VALUE CREATION IN INNOVATIONS CROWDSOURCING. EXAMPLE OF CREATIVE AGENCIES

. Innovations are crucial for most of the companies to survive. However, the concept of innovation has become broader, including new forms of open innovation, such as crowdsourcing. The aim of this paper is to define the business model of a crowdsourcing-driven organization to create value. Empirical research consists of case studies on current crowdsourcing platforms, focus groups with potential crowd members and in-depth interviews with potential customers of creative agencies. Best practices were combined with solutions for closing the most significant gaps in order to create a successful business model. The developed model suggests separating the crowd into free users and an empowered core team and enabling collaboration. Moreover, an innovative motivational model is introduced. Due to a three-step sequence of solution/idea generation, superior value is proposed to the customer. Another competitive advantage should be flexibility and adaptability to the customer’s needs. The paper is original since extended analysis of all crowdsourcing stakeholders is delivered. It also has practical value proposing a business model for creative agencies.


Introduction
Innovations are crucial for most of the companies to survive. Despite several a empts to search for innovations in public, almost all companies are stuck with the rst-mover advantage. However, Lee, Olson and Trimi (2012) state that the understanding of innovation has become broader. In 2003 Henry Chesbrough came up with a concept of open innovation, which states that by sharing their internal knowledge companies could bene t not only nancially, but also boost their knowledge base and accelerate development of own products. Lee et al. (2012) also promote the next level of innovations: "co-innovation is a platform where new ideas or approaches from various internal and external sources are applied di erently to create new value or experience for all stakeholders, including consumers" (Von Hippel et al., 2011 as quoted by Lee et al., 2012, 824 p.).
In 2006, Je Howe came up with the term "crowdsourcing" by combining outsourcing and the crowd. It means outsourcing of the work to the crowd, who would volunteer to perform it in exchange for compensation. Crowdsourcing is a narrower term compared to open innovation or co-innovation, as the la er two encompass any in ows or out ows of innovation in any way, crowdsourcing focuses more on in ows from e orts of single individuals or small groups. In combination with Web 2.0 technology, which enables information to be transferred both ways among many individuals or small groups, crowdsourcing may have cost e cient practical implications.
Even though crowdsourcing has its niche in contemporary industries, it is not very popular due to several reasons. First, crowdsourcing has erroneous perceptions: as cutof-costs activity -even if it is true in some cases, it is not the main focus of creativity aimed crowdsourcing (discussed further in the paper); or as public relationship (PR) campaign -absolutely vital among participants from developing markets, for which it is a brand new phenomenon. Second, it is not equally easy to implement crowdsourcing for an unknown small to medium sized enterprise (SME) or even a larger company acting in emerging economies compared to a well-known large, usually western, corporation. And third, crowdsourcing does not have a well developed model which could create the highest value to all parties involved. Moreover, creative agencies tend to ignore crowdsourcing or, even worse, see it as a threat, not as an alternative to their current business model. e problem of this paper is how to create the value by crowdsourcing innovations for the customers. erefore, the aim is to de ne a business model of a crowdsourcingdriven organization to create value.
Literature sources focus on the development and forms of innovations (Lee et al., 2012;Duarte & Sarkar, 2011;Russo-Spena & Mele, 2012), and motivation to be involved in open innovation communities (Antikainen, Makipaa & Ahonen, 2010). Since the eld of open innovation, co-innovation, crowdsourcing is still fresh, there is li le of empirical research conducted. Some examples involve a survey of innovation intermediaries in France, Netherlands and Finland (Antikainen et al., 2010), or European companies adopting open innovation (Schrol & Mild, 2011). However, there is a lack of research involving all the stakeholders of open innovation, namely crowdsourcing, lack of focus on creative agencies in the eld of value creation for customers. is paper is unique in providing a business model for a crowdsourcingdriven organization. e intended business model is one of the rst a empts to suggest a multi-directional value ow depiction in crowdsourcing initiative run by a dedicated company by combining literature suggestions, best practices and unmet expectations of stakeholders. e model (or part of it) is expected to be de ned quantitatively before its application in an actual venture.
Emerging economies usually lack funding for innovations, therefore conventional forms of innovation and development struggle. ere is a need to implement new innovation harnessing techniques as addition to conventional ones, but not as their replacement. As an example, a few decades ago South Korea was the best on reverse engineering and now its products sometimes surpass western analogues. Bearing in mind Eastern European (including Lithuanian) experience, diverse thinking pa erns and educational background, crowdsourcing may become another success story. On the other hand, due to modest quantity of possible crowd members, crowdsourcing would be more e cient if it was concentrated and managed by few dedicated entities.

