three subsequent sale conversations with the sellers. In total, 52 sales conversations
were conducted. In the first sale, the buyer and seller were not assisted by any IT device. This sale conversation is
called “Traditional”. This sale was considered as a point of reference for examining effect of a blockchain-based
application on asymmetry of information and trust in general. In the second sale, both the buyer and seller were
assisted by the Lite prototype - the prototype that offered only basic information elements (basic information about
the car and a timeline of events), which might be stored in a system without additional elements like evaluations
and the seller’s reputation. This prototype was created to achieve the first objective of this study. The buyer and
seller were asked to start the application on their phones by navigating to a website. They could look at the
prototype for as long as they wanted (as if they were preparing for the meeting beforehand at home) and, once
ready, they could start the sale conversation. Finally, in the third sale, the buyer and seller were assisted by the
Complete Prototype, which included all design components. This prototype was created to achieve the second
objective of this study. Again, the buyer and seller were asked to start the prototype and read through it before the
sale. Table 2 summarizes which design components were included in the Lite and Complete prototypes. While
carrying out the sales in a random order would have been preferable, this was not feasible due to the fact that
information about car quality and reputation were “sticky”: Once a buyer or seller read them in the Complete
Prototype, he would not forget them, and it would affect the subsequent conversations. This important limitation
could be avoided by increasing the sample size (and the combined randomization) or switching to between-subject
design of the experimental setting.
Design components
in Lite
Included in
(1) Basic information about the car and its condition
(2) Car timeline: relevant car-related events (registration, insurance,
accident records, mileage records, etc.) and their sources
(3) Evaluation of car quality from an expert or other potential buyer
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
(4) Articles (in the press or industrial reports) written about the usage of
the specific car model
(5) Cost prediction and life expectancy
(6) Seller’s personal data as well as reviews from previous car sales
Table 2 Inclusion of design components in Lite and Complete prototypes
The experiments were conducted over three days in July 2018 (in total, 14 buyers, two private sellers, and one
professional seller participated in the experiments). The setting (location, tasks) was kept as similar as possible
between evaluations (with the difference that the professional sale took place in an underground car park). The
settings are shown in Figure 8, where buyers and sellers are completing sales in front of the cars used in the study.
Figure 8 Photos from three days of experiments
After each of the sales sessions (traditional, Lite and Complete as described above), the buyers received a survey
to complete. The survey consisted of questions about the buyer’s knowledge of the car and its quality (to capture
knowledge gained or lost in the sessions; based on the statements that were designed for and tested in the study,
as no useful metrics were found in the literature), the buyer’s trust in the seller (by measuring the perceived ability,
benevolence and integrity based on (Fuller et al. 2007)) and the perceived trustworthiness of the overall sales
process (based on (Söllner 2014)). The used questions were always measured on a 7-point Likert scale (from 1
beingI strongly disagree to 7 being “I strongly agree”). The survey questions used in the evaluation can be found
. The presented quantitative results shown below were calculated for N=10 buyers, involved in private sales
only. The four professional sales are excluded from these results, as we believe that the professional sales differ
significantly from private ones in respect to the trustworthiness of the seller and the overall sales process, and,
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
therefore, should be separated. Furthermore, four professional sales cannot provide any significant result as the
number of sessions is not large enough.
Qualitative results were obtained by conducting and analysing semi-structured interviews (Wengraf 2001) with
each of the buyers, which took place directly after all three sessions were finished. Qualitative results include
interviews with all 14 participating buyers (the interviewed buyers are further referred as K1 K14). The interview
guide contained questions about the following aspects: (1) trust, ability, benevolence and integrity of the seller and
how they felt about the sale conversation; (2) car knowledge; (3) the IT artifact: which of the prototypes they
would prefer in the next sale, parts they liked (or disliked) in the Lite and Complete prototypes, reputation features
in the Complete Prototype and how they feel the cardossier can be improved. All interviews were transcribed,
coded and analysed, with a focus on information asymmetry reduction and trust enhancement.
5.3 Evaluation results
In this section, the results of the evaluation are described in detail. We describe the results for reduction of
information asymmetries and trust enhancement separately in the consequent sub-sections. We cover both the
quantitative (surveys) and the qualitative (interviews) parts.
Figure 9 Knowledge about the car
Figure 10 Buyer trust in the seller
Figure 11 Buyer trust in the sales
Reduction of information asymmetries
Figure 9 displays summary statistics that compare the means of the buyer’s knowledge about the car for the
Traditional, Lite and Complete settings. The introduction of the Lite prototype improved the measure, and the
Complete Prototype setting further increased the mean. Altogether, the mean improved as follows: 3.5Traditional à
4.8Lite à 6.1Complete. It is clear that if the buyer used the Complete Prototype in the sale process, the buyer’s
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
knowledge about the car being sold nearly doubles. Two-sided paired-samples t-tests were conducted to compare
the means. There were significant changes (p < 0.01) from the means of Traditional (M.=3.47, S.D.=1.24) and Lite
(M.=4.87, S.D.=1.30), and highly significant changes (p < 0.001) from Lite to Complete (M.=6.10, S.D.=1.07), as
well as from Traditional to Complete, but not significant change from Traditional to Lite.
