Crowdsourcing is a kind of participative online activity in which a large and possibly undefined group of people contribute to the tasks outsourced by requesters through a flexible open call [1
]. Inspired by the financial or non-financial benefits, there is a popular trend for firms or individuals to apply and for solvers to make use of crowdsourcing to handle different possible outsourced tasks, especially problem-solving tasks [2
]. In respect to problem-solving tasks, contest is an extremely widely adopted crowdsourcing form for addressing them [5
]. Furthermore, a large number of online crowdsourcing platforms (OCPs) supporting problem-solving contests have been developed and grown rapidly, such as Hyve, Designcrowd, 99design, and GoPillar. At the same time, millions of solvers around the world have conducted activities on these OCPs and millions of dollars have been transferred from requesters to solvers [5
OCPs act as intermediaries that connect and serve requesters and solvers. Their sustainable development heavily depends on the wide and continuous participation of solvers [7
]. To stimulate solvers to participate, academic researchers have made great efforts on identifying and examining the factors that influence solvers’ participation in a specific task or a specific OCP. These factors mainly include solver motivation [10
], expertise [12
], and cultural background [14
] and external factors like requester and platform fairness [16
] and feedback [18
], task attributes [20
Meanwhile, many practical suggestions were developed for OCPs to improve their designs and services. In those suggestions, researchers all highlighted the importance of serving solvers effectively and gave a few specific pieces of advice. For example, Kohler [23
] considered more than 20 leading crowdsourcing ventures to identify strategies for scaling crowdsourcing platforms. Further, curating existing creators and attracting new creators were two important strategies. Blohm et al. [7
] suggested 21 governance mechanisms to govern platforms and further investigated their effectiveness in 19 platform case studies. Out of these mechanisms, providing proper incentives, giving feedback against solvers’ contributions, promoting socialization, modularizing tasks were effective for activating contributors to solve problems. Johnson and Liew Chern [24
] presented four main crowdsourcing platform design recommendations which were to promote ease of use, attract and sustain solver interest, foster a community of solvers, and show solvers’ contributions after analyzing 12 New Zealand public cultural heritage institutions crowdsourcing platforms.
In addition, in respect to specific problem-solving contests, scholars indicated that appropriate contest design is helpful for keeping solvers’ participation. For instance, Ebner et al. [25
] concentrated on IT-supported idea competitions and suggested widening the topic of the ideas competition and offer an attractive incentive structure in order to keep solvers’ participation. Ren et al. [26
] proposed a top-down process of designing contest which means that design contest in advance with proper IT artifacts in order to address workers’ motives. Oguz Ali Acar [27
] emphasized the importance of activities and technical features that enable one to socialize with other participants, support active participation, and create a participatory experience in idea contests.
Prior studies provide us solid knowledge to understand why (motives) and what factors influence solvers’ participation in a specific task or an OCP in terms of theoretical basis. Additionally, from the practical perspective, some valuable suggestions were developed for improving OCP or a specific problem-solving contest design. Nevertheless, few studies focus on typical and widely used OCPs particularly for problem-solving contests and further identify what these OCPs actually do which aim to serve their solvers. One of the reasons that we are concerned with OCPs is that they are the only places where solvers engage in problem-solving contests. Solvers would search and examine an OCP’s features that they then concern and even compare them with other similar OCPs in making a decision to participate. In other words, what OCPs actually do rather than what they should do catches solvers’ sight first and has a direct impact on solvers’ participation decisions. Another reason is that there is a lack of a whole picture of what OCPs do for solvers in problem-solving contests although the existing studies have recommended some separated measures for serving solvers effectively under specific contexts.