Literature review
Literature overview consists of crowdsourcing related issues raised by various authors.
ose issues are later on combined into one pa ern used to evaluate crowdsourcing based platforms currently available in the market (see Chapter 3).

Open Innovations and Crowdsourcing
Innovation is usually perceived as a positive change in the organizational status quo, therefore improving one or more of its strategic elements (e.g., Luecke & Katz, 2003;Baregheh, Rowley & Sambrook, 2009). Believing that current innovation strategy goes down in its power, Chesbrough came to the concept of open innovation in 2003 (as cited in McKay, 2010). According to Chesbrough, the main reason of open innovation to be employed is that not all innovators work in the company, therefore external R&D might bring additional value to the company and new ways of harvesting it should be sought.
In 2006, Je Howe came up with the so called buzzword "crowdsourcing", a type of open innovation. He de nes it as "the act of taking a job traditionally performed by a designated agent and outsourcing it to an unde ned, generally large group of people in the form of an ''open call" or simply as "application of Open Source principles to elds outside of so ware" (Howe, 2006). Study on crowdsourcing success by Sharma (2010) gives ve critical factors to a ract participation in crowdsourcing initiative, which are shown in Figure 1.
According to Sharma's (2010) model, the success of crowdsourcing depends on motivation of the crowd. Motivation is built by ve factors: crowdsourced solution shall be given; collaboration (including other users/contest participants and administrators of the contest); 75 in terms of respect, liability and ethical issues; in uenced by the involved parties. e success itself also works as an additional factor to increase the motivation of the crowd. Figure 2 represents the parties which could be involved in crowdsourcing. Only two participants are necessary for crowdsourcing -the crowd and the customer (or a purchaser of the crowd's knowledge). However, some individuals could be separated from the crowd due to their speci c abilities to form a closed core team. e customer also has a possibility to choose whether to engage in crowdsourcing activities or  outsource it to a subsidiary/external company. us, the maximum number of e ective crowdsourcing participants is equal to four. Despite the number of involved parties, there are several issues to be assessed while running a crowdsourcing initiative. Among those are: Crowdsourcing om the Perspective of the Crowd Participation of the crowd (and the core team, if applicable) is essential for crowdsourcing, but it should also meet three main requirements to become an e ective tool (Trends E-magazine, 2009): information, submi ing their solutions and pu ing all the diverse ideas together; given.
Ensuring diversity e success of crowdsourcing depends on successful a raction of the critical mass of diverse participants. e crowd may be formed of groups, usually according to their a itudes towards or relationships with the purchaser of the solution (Palumbo, 2009). However, collaboration in a form of discussion is encouraged among these groups. To measure the possible value of group collaboration, Metcalfe's Law could be used, which states that the value of a network increases exponentially for every n-node added to the network (Shapiro & Varian, 1999, p.184).
Making ideas handling easy e most valuable crowdsourcing feature is ability to transcend geographic, political, economic barriers (Sharma, 2010). is enables creation of cross-functional teams in a broad scope of problems. However, diverse cultural backgrounds lead to longer time needed for the nal decision to be taken (Way, O enbacher & Harrington, 2011). erefore, platform's infrastructure should work as a tool for the crowd for retrieval, submission and aggregation of information to compensate the time used for taking decisions. "Rules of the game", including submission guidelines and intellectual property protection, should also be presented clearly (Drummond, 2011).

Incentivizing participation
Motivation and incentives for the crowd's participation should be considered very carefully. e crowd should be perceived as a partner, therefore strong connection between the crowd members and those who conceptualize suggested ideas should be built (Sharma, 2010). Motive alignment study of participants in the SAPiens Ideas Competition (Leimeister, Huber, Bretschneider & Krcmar, 2009) resulted in a comprehensive categorization of motives and incentives for the crowd (Table 1).  Antikainen et al. (2010) also raise an issue of rewarding for groups versus individuals, since until lately the major focus was put on the la er.