In the interviews, nearly all buyers remarked that they learned the most about the car during the Complete sessions,
followed by the Lite sessions and Traditional sessions respectively. Upon being asked in which setting he was able
to evaluate the car most effectively, K8 remarked: “The sale conversations with the IT support were the best. I
had information which was not talked about in the Traditional sale. I went around the car, looked at these things
and asked about them.”
Other buyers also remarked that because of the advanced information elements (3 – 6) that were available in the
last prototype, they were inspired to ask different and new questions about the car in comparison to the Traditional
and Lite settings. They reported that these new types of questions and the resulting dialogue further increased their
knowledge about the car. They also remarked that more information than “just the basics” (such as in the Lite
prototype) must be available. Upon being asked to compare the knowledge gain of the Lite and Complete
prototypes, K1 answered: “Using the cardossier Lite, I did not have as much information in the beginning, and I
had a bad impression upon realizing the high mileage of the car and not knowing more about it. I liked this a lot
more in the cardossier Complete.”
Furthermore, the evaluation of the car (3) by third parties had a big impact on the reduction of information
asymmetry for a lot of buyers. K3 even stated that out of all provided features, it was the most useful one for him:
“I learned the most [about the car] by reading the expert review in the third sale.” Other buyers, such as K6, also
appreciated the reviews provided by non-purchasing buyers: “So here I noticed, in the last one [the Complete
Prototype], the comments that were in it from interested people who have already seen the car... these thumbs up
and down, high mileage thumbs down, other things, like extras thumbs up. It also helped a lot to see the quality
and condition. That jumped into my eye.”
Enhancement of trust
The trust of the buyer in the seller has been increased from both Traditional to Lite, and from Lite to Complete.
On average, the mean improved as follows: 4.1Traditional à 4.9Lite à 5.5Complete. Figure 10 displays summary
statistics that compare the means of the trust of buyers in the seller for the Traditional, Lite and Complete settings.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
Two-sided paired-samples t-tests were again conducted to compare the means. There was a significant difference
(p < 0.05) in the mean of the trust in the seller for Traditional (M.=4.07, S.D.=1.44) and Lite (M.=4.87, S.D.=1.21).
There was also a significant difference (p < 0.01) in the mean of the trust in the seller for Traditional (M.=4.07,
S.D.=1.44) and Complete (M.=5.50, S.D.=1.11). However, there was no significant difference in the mean of the
trust in the seller for Lite (M.=4.87, S.D.=1.21) and Complete (M.=5.50, S.D.=1.11). These results suggest that
both the Lite and Complete prototypes increase the trust of the buyer in the seller, but there is no significant
increase from the Lite and Complete prototypes.
Figure 11 displays summary statistics of the buyer’s trust in the overall sales process (respectively the IT artifacts).
Summarized, the average mean improved as follows: 3.5Traditional à 4.9Lite à 5.7Complete. Two-sided paired-samples
t-tests were again conducted to compare the means. There was a significant difference (p < 0.05) in the average
means for Traditional (M.=3.47, S.D.=1.24) and Lite (M.=4.87, S.D.=1.30). Additionally, a highly significant
difference (p < 0.01) in the average means for Lite and Complete (M.=5.73, S.D.=1.07) was found. Finally, the
average means of Traditional and Complete also differed to a highly significant degree (p < 0.01).
In the interviews, many buyers reported that they felt a stronger sense of trust toward the seller during the Lite
setting and even more so in the Complete setting (when compared to the Traditional setting). For instance, upon
being asked in which conversation he had the most in trust in the seller, buyer K9 answered: “Certainly in the ones
with the cardossier.”
When comparing the trust of the buyer in the seller, the majority of buyers mentioned having higher trust in the
Complete setting over the Lite setting. The reputation mechanism (6) in particular seemed to have a big impact on
the buyer’s trust. For instance, when asked where he had the most trust in the seller, K4 reported: “Of course in
the last one [cardossier Complete]. This is because there were reviews about the seller in the Complete Prototype,
similar to on eBay. If he had good reviews, I also believed that he is a good seller. I did not have that with the
Classic and the Lite.”
Another interesting phenomenon was noticed during the experiments. Buyers used the information in the
prototypes to “fact-check” the seller, which affected their trust. For example, K5 notes: “What I heavily noticed
was that I had the most trust in the seller in the Complete setting. I was able to do a fact check, which increased
my trust in him by a large margin.” - K5. He explains the phenomenon as follows: “He had said two things, the
date of first registration and the mileage, which I quickly looked up again because I thought he said the wrong
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
thing. And then the app has given me the confidence again because he has said the right thing. For smaller things,
I trusted and believed him, and for bigger things, I looked them up.”
Thus, upon passing the “fact check” test, the trust between buyer and seller increases and upon failing, it decreases.
Other buyers reported the same sequence of actions and thought processes. Another important result was that the
Complete Prototype directly affected the seller’s ability, benevolence and integrity. Several buyers mentioned that
they felt these abilities of the seller were stronger in the Complete setting, compared to Lite or Traditional.