In an attempt to fill this gap, this study conducted a content analysis of OCPs’ websites for problem-solving contests to identify what they have done for solvers. These websites include all information provided by OCPs for solvers and act as the intermediaries where solvers directly engage in contests and interact with others. On these websites, what the OCP does is conveyed explicitly by specific service items, standards and regulations, rules, and tools. We name them as service measures that aim to serve, support, regulate, and orchestrate solvers’ participation in this study. Consequently, 14 major service measures that are related to contest management, solver management, and requester management were identified from 25 OCPs concentrating on problem-solving contests. Furthermore, we suggested a framework for demonstrating the relations among these service measures. Our analysis offers a comprehensive list of measures for serving solvers by conducting a content analysis of the OCPs. It, on the one hand, contributes to theoretical researches in the field of crowdsourcing by providing a reference to identify and analyze the factors that impact solvers’ participation in problem-solving contests. On the other hand, the study would contribute to the content analysis method by extending its application into the crowdsourcing field. In terms of practical implications, our findings may act as a guideline for OCPs to improve their solver service and for solvers to make a decision to participate in the appropriate OCPs.
The rest of this paper is organized as follows. Section 2
presents preliminaries related to this study. Section 3
demonstrates methods and Section 4
demonstrates research results. Discussion and implications are illustrated in Section 5
. The final section concludes the study.
The research design involves content analysis of service measures presented by OCPs on their websites. The content analysis method is a qualitative research method that can be used for analyzing all kinds of media texts [54
]. It is considered a common method in communication studies and has been used widely in investigating communication phenomena based mostly on texts on websites [54
]. Application of content analysis began with actual observations and the collection of original documents, and then proceeded to analyze, code, and refine the concepts and categories before constructing the systematic theory [57
]. Figure 1
presents the detailed process of performing content analysis in this study.
3.1. Sample Identification
Web searches were conducted using the terms “crowdsourcing platform”, “crowdsourcing website”, and “crowdsourcing contest” using the search engine Google in May 2020. We first recorded and examined OCP websites for problem-solving contests on the web directly and those that were indicated in web articles such as “top 15 world’s best crowdsourcing platforms and websites|2019 best sites” and “the 4 best crowdsourcing platforms for graphic and product design”. At this stage, a total of 79 websites were retrieved. After deleting those duplicate and defunct, 19 websites that meet all inclusion criteria remained. The criteria included (1) link various requesters and solvers; (2) focus on problem-solving tasks (e.g., graphic design, interior design, software development, data mining and analysis), those focusing on simple task were not included, such as AMTurk; (3) contest is the primary way for problem-solving; (4) crowdsourcing contest is the major value unit, so a few known OCPs like Lego ideas, Threadless, and Kaggle were excluded; (5) money as the major extrinsic incentive. This criterion excludes non-profit OCPs like OpenIdeo and Greenchallenge.
To avoid missing some representative OCPs, this study also collected those investigated in prior studies, such as listed 22 OCPs in [23
], 116 open innovation web-based platforms in [59
], and some examples in [15
]. With the above inclusion criteria, six OCP websites were added to the initial samples. Consequently, the final sample consists of 25 OCP websites that are presented in Table 1
. The information about each OCP was collected directly from its webpages on 26–27 July 2020. We filled the symbol “/” if data could not be found.
Of the selected OCPs, 11 focused on logo design, web design, and graphic design, 7 focused on innovative ideas, concepts, and solutions creation aiming to solve various challenges, 3 served for video production and idea creation, 3 focused on architecture and interior design, and only “Topcoder” worked in software design and build.
3.2. Unit of Analysis and Data Preservation
Selecting the unit of analysis is one of the most basic decisions during implementing a content analysis [61
]. This study took OCP websites as research objects and analyzed what they do for solvers one by one. Each of them is enough to be considered as a whole and to be possible to keep in mind as a context for analyzing. Thus, the suitable unit of analysis is each selected OCP website. With respect to each website, we mainly concentrated on manifest content like visible and obvious components and regulations, policies, and rules in text. Further, as these contents generally do not vary frequently in a period, we observed, extracted, recorded, and coded them directly from the webpages of each website and did not solely download and preserve them. Conduction of the above work was in June–August 2020.