Crowdsourcing from the Perspective of the Company
Companies could engage in crowdsourcing due to various reasons. To check whether crowdsourcing exists among the necessities the company should answer the following questions ( Jouret, 2009): solve it? ose answers of the company originating from the emerging markets most probably would have "Yes, but…" element. en Jouret (2009) suggests another set of questions to determine whether to engage in direct crowdsourcing, or outsource it: What could be o ered as an incentive for the participants?
handling of submissions? Is the company ready to take risk of possible copyright infringements? Only few companies acting in the emerging market could answer positively to the questions above. is leads to a clear need of an intermediary where crowdsourcing activities could be outsourced to.
Crowdsourcing from the perspective of the creative agency as an intermediary e creative agency may be de ned as an organization which creates intellectual property (IP) for pro t. Due to human resources limitations, creative agencies usually focus on a speci c industry: e.g., Marketing/Communications, Web Design/ Development, Multimedia. However, the creative agency could also be perceived as an organization with a main goal of creating new knowledge by using knowledge management principles. To accomplish such transformation, creative agencies should change a itudes towards crowdsourcing, focus on sustainability and ful l certain obligations for a business model. In that way, crowdsourcing is expected to become more professional with business-like outlook between the customer and the crowd (Parpis, 2009).
Growing popularity of crowdsourcing is seen as a threat for current creative agencies, because crowdsourcing practices are developed the best in the same industries creative agencies are working in (Winsor, 2009). However, John Winsor (2009), as CEO of an advertisement agency based on crowdsourcing principles, believes that all professionals should employ crowdsourcing as a tool which pushes creative agencies to transform current and develop new business models. Moreover, usage of crowdsourcing is a desired feature of the customers who would like to take part in strategy formation of their beloved companies (Noam Buchalter as cited in Murphy, 2009).

Research Problem De nition
Since the main problem with crowdsourcing is disability to create value in the way it could be main revenue stream for the company and become a regular way in the market for obtaining innovations, the research is aimed at creation of the business model for a crowdsourcing-based company to create value by combining best practices and addressing unmet expectations from all the parties involved.

Selection of the Model Framework
e paper is based on the business model canvas suggested by Osterwalder, Pigneur, Smith, and 470 other practitioners (2010). It is worth mentioning that the business model itself is neither questioned as a concept in this paper, nor is it intended to present the selected as the most appropriate one. Alignment of the critical factors of crowdsourcing success proposed by Sharma (2010) (Figure 1), with the business model canvas by Osterwalder et al. (2010) provides a foothold for our research model (see this combination in Figure 3). e essence of the combination is that human capital is treated as partners, and activities are held to build the trust. Infrastructure is the platform itself, which is also treated as a resource and the main collaboration channel. Vision and Strategy re ects the value proposition for both: the users and the purchasers of the solution. e di erence of cost and revenue streams, to be more speci c, the di erence between positive and negative cash ow could be de ned as success of crowdsourcing.

Human Capital
Human resources consist of in-house employees and the crowd. On the other hand, in crowdsourcing it is critical that the crowd is visualized as a partner. Since the crowd has diverse skills, abilities and the current level of professionalism, it could be split into two or more levels, where di erent roles of users are authorized to use di erent features of the platform. e possibility to collaborate between levels or among teams, if these are applicable, should be evaluated in terms of e ectiveness as well.

Linkages and trust
Crowdsourcing should be a fair game. Clear information about odds of winning, selection criteria should be stated. Another very important issue is intellectual property protection. Rights and ownership of work a er the submission and a er the end of contest should be clearly de ned. Support of well-known corporate stakeholders gives more trust to the users. Government is perceived as stability warranty. Any association with or support from governmental institutions is more likely to add trust. Previous success stories or testimonials also add some trust elements.
In astructure e organization of ideas is represented by the platform. e major focus shall be allocated to presentation of primary material, which is used by crowd for elaboration on ideas. In general, the platform should be user-friendly and capable to process and evaluate submi ed ideas. From the perspective of the purchaser of ideas, additional capabilities, like an environment for an iterative approach to service innovation, are also important.

Vision and Strategy
e best disclosure of value proposition is mission, vision and objectives statement. However, here comes value for two parties, the crowd and purchasers of the endproduct. e community requires an incentive as a reason to participate. is includes needs, aspirations, motivation and must remain the most important consideration while developing the crowdsourcing initiative.
Value proposition for companies concerns newness and customization, design advantages and cost reduction, and convenience. e main idea is to provide a lowfriction, cost-e ective environment for collaboration.