An important finding was identified in interviews with the buyers and sellers in both the requirements engineering
phase and in the experiments: The information provided by the cardossier prototype is only useful if its correctness
(meaning that information is verified to be accurate or truthful) is guaranteed. Otherwise, as some interviewees
argued, its usefulness would be nullified, as such unreliable information is only as valuable as the word of a seller
trying to sell them a lemon(in other words, not useful at all). This point is further supported by the fact that some
participants mentioned that they were unable to cope with the massive amount of information available at their
fingertips”. It follows that in the used car market, more information is not always better.
As one way to enrich the quality of information, the experiment candidates noted that they liked the design of the
timeline (2) in both prototypes. This was because each item in the timeline featured a small box showing that the
information was verified (using a green check-mark) and where the information was collected (by providing the
information source, e.g., the name of a garage where a service was conducted).
6 Discussion and Conclusions
The two research questions of our study were:
RQ1: How can the provision of car usage data in a blockchain platform reduce information asymmetries and
enhance trust in the used car market?
RQ2: How can further relational data be included to reduce information asymmetries and enhance trust in the used
car market?
The main source of information asymmetries has been identified to stem from the fact that the seller has knowledge
about the car being sold that the buyer does not (Akerlof 1970). The findings show that both the Lite prototype
and the Complete Prototype managed to increase the buyer’s knowledge about the car by an impressive margin in
both the private and professional sale experiment. In the private sale experiment, significant results were found.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
In both experiments, the Complete Prototype managed to outperform the Lite prototype by roughly the same
margin as the Lite prototype outperformed the Traditional setting. Thus, we can conclude that the cardossier
prototype successfully reduces the information asymmetries between buyer and seller. These results imply that the
buyer learned more about the car being sold and had better valuation success with both cardossier prototypes. As
the Lite prototype focused on providing only basic information elements (e.g., date of first registration, mileage)
about the car, this indicates that one way to reduce the information asymmetry is through providing basic
information elements about the car, as the literature suggests (Ba and Pavlou 2002; Mishra et al. 1998). Based on
the results, it follows that there should actually be more focus on the quality of the information than the quantity.
For this, the authentication property of blockchain technology is helpful (Miscione et al. 2018). However,
correctness of such data must be guaranteed (and backed by nodes, operating the system, or system designers) and
properly communicated to the user to enable the positive outcome. In the case of the cardossier prototype, we
added product-related information such as evaluations of the car from third parties (like experts), which provide
additional validation of the existing raw data from the timeline. Additionally to validation, these evaluations help
in making sense on the business level and help the end user, who lack experience in car sales, more accurately
make his/her decision. This finding suggests that applications that utilize data from blockchain platforms still
should include some third-party human actor that can reflect on quality of a product (in our case the car). With
this, we can also conclude that trust in the IT artifact itself (Söllner et al. 2012) plays an important role not only as
a mediator, but also as a trustee to build up initial trust in the system itself, and should be further explored in this
In our study, we also looked how product-related information, such as data analysis for costs prediction and life
expectancy, may be reflected in the interface. Though probably not crucial for trust building, these elements have
an important role for reduction of information asymmetries and offer a powerful mechanism for decision-making
on purchase of a car. These elements close the gap in knowledge and expertise of an average car buyer, addressing
the problems, mentioned by Zavolokina et al. (2019). Furthermore, in this study we show, how the requirements
for a blockchain-based application in the used car market, proposed by Zavolokina et al. (2019), can be actually
implemented in a prototype and, thus, evaluate their usefulness and limitations.
Furthermore, the fact that person-related data such as reputation data is stored in the system may hinder the
participation of car owners, who may be unwilling to have negative (and sometimes untruthful) information about
them stored. With this, the privacy issue should be considered as well, concerning what personal data should be
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
stored in the blockchain and how a car owner may retain control over this data. Furthermore, a question of
reputation transferability (Kokkodis and Ipeirotis 2015) between platforms in the context of a car sale becomes
even more relevant. A private seller is unlikely to sell many cars in a short period of time; therefore, his/her
reputation cannot be based on a large number of transactions, but may be provided by other platforms (e.g., auction
sites like eBay, or social media). To make this reputation transfer possible, blockchain technology may be
beneficial due to its distributed and decentralized operation. These are the questions that should be further studied.