3.3. The Pilot Study
For all units of analysis, the authors first browsed and acquainted each unit’s major components and their relations. Generally, five kinds of content, that is, contests and their management, solvers and their management, requesters and their management, OCPs’ basic and operation information, community/forum were presented on the OCPs’ websites. From these contents, we checked and extracted the service measures developed by OCPs for solvers. Furthermore, based on the components of a crowdsourcing system and factors influencing solvers’ participation demonstrated in Section 2
, we divided the extracted service measure into three top categories. They are related to contest management, solver management, and requester management. Subsequently, under each of the categories, specific texts and descriptions of measures were observed, collected, coded, and further classified into sub-categories.
3.4. Coding Guide Development and Coding Procedures
We first reviewed six representative websites including “Eÿeka”, “HYVE”, “99designs”, “DesignCrowd”, “Topcoder”, and “Gopillar”, and further drafted a coding guide. Since the number of OCPs in this study is not too large, so our coders were working together to review, refine, and retest to develop consistent definitions, examples, and codes iteratively. The codings subject to disagreement were revisited until the agreement was reached. The specific procedure of coding content of a single website was as follows.
Step 1: Observing and coding related content on the selected websites one by one. We selected “Eÿeka” as the first website for content analysis. The manifest content on its webpages was observed, examined, and recorded in an Excel sheet. In accordance with defined categories at the stage of a pilot study, we divided the content in terms of contest, solver, and requester management. For content such as illustration pictures of a specific contest and videos illustrating how the platform works were not recorded but were reviewed by the authors. On the other hand, we did not review all contests published on “Eÿeka” but just chose five ongoing contests and five completed contests when we investigated its contest management.
Step 2: We viewed the recorded text as a meaning unit and further read, condensed, and wrote notes and headings which reflect its core content, that is, codes. Then, the codes with similar meanings were integrated together. Subsequently, a category was formed to group the list of codes and named using content-characteristic words.
Step 3: It is noted that the content that needs to be analyzed is determined by research aims. It means that not all text should be coded step by step. We mainly paid attention to the content that OCPs presented to solvers. For example, the intellectual property rights (IPR) policy on “Eÿeka” has thousands of words. We did not code all these texts because some of them may not be concerned by solvers. The main texts that are relevant to who owns solvers’ intellectual properties under what conditions were taken into account in this paper. So, the code is “requesters could own IPR of a solver’s submissions only if they have paid for them”, the sub-category is “Ownership transfer and use of solvers’ submissions”, and the category is “service measures related to contest management”. The above meaning units, condensed meaning units, codes, sub-categories, and categories were recorded in the Excel sheet. The coding process was performed, discussed, and refined by the authors together.
After coding the content of an OCP’s webpages, a content analysis of another OCP was conducted. The coding procedure was the same as the aforementioned steps. The previous coding results may be used as a reference to the new coding work. In turn, the new coding outcomes updated the previous coding results. Updates may include changing the names of some codes and sub-categories or adjusting their categories belonging.
Furthermore, to capture the final list of service measures, we merged the new coding schema into the previous schema. In particular, new codes or sub-categories would be added one by one into the global schema. This merge was conducted in a top-down way. First, we added new coding schema by categories, which means updating codes and sub-categories under a category at a time. Second, for each category, we examined if there were new codes in the new coding schema. If so, we inserted them into the sub-categories that it belongs to in the previous coding schema directly. If there are new sub-categories in the new coding schema, we merged them into the previous coding schema. Thus, it was an iteration process to gain the final coding outcomes.
4. Results Analysis
The service measures that are related to contest management, solver management, and requester management are illustrated as follows.