External environment (purchasers)
e purchasers are concerned about external ideas incorporation into the strategic planning process. However, Make/Buy/Partner decision is made by evaluating internal capabilities to exploit crowdsourcing: lack of in-house professional team for ideas supply or any network with suppliers, vendors, competitors or inventors minimize the likelihood of Make decision; the scope of crowdsourcing usage and that the company o ers in exchange for ideas (cash, workplace, long term partnership) usually determines either Buy or Partner decision.
Success e success of crowdsourcing depends on participants and their willingness to share ideas. e essence of the business model, however, is to get the maximum for the concept from the purchaser and to pay as li le as possible to the crowd. e main cost lines of the platform include cash incentives, platform acquisition and the company's maintenance costs.
ere is no possibility to nd out exactly how much the platform owners get from the project. But it is possible to identify the revenue model like one-time customer payments or recurring transactions due to post-purchase customer support, or both. ese details could be used to determine pricing strategy, whether it is value driven or o ers only cost-saving possibilities.
Methodological Approach e nature of the research object determined qualitative empirical research. Deductive approach is conducted rst to match the pa ern suggested by literature review with that currently available in the market. Later, the importance of retrieved best practices is tested with direct stakeholders of crowdsourcing initiative. Moreover, the parties involved are surveyed to extract either already satis ed or still open needs and expectations of potential customers (Table 2).

Case Studies
Case studies are conducted in order to compare the pa ern of activities obtained from the literature review with the ones currently used by crowdsourcing platforms. e aim of this research part is to extract current best crowdsourcing practices and to identify areas for improvements. e crowdsourcing platforms selected for case studies have di erent approach and di erent target audience: awards); comments; Case studies on selected crowdsourcing platforms are held by lling in prede ned research questions matrix (see Table 3) which is prepared according to the selected business model framework (see Figure 3)

Focus groups
Focus groups are aimed to rank practices extracted from literature and case studies (therefore are conducted a erwards); to discover needs and unmet expectations of possible platform users. Time and access limitations led to the selection of convenience sampling to form the groups: directly responsible for improvements of the processes related to nancial operations; 12 in total addressed with an invitation, 4 responded; the main aim was to evaluate a itudes of internal employees towards crowdsourcing; authors as creative and innovative individuals (co-workers or collaborates in some way); 8 in total addressed with an invitation, 2 responded; the main aim was to identify value creating activities for potential participants in crowdsourcing initiative; by the initiative manager as most suitable for the focus group; the main aim was to clarify the needs of platform users for solving NGO's problems and acquirers of NGO tailored solutions. Participants are provided with suggested topics, the objective, and the structure of the discussion. Questions were raised (see Table 4) to re ect the key business model framework areas which were not disclosed to participants.

In-depth interviews
In-depth interviews are performed to identify current and potential uses of crowdsourcing in di erent companies/organizational bodies as well as determine possible ways of collaboration, to discover their needs and expectations. Convenience sampling is used for the selection of respondents. However, di erent companies in terms of their size and activities were selected: which innovations were the most important competitive advantage and crucial to survive; research and analysis company, for which innovations were the way to ensure their market share, but these are shared globally; governmental institution, for whom improvements and consultations with society were a duty. In-depth interviews questions were designed to address Customer Relationships and Customer Segments elements of the business model framework mostly, as these were not assessed in previous research steps. Suggested topics (Table 5), the objective, and structure of the discussion were provided in advance.

Results of the Platforms Case Studies
Platforms case studies were held in order to compare the pa ern of current activities with the one obtained from literature review, to disclose existing gaps. e results (see summary in Table 6) are presented in the same logic as the questions were ordered in the questions matrix (see Table 3).

Human Capital
Well known platforms are capable to a ract thousands of users (HC1.1 -see categories in Table 3), but in most cases these fail to make them collaborate. Metcalfe's law states that potential of collaboration grows exponentially with every additional node, but with limited interaction between nodes it becomes useless (HC1.2). On the other hand, this emphasizes quantity rather than quality, because lots of participants create huge amount of information, and simultaneously a problem of processing it e ectively. Another problem is useless mind ow (HC1.5) in the platforms due to lack of proper user and submission tracking. Some platforms are very easy to join, like Dell's IdeaStorm, which creates overload of users and information.

Linkages and Trust
In all cases except Dell's IdeaStorm, participants try to meet the demand from the customer, not to supply ideas despite the demand (LT1.1). As it is shown by example of InnoCentive or TopCoder, the more complex the challenge is, the be er criteria of selecting the winner are described (LT1.2). All the examined platforms act only as a medium for collaboration between the solution purchaser and the crowd and give almost no e orts in creating any other value by them.
All platforms propose quite adequate direct rewards for the solved challenges (LT3.1). And the last thing causing trust in the platform is participation of governmental and non-governmental organizations (NGO) as customers of the project (LT4.1).