The Complete Prototype focused on providing advanced information elements and market transparency (such as
interactive elements showing estimated costs; provision of car evaluations, featuring reviews by independent
experts; and non-purchasing buyer ratings), as well as a reputation mechanism, and therewith managed to further
improve the car knowledge of the buyers. This leads us to the conclusion that by adding such advanced information
elements as well as a reputation system, we can further enrich the buyer’s knowledge and thus reduce the
information asymmetries, improving the situation in ‘The Market for Lemons’ (Akerlof 1970). Using the insights
from the research at hand, we further study effects of the cardossier application on market transparency in the
subsequent study (Bauer, Zavolokina, and Schwabe 2019). However, the interplay between the elements must be
studied further: is market transparency together with a reputation mechanism stronger than these elements in
isolation? And what role do they play in the trust relationships? This also brings us back to the question of what
value blockchain technology creates. While prior literature focused on the value, blockchain technology generates
to businesses (Bauer, Zavolokina, Leisibach, et al. 2019), and to markets (Bauer, Zavolokina, and Schwabe 2019),
in our study we were able to examine the user perspective in a face-to-face used car sales scenario. In our study,
we were able to focus more on the personal interaction between used car buyers and sellers, and their trust
relationships. We can conclude that, firstly, it is not blockchain technology per se which brings value, but an
application built on top of it. The technology itself, however, contributes to the availability and the trustworthiness
of car-related data provided. While there may be other sources of car-related information, such as CarFax report
or the ones provided by car manufacturers, they were not accessible in the context we examined (not because of
the study design, but in general in the country, where the study was conducted; therefore, the participants of the
study were also not familiar with them). However, from the user perspective, the trustworthiness and authentication
property (Miscione et al. 2018), should be communicated to the user in a straight-forward way (for this,
explanatory tools, such as FAQs, videos, are helpful). If this condition is fulfilled, the users appreciate
authenticated data a lot. With this, we confirm that trust is the main value that blockchain technology creates
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
(Fleischmann and Ivens 2019), but this ‘trustfulness’ of data should be established not only on the technical level,
and on the user level (e.g. by communicating). What other factors contribute to and make blockchain-based data
trusted should be studied further.
The findings show that both the Lite and Complete prototypes managed to increase the buyer’s trust in the seller.
This effect was significant for the Lite setting and highly significant for the Complete round. Interestingly, there
was no significant change in the buyer’s trust in the seller from the Lite to the Complete setting. This implies that
the trust relationships were already improved by the reduction of information asymmetries (making the basic data
and the events, provided by an external party in the cardossier, available in the digital form). The interviews
conducted following the experiments can help explain this result. Several buyers used the information in the
prototype to “fact-check” the seller. In a first step, they looked for and read specific information, such as the car’s
mileage, in the prototype. Then, in the sale conversation itself, they tested the seller by “fact-checking”. Using the
example of the mileage, they asked the seller what the car’s mileage was. Several buyers mentioned in the
interviews that if the mileage given by the seller matched what they read in the prototype, they would see him as
more trustworthy. Similarly, if the seller gave a different answer, they would lose trust in the seller. However,
results suggest that the additional information elements improve the overall purchase experience of buyers.
To summarize, this study contributes to the literature in several ways. First, it contributes to the stream of literature
on solutions helping to resolve information asymmetries and enhance trust in inter-personal relationships by
introducing a design of such application and discussing its usage. Results of this study can be transferred to markets
with similar conditions: i.e. where history and quality of a physical good is important and, if hidden, cause mistrust
between buyers and sellers. A good example of such can be real estate market. Second, it contributes to the
literature on blockchain-based applications, making a conclusion that, from end users’ perspective, ‘raw’ data from
a blockchain system does not suffice to resolve the mentioned problems, additional support for sense-making is
needed. We show, how this may be reflected in an application design.
7 Limitations
It is important to recognize the following limitations. Firstly, the limitation resulting from the use of the main
DSRM approach. Since the goal is to find comprehensive solutions to both the information asymmetry and trust
problems, the results cannot be attributed to isolated, simple factors such as colors or pictures used in the
prototypes. Further limitations also result from the evaluation itself. One limitation was that the experiment setting
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
was near-realistic. This setting was selected purposely, but cannot be compared to laboratory conditions, where
most factors can be controlled easily. Due to the complex nature of a car sale, this was the best available option.
Another limitation was that the test participants (with the exception of the professional car seller) were all hired
and received monetary compensation for acting out a scenario in which they were tasked to buy a car. Due to the
stickynature of information, the order in which the three sale conversations were conducted was always the
same. This could lead to buyers changing behavior, as well as remembering information from previous
conversations and building more trust through extended exposure to the seller. Furthermore, we evaluated the
proposed design component as a whole; the separation existed only between the Lite and Complete prototypes.
Therefore, it is hard to isolate the components that had the greatest impact on the results. We also recognize that
the measures used were focused more on the perceived and subjective feelings of the participants, as they were
asked in the survey, rather than any other possible evaluation strategies to capture gained knowledge or
experienced trust. Finally, the sample size for both experiments is small, so the offered results have some
limitations in generalizability.
Akerlof, G. A. 1970. “The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism,” The
Quarterly Journal of Economics, pp. 488500.
Ba, S., and Pavlou, P. A. 2002. Evidence of the Effect of Trust Building Technology in Electronic Markets:
Price Premiums and Buyer Behavior,” MIS Quarterly, pp. 243268.
Bauer, I., Zavolokina, L., Leisibach, F., and Schwabe, G. 2019. “Exploring Blockchain Value Creation: The
Case of the Car Ecosystem,” in 52nd Hawaii International Conference on System Sciences.
Bauer, I., Zavolokina, L., and Schwabe, G. 2019. “Is There a Market for Trusted Car Data?,” Electronic
Markets, pp. 115.
Brousmiche, K. L., Heno, T., Poulain, C., Dalmieres, A., and Hamida, E. B. 2018. “Digitizing, Securing and
Sharing Vehicles Life-Cycle Over a Consortium Blockchain: Lessons Learned,” in New
Technologies, Mobility and Security (NTMS), 2018 9th IFIP International Conference On, IEEE,
pp. 15.