4.1. Service Measures Related to Contest Management
(1) Launch various contests
Contests act as the space where solvers conduct activities and gain benefits. The sustainable launch of various contests is a critical way to attract and keep solvers’ engagement. Most of OCPs provided varieties of contests while several OCPs especially “logomyway”, “Guerra creative”, and “48hourslogo” offered specialized contests, that is, logo design. Additionally, in terms of the number of new contests, the OCPs concentrating on complex/innovative problems averagely have fewer new contests and thus update less frequently than the OCPs which mainly focused on logo design, graphic design, and video production. For instance, on “eÿeka”, one to four new contests were published in a week, while 171 ongoing contests were presented on “Designcrowd”. Finally, all OCPs would send notifications to solvers once a new contest in which they may be interested was launched.
(2) Categorize and navigate contests
This service measure could facilitate solvers to search and choose contests in which they are interested. The criteria for filtering contests commonly consist of ongoing/completed, skills, industries, types, prizes, and contest levels such as base, gold, and platinum. 15 (60%) OCPs provide this service measure. However, the other 10 (40%) OCPs, including “Guerra creative”, “48hourslogo”, “Logomyway”, “Open innovability”, “Ideaconnection”, “Challenge.gov”, “Userfarm”, “Zooppa”, “GoPillar”, “Arcbazar”, just list all contests without assortment.
(3) Have a contest illustration framework
A structured framework for illustrating contests not only helps requesters to define their contests accurately but also helps solvers to make sense of the contests. This service measure is important and necessary because the clarity and accuracy of contest illustration have an influence on solvers’ willingness to engage and even submissions quality. All of OCPs develop a structured framework for contest demonstration. In general, context, problem, time scheme, total prize were mandatory to be illustrated in the framework, and requester identity and preferences, exemplars, submission evaluation criteria and jury, discussion board were optional and varied on different OCPs. Furthermore, most OCPs assigned their community managers/experts to review and improve a contest demonstration. At the same time, some tips or success cases of designing contests were offered to requesters by some OCPs.
(4) Manage submissions
It mainly includes management of submission submitting, evaluation, disclosure, and ownership transfer and use. Table 2
presents some specific features of this service measure.
As seen in Table 2
, all OCPs provided a structured guideline for solvers to answer requesters’ questions and demonstrate their submissions clearly. This kind of guideline always was specialized as a customizable input format.
For a specific contest, estimation of evaluation jury, process, and criteria is helpful for increasing solvers’ perception of contest justice and trust in requesters and understanding of how to develop submissions. However, only “eÿeka”, “Open innovability”, “HYVE”, and “Challenge.gov” requested requesters to publish a jury and its members who would be responsible for scoring submitted submissions and determining the wins. Several OCPs, like “eÿeka”, “Open innovability”, “Zooppa”, “Topcoder”, arranged their community managers or technical experts to review submissions first and then give suggestions to requesters before making final decisions. In addition, seven OCPs asked requesters to disclose a few evaluation criteria even a scorecard such as on “Topcoder”.
After a contest is completed, all of the OCPs would announce their winners. Nevertheless, not all OCPs disclose all submissions of a contest to the engaged solvers, even though it is an important way to embody fairness and for solvers to learn and gain skills. Specifically, four OCPs, “Open innovability”, “Herox”, “Challenge.gov”, “Topcoder”, never exhibited submitted submissions in a contest to engaged solvers or visitors. The other OCPs disclosed submitted submissions but with some preconditions. For example, “Crowdspring”, “Crowdsite”, and “Freelancer” disclosed submissions of a contest only if it is not private, which is set by requesters. Disclosure of a submission in a contest on “ZBJ.COM” and “EPWK” is up to solvers themselves.
Finally, regarding ownership of a submitted submission, all OCPs published their IPR policy which explicitly stipulates who owns solvers’ submissions and conditions of using those submissions. In general, solvers need to transfer their submissions to the requesters who have paid for them. Otherwise, solvers own the IPR of their submissions.