In astructure
Ease of use correlates with the number of possible functions and primary material presentation capabilities (IN2.1, IN2.2, IN2.3). However, there were some well-balanced examples like CrowdSPRING or InnoCentive, and some imbalances, like IdeaBounty. Presentation of primary material is mostly based on "text only" principle, only part of platforms has capabilities for a achments or presentation of visual content (IN2.3). Crowd-SPRING simply gives a pre-de ned template according to challenge type for the customer to ll in and posts it without edition. Evaluation is put outside the platform (IN3.2), therefore it makes participants guess the real dimensions of the evaluation (IN3.1) and also wastes resources of human capital, especially when mind ow is intensive.

Vision and Strategy
While talking about platforms as a source of experience (VS1.1), IdeaStorm or Victors&Spoils have some practices in place. Mentors, if available, guide towards a particular way, unfortunately, they do not help with individual feedback (VS1.2). And there is no feedback from peers (VS1.4), except Dell's example. Having very limited feedback, it is hard to say that platforms are acting as incubators (VS1.3), where good ideas can grow.
Monetary prizes are a ractive enough (VS2.1). In most cases the winner is a single person or a group. Non-monetary prizes are also available (VS2.2), but, in the same way as career opportunities (VS2.3), are rarely used in practice. Still, Victors&Spoils platform suggests the model of making a career within the platform.

Success
All platforms present themselves as value driven (SV1.1) and propose an award according to value-in-use pricing. E.g., IdeaBounty insists that it is working on a free market principle, that the more signi cant the award is, the be er are ideas to choose from. e cost of mediation depends on the platform (SC2.1), varying from xed rate plus 15% of the award to 100% of the award. Victors&Spoils employs an option where the project is owned by the intermediary and only part of the money for winning ideas is shared with the crowd. None of the platforms provide post purchase support (SC2.2).

Aggregation of the results of the case studies
To sum up, even though platforms are orientated to di erent target audience, the means they are employing do not di er a lot. Users are treated as individuals, not as an integral source of knowledge, therefore their collaboration is minimal. Incentive models do not vary much, either, and lack of trust building activities usually lowers the e ect of motivational options applied. Although capabilities of platforms are di erent and represent o ered services quite neatly, they have a common drawback -knowledge submi ed by the users is not processed by an intelligent creature. Figure 4 shows generic knowledge exchange model currently available in the market. e elements distinguished by dashed lines are seldom used. e major downside of the current business model observable in the market is that knowledge which goes through it is not enhanced in any other way than useless information ltering. is means that the intermediary acts only as a collector of individual thoughts of the crowd members, which are passed to the customer "as is". Nothing is done inside the platform to aggregate, summarize or evaluate this knowledge, the knowledge is not even a empted to be transformed to the commercial product which could be sold to the customer a erwards. Table 6 summarizes the key ndings of case studies and identi es the observed gaps from literature suggestions.

Focus Group Research Results
e aim of this research part is to evaluate the signi cance of practices extracted from literature and case studies, and discover real needs and unmet expectations of possible platform users. e results (Table 7) are presented by pre-de ned categories: rst goes motivation, the second one is risks and obstacles which decrease the motivation, then expected "rules of the game" are presented, and the last one is required technical capabilities of the platform.
Motivation e main motivational issue was that youth values experience (including feedback and career opportunities), older people are satis ed with acknowledgement, but money retains relevancy as well (Group 3). First of all, any commercial implementation of the idea is expected to be rewarded with cash or any other direct and tangible bene t. Direct motivation could also be given in a form of salary, if the user submits a certain number of ideas for a pre-determined period of time (Group 2). e other motivation option is freely gained experience. It could be expressed in several ways: sharing of perspectives through the eyes of a worker from another industry (Group 3), a possibility to use the skills one has, and a possibility to implement the idea with professionals.
Feedback from experts is perceived as necessity, especially for students (Group 1). Feedback from peers, however, might be very subjective, but some kind of discussion would be valuable.

Risks and Obstacles
e most common problem in crowdsourcing is willingness to own ideas even if one does not have any possibility to implement them. e risk of ideas the or loss prevents people to share ideas publicly or even privately with a potential investor. One of the possible solutions how to deal with the IP concerns is to make the contest in several levels (Group 1). As the rst step, all ideas could go public, but those are just raw material for real projects. In the second step submissions could be private and all participants should get some symbolic appreciation from the host of the idea sharing platform (Group 2).
Since people are participating on a voluntarily basis, lack of responsibility is always present. Without proper motivation, crowdsourcing may be perceived as a cheap way for companies to get ideas (Group 1).
To ensure capabilities for assignment, the platform should have a clear segmentation. Only people with expertise make the discussion e ective. However, fresh ideas are always relevant, thus students may be employed. From the platform's perspective, the customer should give clear criteria to the intermediary (the platform owner), who on his own behalf should prepare a task for the crowd (Group 2).