Fleischmann, M., and Ivens, B. 2019. “Exploring the Role of Trust in Blockchain Adoption: An Inductive
Approach,” in Proceedings of the 52nd Hawaii International Conference on System Sciences.
Fuller, M. A., Serva, M. A., and Benamati, J. “Skip.” 2007. “Seeing Is Believing: The Transitory Influence
of Reputation Information on e-Commerce Trust and Decision Making,” Decision Sciences (38:4),
pp. 675699.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
Granados, N. F., Gupta, A., and Kauffman, R. J. 2006. “The Impact of IT on Market Information and
Transparency: A Unified Theoretical Framework,” Journal of the Association for Information
Systems (7:3), p. 7.
Granados, N., Gupta, A., and Kauffman, R. J. 2008. “Designing Online Selling Mechanisms: Transparency
Levels and Prices,” Decision Support Systems (45:4), pp. 729745.
Gregor, S., and Hevner, A. R. 2013. “Positioning and Presenting Design Science Research for Maximum
Impact,” MIS Quarterly (37:2), pp. 337355. (
Hevner, A. R., March, S. T., Park, J., and Ram, S. 2004. “Design Science in Information Systems Research,”
MIS Quarterly (28:1), pp. 75105. (
Jøsang, A., Ismail, R., and Boyd, C. 2007. “A Survey of Trust and Reputation Systems for Online Service
Provision,” Decision Support Systems (43:2), pp. 618644.
Kokkodis, M., and Ipeirotis, P. G. 2015. “Reputation Transferability in Online Labor Markets,” Management
Science (62:6), pp. 16871706.
Levy, Y., and Ellis, T. J. 2006. “A Systems Approach to Conduct an Effective Literature Review in Support
of Information Systems Research.,” Informing Science (9).
March, S. T., and Smith, G. F. 1995. “Design and Natural Science Research on Information Technology,”
Decision Support Systems (15:4), pp. 251266.
Mayer, R. C., Davis, J. H., and Schoorman, F. D. 1995. “An Integrative Model of Organizational Trust,”
Academy of Management Review (20:3), pp. 709734.
McKnight, D. H., and Chervany, N. L. 1996. The Meanings of Trust.
Miscione, G., Ziolkowski, R., Zavolokina, L., and Schwabe, G. 2018. “Tribal Governance: The Business of
Blockchain Authentication,” in Prepared for the Hawaii International Conference on System
Sciences (HICSS).
Mishra, D. P., Heide, J. B., and Cort, S. G. 1998. “Information Asymmetry and Levels of Agency
Relationships,” Journal of Marketing Research, pp. 277295.
Notheisen, B., Cholewa, J. B., and Shanmugam, A. P. 2017. “Trading Real-World Assets on Blockchain,”
Business & Information Systems Engineering (59:6), pp. 425440.
Nunamaker, J. F., Briggs, R. O., Derrick, D. C., and Schwabe, G. 2015. “The Last Research Mile: Achieving
Both Rigor and Relevance in Information Systems Research,” Journal of Management Information
Systems (32:3), pp. 1047. (
Peffers, K., Tuunanen, T., Rothenberger, M. A., and Chatterjee, S. 2007. “A Design Science Research
Methodology for Information Systems Research, Journal of Management Information Systems
(24:3), pp. 4577.
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
Pries-Heje, J., Baskerville, R., and Venable, J. R. 2008. “Strategies for Design Science Research
Evaluation.,” in ECIS, pp. 255266.
Resnick, P., Kuwabara, K., Zeckhauser, R., and Friedman, E. 2000. “Reputation Systems,” Communications
of the ACM (43:12), pp. 4548.
Rezabakhsh, B., Bornemann, D., Hansen, U., and Schrader, U. 2006. “Consumer Power: A Comparison of
the Old Economy and the Internet Economy,” Journal of Consumer Policy (29:1), pp. 336.
Riege, C., Saat, J., and Bucher, T. 2009. “Systematisierung von Evaluationsmethoden in Der
Gestaltungsorientierten Wirtschaftsinformatik,” in Wissenschaftstheorie Und Gestaltungsorientierte
Wirtschaftsinformatik, Springer, pp. 6986. (
Seebacher, S., and Schüritz, R. 2017. “Blockchain Technology as an Enabler of Service Systems: A
Structured Literature Review,” in International Conference on Exploring Services Science, Springer,
pp. 1223. (
Söllner, M. 2014. Deriving Trust Supporting Components for Ubiquitous Information Systems, (Vol. 1),
kassel university press GmbH.
Söllner, M., Hoffmann, A., Hoffmann, H., Wacker, A., and Leimeister, J. M. 2012. Understanding the
Formation of Trust in IT Artifacts, presented at the International Conference on Information
Systems, Association for Information Systems. (
Söllner, M., and Leimeister, J. M. 2013. What We Really Know About Antecedents of Trust: A Critical
Review of the Empirical Information Systems Literature on Trust,” Psychology of Trust: New
Research, D. Gefen, Nova Science Publishers.
Swan, J. E., and Nolan, J. J. 1985. “Gaining Customer Trust: A Conceptual Guide for the Salesperson,”
Journal of Personal Selling & Sales Management (5:2), pp. 3948.