(5) Have a reward and charge system
Rewards and fees intimately associate with benefits that solvers can gain on an OCP and thus have their great attention. The monetary/physical prize and non-monetary prize are two major kinds of awards. As shown in Table 3
, all OCPs offered monetary/physical rewards. Non-financial incentives were introduced by 13 OCPs, these are creativity points and/or activity points. They were calculated based on solvers’ performance and behaviors including participated contests, number of submitted submissions, won submissions, and review scores by others. In practice, various featured terms were used by different OCPs such as reputation scores on “Crowdspring”, quality assessment scores on “Designhill”, and ideation and production points on “Tongal”.
Another feature refers to multiple opportunities for solvers to gain monetary/physical prizes. Fifteen OCPs suggested requesters set multiple winners in their contests. Most of these OCPs mainly concentrated on contests of generation and development of innovative ideas, concepts, and solutions. In contrast, the OCPs focusing on logo, web, and graphic design suggested requesters set one winner in a contest. Nevertheless, requesters on these OCPs can also set extra winners in their contests if they pay additional prizes. Except for prizes awarded by requesters, the OCPs including “Designcrowd”, “eÿeka”, “HYVE” may award several non-winners in a contest in order to praise their positive behaviors. For example, “HYVE” may select a couple of members as the most valuable participants (MVPs) and award them with money for their helpful comments, great and diverse ideas, or some kinds of special commitments. Finally, the affiliate prize acts as another reward source that a solver may get on some OCPs like “99designs”, “Designcrowd”, “Crowdspring”, “Crowdsite”, “Logomyway”, and “Tongal”. It is related to the number of requesters that a solver refers to launch a contest successfully.
Additionally, solvers may be charged by a few OCPs when they gain some prizes in contents. The charged fees include membership fee, service fee, and possible taxes. Specifically, two OCPs, that is, “Freelancer” and “GoPillar” charged some money if a solver tried to be a premium member who would enjoy more privileges than regular members. With respect to the service fee, it would be charged from solvers who won a contest or engaged in an invite-only contest. Commonly, the amount of service fee depends on OCPs, such as 15% of won prize on “Designcrowd”, 10% on “Logomyway”, “Freelancer”, 3% on “Arcbazar”, 20% on “EPWK”. The last kind of fee a winner was likely to bear is possible taxes and other possible costs that occurred during payment. It relies on winners’ geographic location and the payment method they adopt.
(6) Give feedback
The feedback from requesters, OCPs, and other solvers can promote solvers’ participation intention and even quality of submissions. All of the sampled OCPs suggested and supported requesters giving feedback against solvers’ actions and submissions through various approaches. They included ratings, commenting, discussion boards for publishing news and communicating with each other, declining unsuitable submissions, and even designing markup tools to draw directions on a specific submission.
To give feedback or not is decided by requesters. Unfortunately, some solvers who especially are not winners do not get any feedback from requesters. According to our investigation, only “eÿeka” and “Open innovability” guaranteed that solvers, and at least winners, can get feedback from requesters.
Another kind of feedback is given by OCPs. It always presents as community managers’ reviews, scores, or ranks of a solver’s submission. This feedback provides requesters a reference when they make final decisions on the winners of their contests. However, for a solver, it is rarer to acquire feedback from OCPs than from requesters in practice.
The last kind of feedback comes from other solvers or visitors. The OCPs including “Guerra creative”, “Freelancer”, “Innocentive”, “Herox”, “Tongal”, “Userfarm”, “Zooppa”, “GoPillar”, “Arcbazar” allowed solvers to give their feedback against each submission. This feedback includes votes/likes, commenting, or communicating in the discussion board.
4.2. Service Measures Related to Solver Management
(1) Offer various benefits to solvers
The OCPs offered solvers multiple benefits in order to satisfy their expectations and further attract and keep their participation. The main benefits appeared as the opportunities to earn money, practice skills, have fun, acquire lots of new work, help others, demonstrate expertise, gain reputation, develop career, merge into a community. Therein, the former five benefits are highlighted by the OCPs.