"Rules of the game"
First of all, challenges should be separated according to the expected nal result. Crowd is more interested in creative, not technical challenges -the technical ones should be le for the experts of the eld (Group 1). Users are about to be separated to segments (Group 1). Access limitations should be applied by spli ing the contest into several stages (Group 2). To prevent premature loss of interest, access to the rst stage of challenges should be granted for all, and di erent ways to get to the second stage (as a short-term incentive) should be ensured.
Technical capabilities e platform itself should be easy to use and easy to join. On the other hand, access restrictions should be employed to prevent anybody and everybody from entering and ruining the competition (Group 3). e task should be presented in a clear structure to save time and prevent misinterpretations. If the task is good, the answers/solutions will be good as well (Group 1). e company's capabilities to implement the nal product should be stated to avoid unreal suggestions (Group 3), and needless e orts.
As an additional revenue model, traditional approach of ideas bank should not be discarded. Companies may be paying for the access to that knowledge base as well (Group 2).

Aggregation of the results of the focus group surveys
To sum up, cash is still a really relevant motivational option for the users of the platform. Other options of incentives are barely addressed in the current market: experience gaining in the way of feedback, knowledge transfer, mentorship and acknowledgement. Having this in mind, the desired value ow among the users of the platform is shown in Figure 5. is chart completely represents the element of Value Propositions of the business model for the users of the platform. e value ow was designed by making assumptions that users from di erent levels are collaborating between each other and provide mutual value. Due to collaboration among peers of the same level, value is created and shared within this level as well. e intermediary's activities towards its users are aimed in the way that all value goes to the users in exchange for fresh ideas. e core team gets more value to become a desirable target for the crowd's members to pursue.

In-depth interviews results
e main objective of the interviews was identi cation of current and potential uses of crowdsourcing in represented companies/organizational bodies and discovery of possible ways for collaboration. e results are presented by categories in a particular structure (see Table 5).

Value of Innovations
Small companies acting in niches or by a non-conventional business model usually live from innovation or are innovations themselves (Interview 1). Approach of large companies towards open innovations is problematic. R&D departments do not tend to share any part of their knowledge with outsiders due to the possibility or underlying games by competitors (Interview 1).
Innovations and improvements in governmental solutions are driven by expectations of the society. Various commi ees are formed by members originally working in consulting or similar business, public associations, etc. Members are selected according to the purpose of the commission and are working on a voluntarily basis; they get no other than emotional compensation, e.g., acknowledgement (Interview 3).

Current Uses of Innovation Sources
Small to medium sized companies are really good at sharing information inside the company, because it is crucial for them and it is easier with a smaller number of employees. A forum-like on-line discussion place is perfectly suitable for such type of a company. However, such companies use secondary data and rarely are involved in any public request for primary material, designed specially for them (Interview 1).

FIGURE 5. Desired value ow between user and intermediary
Source: created by the authors e most common application of crowdsourcing in large companies is the usage of small groups of selected people, usually customers, to get some ideas or test the market options. As a platform, social networks like Facebook or tiny applications in own site, are used. From the company's perspective -one cannot create a suitable nal product without knowing enough internal information which is con dential.
ere are three key sources for suggestions on new legislative projects: individual expertise of a regular employee, various commi ees and boards, and collective knowledge from society (Interview 3).

Partnership Possibilities
For small companies outsourcing means signi cant cost savings, because they give away considerable part of their internal labour force (Interview 1). Large companies are also keen on outsourcing when it is more e ective than having an in-house specialist (Interview 2).
To become a partner of for-pro t corporations, crowdsourcing intermediary should have a clearly de ned market in terms of industry and target audience to: (1) have highly skilled insiders (or outsiders) that are able to prepare the nal product for the customer, (2) be able to propose real innovations and gain trust of the customers R&D department (Interview 1). Figure 6 presents a generic pa ern for the desired value ow between the intermediary and its customer. Despite higher amount of value streams from the intermediary to the customer, there is only one intentional -Fresh and Professional ideas. Others come from the nature of crowdsourcing unconsciously. On the other hand, it is not always all three inbound value streams coming to the intermediary. Intermediary-crowd value exchange is shown in Figure 5. FIGURE 6. Desired value ow between intermediary and its customers Source: created by the authors As the presentation of the results show, the needs di er in all the three segments; therefore Table 8 gives rather a summary of the interviews ndings than an aggregation. Discussions e research results show clearly that the expectations of users and customers are far from being met by the existing platforms. ese also fail to implement signi cant part of tweaks proposed by scholars in literature. e major gaps not closed yet are as follows:

Aggregation of the results of the interviews
1. collaboration among the users of the platform and constructive feedback; 2. value creation in the intermediary internally to create a professional and complete nal product; 3. empowerment of certain level users, thus giving them additional motivation; 4. involvement of the idea author in further development of the idea; 5. growth of users' competence and career possibilities; 6. facilitation of governmental duties to build the trust; 7. honesty and clearance of the "rules of the game"; 8. low quality of primary material given for the crowd; 9. missing general capabilities to evaluate idea inside the platform; 10. pro ling and networking possibilities; 11. usage of other than regular WWW channels to increase time spent on the platform; 12. usage of all media (not text only) for communication; 13. dynamic proposals for customers to meet their needs; 14. supply rst approach is not considered as additional service; 15. other than cash incentives are under-evaluated; 16. barter market with customers, especially not-for-pro t ones. As it is shown by the Generic Knowledge Exchange Model (see Figure. 4), all existing platforms of crowdsourcing are basically a space where the purchaser meets a solver, which makes them simply a next generation of web forums, but not a real innovation tool. To create a successful business model for the crowdsourcing intermediary, aforementioned 16 open gaps are mapped on the business model framework (see Figure 7). erefore the business model for the crowdsourcing aimed creative agency is created (see Table 9) in brief; a comprehensive model is provided in Appendix A. e model accompanied with key resources (the platform and internal employees, further referred to as internal HR) is capable to provide superior value for both -the customers of the intermediary and the users of the platform, or the crowd. However, Internal HR salaries; Lawyers in case of con ict; Sales & marketing of the platform.

Revenue Streams
Value-in-use pricing. Extras cost extra.
Fine and kept advance payment for dropping the project Direct advertising on the 1 st level of contests.
Fees for "branded" space, and browsing through ideas bank & portfolios.
Barter market (especially with NGOs) Source: created by the authors the main challenge is not only to develop the platform and nd appropriate employees, but also educate the customers so that crowdsourcing might create a satisfying nal product, equal to the product outsourced from any other company or even be er. Further in this chapter, each part of the model is described in detail.

Key Partners
First of all, users should not be treated as a resource as it is common currently. Users provide the most valuable resource for the company, therefore they should be considered as key partners. To provide a be er organization of users, they should be separated into two groups -free users, or rst level users -anyone who joins the platform; and the core team or the second level users, who are carefully selected from the rst level users by internal HR to help in elaboration of primary ideas to the nal product. erefore, each product goes through two stages -brainstorming among the rst level users and elaboration of the selected ideas in the core team. e nal product is generated by internal HR using suggestions of the ideas selected by the core team. While rst level users ensure diversity of the crowd, second level users perform higher quality work. More capabilities should be given to the core team members: 1) empowerment: they are acting as moderators and mentors simultaneously; 2) possibility to team-up with other members.

Key Resources
Two types of property are considered as key resources -its internal HR and the platform itself. Internal HR is really valuable, therefore an expensive resource. First, internal HR is responsible for preparation of the primary material for the rst stage of the idea generation process as well as guidelines for the core team in the second stage. Another responsibility of internal HR is to nalize the product for the customer, if necessary. e platform itself does not cost much while in use, but initial investment is quite signi cant here. e platform should clearly de ne the users' roles and responsibilities. It should also provide di erent ways of primary material presentation as well as transcendence of collaboration barriers.

Key Activities
Very basic requirement for trust is insurance of IP protection. First it should be done technically by preventing the and regulating work of the search engines' bots. Another means is to prepare legislative agreements of juridical power.
Another requirement for building the trust is honesty. at one obliges clarity of the"game" rules. Conventional terms of use should go together with a simple explanation how the platform works and how submissions are handled in terms of IP.

Value Propositions
Ability to choose is a motivation itself. erefore, each user should be capable to rank his motivational options in the pro le. Possible motivation options should be di erent for free users and the core team.
Customers, on the other hand, also have some motivation options, which stimulate their choice of the company as an external service provider. ey get not only a lower costs product, but the possibility to choose from a wide range of "fresh" ideas.
For be er understanding of value ows, a joint model of those desired by potential users ( Figure 5) and desired by customers of the crowdsourcing intermediary ( Figure 6) is presented in Figure 8. Moreover, since advertising in the rst stage of competitions is approved, additional revenue stream from third parties is marked with do ed arrows.
e success of business model could be very simply explained by looking at this value ow chart: the company will make pro t as long as inbound value ows exceed outbound ows in terms of cash, and this di erence is higher than the costs of the company maintenance.