Venable, J., Pries-Heje, J., and Baskerville, R. 2012. “A Comprehensive Framework for Evaluation in Design
Science Research,” in International Conference on Design Science Research in Information
Systems, Springer, pp. 423438.
Wengraf, T. 2001. Qualitative Research Interviewing: Biographic Narrative and Semi-Structured Methods,
Zavolokina, L., Miscione, G., and Schwabe, G. 2019. “Buyers of Lemons: Addressing Buyers’ Needs in the
Market for Lemons with Blockchain Technology,” in 52nd Hawaii International Conference on
System Sciences.
Zavolokina, L., Ziolkowski, R., Bauer, I., and Schwabe, G. 2020. “Management, Governance and Value
Creation in a Blockchain Consortium,” Forthcoming in MISQE.
Ziolkowski, R., Parangi, G., Miscione, G., and Schwabe, G. 2019. “Examining Gentle Rivalry: Decision-
Making in Blockchain Systems,” in Proceedings of the 52nd Annual Hawaii International
Reducing Information Asymmetries and Enhancing Trust in the ‘Market for Lemons’
Conference on System Sciences, Maui, Hawaii, USA, January 8.
... The cardossier platform leads to increased data quality [14], new business models [15], and increased market transparency [16]. It ultimately moves more used car business from the garage to online platforms [17]. Used-car dealers may use the certified data of the cardossier platform (1) to apply advanced analytics to evaluate the car state and value, (2) to buy used cars on online platforms, and (3) to offer advanced warranties [18]. ...
... A cardossier has been shown to increase the data quality [14] and, thus, is a good foundation for well-built AI models. This AI then can be used to extend the online platforms introduced in [17] with the new functionalities. ...
The used-car market is notoriously untrustworthy and shady. Certified data has been shown to help mitigate the information asymmetry, one of the major factors to an untrustworthy market. In recent times, more and more used-car dealers have had problems surviving in this competitive data-driven market. In this study, we conduct 12 interviews with used-car dealers and several meetings and workshops with employees and executives from the AMAG Group, one of the largest automotive companies in Switzerland. This creates insight into current problems for used-car dealers and how artificial intelligence can help. The problems can be abstracted to the problem of high transaction cost and its subcategories. In reducing transaction costs by utilizing artificial intelligence, new secondary problems arise. People need to trust the certificate, the analytics, and the predictions. Additionally, the data and analytics need to be transparent and understandable, and privacy concerns must be addressed. The implications of this study are manifold. First, we define the problems for used-car dealers on the used-car market and introduce artificial intelligence approaches to the current data-driven used-car market. Afterward, we stress that artificial intelligence needs to follow a human-centered perspective and be designed for trust.
... We acknowledge that multi-party certification could be delivered by any (existing or future) technology similar to blockchain that would address the needs of its providers in terms of power decentralization and unique digital ownership. Figure 1) While the first two subsystems concern the development of multi-party certificates, the last subsystem, the Buyers and Sellers of Used Cars Subsystem (marked in white in Figure 1) concerns the use of multi-party certificates: The social actors that are intended to use multi-party certificates are the sellers and buyers of used cars [9,64,65]. These are confronted with problems of information asymmetry during used-car sales (input) [3]. ...
... Furthermore, it can mitigate adverse selection 1 The key-difference between an organizational field and an ecosystem is that the former includes actors producing a similar product or service (they belong to the same industry) while the latter is a cross industry network of producers of many different types of goods and services that can be mixed/coupled into different combinations (Iansiti and Levien 2004). effects and reduce the moral hazard of good market participants (Notheisen and Weinhardt 2019;Zavolokina, Schlegel, and Schwabe 2020). Similarly, the aspect of the risk mitigation can ensure the absence of any collapse of the consumption of art products and/or services should the transaction involve an artwork which turns out to be a fake. ...
Blockchain technology is currently stimulating a broader process of social and industrial transformation impacting several potential areas of adoption (e.g. transactions, tracking and tracing solutions, authenticity, etc …). The acceptance of this technology represents a major challenge for the art ecosystem. The literature around relations and applications of blockchain in the art market is fragmented and fails to provide an understanding of the current/potential opportunities of this application. This paper explores how blockchain is being adopted in the art sector. By using a perspective of mobilizing organizational and institutional field theory, this study performs an explorative qualitative analysis based on semi-structured interviews with 15 experts from the art market. The findings underline the complexity of effective implementation of this technology, the contradictory positions regarding the benefits and business opportunities it offers, as well as the market players' resistance to change and trust. The theoretical and practical implications are also discussed.
... -Setting up the efficient price in markets because information asymmetries are eliminated among stakeholders 54 . ...