Different OCPs emphasize the benefits differently. For example, “HYVE” claims several kinds of benefits for solvers while a few OCPs just present one or two benefits. In addition, the OCPs, including “Guerra creative”, “48hourslogo”, “Crowdspring”, “Crowdsite”, “Freelancer”, “Tongal”, “Userfarm”, “Cad crowd”, “ZBJ.COM”, and “EPWK”, did not demonstrate benefits for solvers apparently. The benefits always are claimed using clear slogans or demonstrations on the OCPs’ homepages or the sector of “for solvers/creators/designers”.
(2) Have different participation channels
Two channels, that is, self-selection and invite-only, were always provided for solvers to participate in contests. The former is a predominant one, which refers to solvers selecting contests freely when they are allowed to enter. In contrast, a solver could register a contest only after being invited by requesters. For the second channel, there is another form called the one-to-one project. Requesters always invite solvers according to their skills, ranks, creativity scores, or activity scores. In some cases, OCPs would recommend a couple of suitable solvers to requesters based on their past performance, behaviors, and even membership levels.
(3) Develop a solver show system
The solver show system offers solvers a stage to display themselves on the OCPs. This system aids requesters and others to acquaint themselves with a solver and stimulates solvers to gain reputation or fame in the communities. In this study, we divide this system into three sub-systems that are the personal information system, solver rank system, and solver showcase system.
The personal information system was equipped by all the OCPs although they may be different in the presentation of a solver’s facets. In general, two types of information are presented in this sub-system. One is the solvers’ basic profiles including username, location, education, skills, work experience, member since, following, and followers. The other is the solvers’ activities and performance indicated by participated contests, submitted and won submissions, earnings, placement, and creativity/activity scores. Furthermore, some OCPs allowed solvers to run their shops with the services with specific price tags. These services demonstrate a solver’s preference, skills, and works they are capable of and good at.
For those solvers who are active and have good performance, they would be added to the top solver list. The top solvers on one hand could acquire more opportunities to capture requesters’ attention when they are going to invite solvers to contribute to their contests. On the other hand, they could get more benefits such as faster payout processing, prioritized support, and increased visibility. In addition, 15 OCPs, including “Guerra creative”, “Designcrowd”, “Crowdspring”, “Designhill”, “Logomyway”, “Freelancer”, “eÿeka”, “Innocentive”, “HYVE”, “Tongal”, “Cad crowd”, “Arcbazar”, “Topcoder”, “ZBJ.COM”, and “EPWK”, had a function of filtering the top solvers in terms of industries, creativity and activity scores, number of participation and won times, while the other OCPs just enumerate these solvers without assortment.
Of those top solvers, OCPs may interview some of them who have performed very well recently or in the long run. Six OCPs published success stories of those solvers as a way of praising them and stimulating other solvers. For instance, “99designs” developed the 99awards that showcase the best-of-the-best created by its talented and diverse designers from all over the world. “eÿeka” published creators of the month and shared their stories. “Tongal” presented an annual celebration of the brilliant people in the community and the outstanding creative works they have completed each year.
(4) Grade solver and serve them differently
On a few OCPs, solvers were assigned different levels based on two principles, that is, performance-based and member-based. For example, for the performance-based principle, solvers were graded into top, mid, and entry-levels on “99designs”, supernova, mega star, super star, rising star on “Guerra creative”, and level-1 to level-3 on “Crowdsite”. Under this principle, the number of won contests and submitted submissions, and rating scores are the major indexes to evaluate and estimate solvers’ grades. The other principle is member-based which means that the OCPs set grades of solvers based on their membership. To be a member, solvers are charged with some money by OCPs such as “Freelancer”, “Zooppa” and “GoPillar”.
Solvers with different grades own different rights. For instance, “99designs” rewarded solvers with high levels of additional benefits including faster payout processing, finalist payments, beta testing opportunities, prioritized support, increased client visibility across the platform. “Designhill” allowed pro-designers to participate in pro-contests, get a nice badge on their portfolios, and the first to test out exciting new features.
(5) Develop a support system
The solver support system was developed to support solvers to conduct activities. It comprises community support, technique support, and participation support.
Community blog/forum/social media generally has three functions: publishing the latest news, policies, new contests, stories of success contests and featured authors, facilitating communication among solvers and requesters, and presenting scientific articles and reports that are related to design ideas and skills. It may aid solvers in having a feeling of being a member of a community which is one of the motivational factors that stimulate solvers’ engagement and is also an important benefit claimed by some OCPs for solvers. All of the OCPs had their social media accounts on Facebook, Twitter, Instagram, or wechat. However, only 21 OCPs not including “Ideaconnection”, “HYVE”, “GoPillar”, and “Arcbazar” developed a community blog or forum for solvers. Further, only “ZBJ.COM” and “EPWK” developed an APP for solvers to participate and contribute to date.
Commonly, in the community blog/forum, some creative techniques, tips, and tricks were presented to solvers. These can assist solvers in developing submissions easily and effectively and train and improve their skills. In general, text, exemplars, or successful contests/submissions act as forms of tutorials provided for solvers. For example, “Crowdspring” presented some articles like “the 9 biggest packaging design trends of 2020” and “5 packaging design mistakes and how to avoid them”. “99designs” gave some examples of logo and t-shirt ideas in its designer resource center.
4.3. Service Measures Related to Requester Management
The OCPs spoke extensively about what and how they serve requesters on their websites. However, some of them were emphasized in the aforementioned solver and contest management and some, like providing different price packages for contest launch, are of little relevance to solvers service resulting in them being ignored in this study. We mainly shed light on the following three service measures that aim to ensure the trustworthiness, transparency, and justice of requesters.
(1) Disclose requesters’ identity
Disclosure of a requester’s identity is a key factor that may influence solvers’ perception of his/her trustworthiness and fame. Out of the selected OCPs, three OCPs, that is, “Designcrowd”, “Logomyway”, and “Arcbazar”, did not disclose any information about requester identity in contest illustration. A large proportion of OCPs told solvers requesters’ nicknames or ID numbers but with recent and overview activities including how many contests were held, review scores, last seen, and member-since. In contrast, a few OCPs like “eÿeka”, “Innocentive”, “HYVE”, “Zooppa” noted that disclosure of requester identity is helpful for promoting solvers’ participation. Consequently, the identity of requesters and what they have done related to the contests were always presented on these several OCPs. To some degree, the OCPs who disclosed requester identity were apt to serve requesters who are big or famous companies, while those who did not do this tended to serve small and medium-size enterprises or even individuals.
(2) Authenticate requesters
Requester authentication aims to verify the authenticity of requesters. Some OCPs required requesters to authenticate their email address, telephone, and even business license. These authentication items would be marked in their launched contests. In addition, solvers can also make a judgment against a requester from the disclosure of his/her identity and OCPs’ declaration about what kind of requesters they serve. In general, the OCPs such as “eÿeka”, “Innocentive”, and “Challenge.gov” always served large companies, progressive and innovation-driven companies, government agencies, and nonprofit organizations. The OCPs focusing on interior design and graphic design especially logo design always served small businesses, startups, and individuals. Engaging in contests held by big and famous companies or government agencies not only makes solvers have a feeling of security in the virtual environment but also stimulates them to gain reputation and acquire certification for future development.
(3) Require prize guaranteed
This service measure intends to ask requesters who plan to launch a contest to pay contest’s prizes to the OCPs before it opens to solvers. Solvers receive prizes from the OCPs if they are selected as winners in a contest. Generally, there is a label of “guaranteed” on the contests which have been paid in advance. Such a measure is helpful for avoiding a cheating situation of requesters delaying or even not paying the prize after a contest is completed, and lets solvers participate in a contest without hesitation. All of the OCPs suggested requesters set their contests as guaranteed contests and most requesters followed this request according to our observation on their websites.