Full-text available
Such a disruptive technology as the blockchain platform has a potential to provide value-added contributions towards multiple-disciplinary applications these days. Blockchain technology, considered as the second generation of the Internet in the digital age, has been applied in practice for various industry such as finance, healthcare, tourism, retailing, manufacturing, education, public sector, supply chains management, agri-food industry and so on. In which, supply chain operations, particularly agri-food management, becomes a prosperous beneficiary from blockchain technology. In the local context, Vietnam has been achieved the competitive advantages in agricultural production among the world in some agri-food products such as black peppercorn, cashew nuts, coffee, coconut, rice, and rubber. Furthermore, Vietnam is considered as one of pioneers in implementing blockchain technology for traceability and transparency of agri-food products across Association of South East Asian Nations (ASEAN) countries. However, design and implementation of the blockchain technology application in agri-food production in the case of Vietnam agriculture industry have been confronted with opportunities and challenges, besides its strengths and weaknesses in nature. This paper aims to concentrating on the analysis of Strengths, Weakness, Opportunities and Threats (SWOT) of blockchain-applied supply chain relevant to agri-food products. In addition, the application of blockchain technology in supply chains has been expected to promote the sustainable development of Vietnam agriculture, in the accordance with the United Nation Sustainable Development Goals (SDGs) in the 21st century. We adopt the research method of interpretive approach to synthesize the literature and reasonably induce research conclusions. We finally prescribe some policy recommendations and implications for agri-food stakeholders such as regional policy makers, agri-food based blockchain platform designers, executives, and farmers as users in relevant to facilitate the blockchain technology for agri-food chains in Vietnam agriculture.
... where additional trust support in an application design and sense-making of blockchain-based data are favorable [54]. This simple scenario gives an idea of a setting in which the cardossier platform is used by a car buyer: ...
Full-text available
Trust is a crucial component for successful transactions regardless of whether they are executed in physical or virtual spaces. Blockchain technology is often discussed in the context of trust and referred to as a trust-free, trustless, or trustworthy technology. However, the question of how the trustworthiness of blockchain platforms should be demonstrated and proven to end users still remains open. While there may be some genuine trust in the blockchain technology itself, on an application level trust in an IT artifact needs to be established. In this article, we examine how trust-supporting design elements may be implemented to foster an end user's trust in a blockchain platform. We follow the design science paradigm and suggest a practically useful set of design elements that can help designers of blockchain platforms to build more trustworthy systems.
... Figure 34 Architecture of the cardossier platform Though the project has a number of facets which are crucial considerations for the design of the platform (like platform governance , business model (Bauer, Zavolokina, Leisibach, et al. 2019), incentive system , blockchain consortium management (Zavolokina et al. 2020)), we focus on its application level and the perspective of its end users: car buyers, willing to consult the cardossier to assess the quality of a used car they intend to buy. Prior studies report on the problems and needs of car buyers (Zavolokina, Miscione, et al. 2019), where additional trust support in an application design and sense-making of blockchain-based data are favorable (Zavolokina, Schlegel, et al. 2019 ...
Full-text available
The automotive market is in the top three markets with least trust from consumers (followed by banks, and telecommunications industry) (Lades 2017). In particular, in the used car market, consumers encounter such problems as: a car in a poorer condition than initially indicated, accident damage that was not previously disclosed, little or no documentation provided, and fraud. Nobel prize winner G.A. Akerlof described the market for used cars as an example of the problem of information asymmetries and resulting quality uncertainty (Akerlof 1970). In my dissertation, I look at how an application on top of blockchain technology, which is claimed to bring trust by design due to its inherent characteristics, may be designed as a novel solution to resolve problems in the ‘Market for Lemons’ and, thus, enhance trust between buyers and sellers. While there may be some genuine trust in blockchain technology itself, on an application level, trust in an IT artifact needs to be established. This need for trust establishment is at heart of this dissertation project. This cumulative dissertation is structured in three parts. In part I, I explore the characteristics of blockchain platforms and relevant aspects for their design (i.e., the data quality aspect and token design). In part II, I propose a design to address problems associated with a of lack of trust, caused by information asymmetries, between car buyers and car sellers with the help of an IT artifact: a blockchain-based cardossier. The cardossier application reflects a car’s history over its whole lifecycle (from production to utilization). The proposed design relates to the end-user-facing frontend application. In this part, I focus on the problems and needs of used car buyers, the requirements placed for the design of the application, the actual elements of the application design, and the effects of the application’s usage in the ‘Market for Lemons’. The results show that the problems buyers experience are a paradox of having a large variety of information sources available about brands, and the quality characteristics and performance of certain models; while information about a specific car is not transparent and often even hidden. Though providing more information about a used car leads to fewer information asymmetries in general, advanced information elements are beneficial in improving the situation further. As for trust, such a system enhances trust between buyers and sellers and, in general, makes the overall purchase process more trustworthy. However, to achieve these positive effects, the quality of the stored information should be guaranteed and properly communicated to the end-user. Additionally, exploring the use of the cardossier, embedded into an online used car sales platform, several groups of factors (technology-related, data-related, design-related, user-related) that play an important role in trust formation were revealed. Furthermore, through adopting the perspective of technology affordances, we identified blockchain platform affordances related to the trust in a blockchain platform. In part III, the transfer of a research project into practice, preparation for market entry, associated tensions the blockchain consortium had to face while implementing the platform are discussed along three themes: consortium management, business value, and governance. Altogether, the three parts of the dissertation provide better understanding of blockchain platforms, applications built on top of them and their usage in the ‘Market for Lemons’. Thus, this dissertation contributes to the body of knowledge on trust design in the context of blockchain technology and applications on top of it, and also to the broader literature on examining the ‘Market for Lemons’ and solutions to its problems. As blockchain applications are still being developed, and as we are just beginning to understand the potential of the blockchain technology, such research has scientific value in terms of design and organizational implications.
Full-text available
In recent years, an increasing number of blockchain consortia have emerged. However, little is known about how these consortia are developed and what tensions emerge in such collaborations. We describe the evolution of the cardossier blockchain consortium in Switzerland, which is building a system for managing car data and seeking to improve collaboration between players in the car-related ecosystem. From our involvement with the cardossier project, we have gained insights that will be valuable for enterprises considering whether to join a blockchain consortium.
Full-text available
Many marketing exchanges are characterized by an information asymmetry between suppliers and customers. Specifically, customers are faced with both adverse selection and moral hazard problems that involve, respectively, uncertainty about supplier characteristics and the risk of quality cheating. Drawing on prior research, the authors propose that agency problems in a customer relationship can be resolved by means of customer bonds and price premiums, which serve as signals and supplier incentives, respectively. The authors also propose that adverse selection and moral hazard problems exist in relationships between suppliers and their employees. Similar to the customer relationship, these problems can be addressed with signals and incentives of various kinds. The authors present hypotheses regarding the agency problems in both of these relationships and test them empirically in the context of automotive service purchases. Data obtained from 287 service managers support the hypotheses. The data also suggest that institutional differences across service outlets (e.g., ownership structure and size) influence how the two types of agency problems are managed.
Conference Paper
Full-text available
The blockchain comes with the promise of being a disruptive technology with the potential for novel ways of interaction in a wide range of applications. Although scholarly interest in the technology is growing, a broad analysis of blockchain applications from a governance perspective lacks to date. This research pays special attention to the governance of blockchain systems and illustrates core governance decisions on 15 blockchain systems from four application domains. Based on academic literature, semi-structured interviews with representatives from those companies, and content analysis of grey literature, different blockchain governance decisions have been derived and their enactment described. The identification of them enriches the scarce body of knowledge on blockchain systems with a better understanding of how key governance decisions are enacted in practice.
Conference Paper
Full-text available
Blockchain is expected to create a variety of new opportunities for businesses. Yet, little is known about how companies can exploit business value from the technology. However, without a clear understanding of how, and corresponding adaption of business practices, the realization of value is doomed to failure. Hence, we contribute to this gap by analyzing and explicating the specificities of value creation from blockchain in the ecosystem of a car. In the course of an exploratory case analysis we conducted interviews and workshops with industry and blockchain experts from five diverse stakeholder groups. In brief, we provide early evidence that (1) blockchain enables value creation through: Distributed Product Innovation, Controlled Customer Intimacy and Shared Operational Efficiency. Further, (2) we derive guidelines and discuss learnings for other businesses aiming to leverage value from blockchain technology.
Conference Paper
Full-text available
The automotive market is in the top three markets with the least trust from consumers. In particular, in the secondhand car market, consumers suffer from such problems as the car being in worse condition than initially indicated, accident damage that is not disclosed, fraud, etc. Akerlof, described the market for used cars as an example of the problem of information asymmetries and resulting quality uncertainty. In order to cope with quality uncertainties, used car buyers actively engage themselves in information seeking. Blockchain technology promises to automatize the tracking of cars through their lifecycles and provide reliable information at any point in time it is needed. In our study, we investigate the problems car buyers face during information seeking and propose requirements for the design of a blockchain-based system to address these.
Conference Paper
Full-text available
The blockchain technology offers a novel mode of distributed authentication, which does not depend on a central authority. We consider this novelty against established governance modes. We illustrate our argument by paying special attention to blockchain-based authentication functions in the empirical domain of land registries across the world. Based on interviews with representatives from organizations deploying blockchain, and content analysis of related grey literature, we discuss established governance idealtypes against what the rivalry that cryptocurrencies and blockchains bring to digital settings. After referring to market, hierarchy, network, and bazaar, we conclude outlining the prospects of a different , blockchain-related governance mode called 'tribal' that better captures the 'togetherness' which rivalry originates.
The used-car trade is characterized by information asymmetries between buyers and sellers leading to uncertainty and distrust, thus causing market inefficiencies. Prior research has shown that blockchain offers a solution: a transparent, trustworthy and verified car history that addresses these issues in the market for ‘lemons’. Yet, whether or not there really is a market for trusted car data remains an open question. In particular, it is unclear if trusted car data increases transparency in the market for lemons and how market participants value increased transparency. Hence, through a market game with 50 participants, we explored the effects of trusted car data on the sales price of the cars, and the relative revenue of buyers and sellers. Additionally, we conducted interviews with the participants to elicit the perceived customer value. The results show that blockchain enables an increase in transparency and creates value for both buyers and sellers.
2001 introduction to in-depth semipstructured qualitative interviewing and to BNIM in paerticular. Unique in its conceptual coherence and its level of practical detail, it cov ers a full spectrum from the identification of topics and research questions, to the interviewing, to the answerin g of research questions, the compring and theorising of cases an d to strategies of writing-up presentations.