Customer Relationships
As mentioned previously, the nal product given for the customer is comprehensive and professional. Interim reviews, where the customer participates in the selection of ideas at the rst or second stage, could be held upon request of the customer. Another virtue of the platform owner is to make dynamic value propositions according to the needs of the customer. To enable it, the platform itself should have customization capabilities. One tough task is to promote the crowdsourcing industry itself. Despite current crowdsourcing approaches, companies still tend to create value internally only.

Customer Segments
Due to dynamic propositions, the intermediary should be able to serve various segments of customers: from occasional users with the need of very basic service to governmental institutions with large set of regulations and other requirements to be ful lled. Another group of customers is small companies which are looking for nonconventional solutions. Such companies may look for primary data instead of secondary data they usually base their strategy upon. e last signi cant segment is governmental institutions and NGOs.

Channels
In addition to a conventional text based on-line communication, audio and video media could be used. Moreover, interactive media, e.g., simulations, could be used in order to let "touch the model", but not disclose the commercial secrets. Di erent access capabilities, like smartphone applications accompanied by mobile internet access would increase the time spent on the platform as well as the number of active users.

Cost Structure
Main costs of the intermediary consist of acquisition and maintenance of key resources. First, development of the platform itself would be a major part of initial investment. Also maintenance of the platform would be a part of variable costs. e biggest part in variable costs would be for internal HR salaries and other incentives. On the other hand, dynamic nature of external human resources (the crowd) allows cu ing the costs signi cantly, as it is relatively easy to adjust its size to existing demand. However, cash compensation to the core team should be considered anyway.

Revenue Streams
Value-in-use pricing approach would be used in the model, still keeping the space for cut of costs. Besides a regular challenge price, the customer could be charged for extra services where some part of work should be outsourced by the intermediary. e government could be charged exclusively for feasibility studies and building of external commi ees. NGOs most probably could engage in barter market -services for them could be provided in exchange of internal knowledge, which is used later to build internal knowledge base or for the advertising space of the platform.
In order to diversify revenue streams, both demand (purchaser asks) and supply (crowd o ers) based crowdsourcing approaches are employed (see Figure 9). Moreover, companies are encouraged to share internal knowledge in exchange for discount or similar bene t. By using this knowledge, internal knowledge base is supposed to be developed and used as incentive for the users of the platform in that way saving costs on cash incentives. Compared to forum-like knowledge exchange model (see Figure 4), the developed business model is a way more stable than these currently available in the market.

Conclusions, Limitations, Further Research Topics and Practical Implications
Crowdsourcing gained popularity due to favourable circumstances present in the market. Global cut-of-costs policies, reached limits of traditional business models and demand for non-conventional solutions led companies to search for alternatives. Due to the nature of crowdsourcing, not all companies are capable to engage in direct application of it, so crowdsourcing intermediaries jumped in the market. However, these are struggling with further development due to absence of a business model which would enable value creation, while innovations are crowdsourced by intermediaries for their customers. is laid back approach of crowdsourcing intermediaries not only leads to under-use of crowdsourcing in western business environment, but also is the reason why the word crowdsourcing itself needs additional explanation for most entrepreneurs in emerging markets. e results of empirical research have shown the major gaps with expectations and best practices of current crowdsourcing intermediaries. By combining the research results, the best current practices and those suggested by literature, a comprehensive business model for the innovations-aimed crowdsourcing-driven creative agency is developed. Since the aim of the paper was to create a successful business model, the success is de ned by the ability to capture more value in terms of cash than it costs to maintain the company. In other words, the di erence between inbound value ows and outbound value ows should exceed the total costs of the creative agency. e main limitations of the paper exist due to the type of the survey. erefore since in-depth interviews were held with only one individual from a separate segment, generalization of the results should be made with caution. On the other hand, qualitative research serves as a foundation for further quantitative research to obtain certain values or measures of variables included in the model. ese could encompass compensation issues, human capital required as well as proper distribution of incentives to a ract the crowd, but simultaneously keep the intermediary healthy having in mind that crowdsourcing success is not guaranteed in each and all cases.

Key Partners -users that provide knowledge
Di erent level users: 1 st level -anyone and everyone; 2 nd level -core team (selected).
Full Awareness.
Both demand rst and supply rst approaches.

Key Resources -platform & internal HR
Internal HR for: Preparation of primary material; Guidance of the core team; Aggregation of knowledge to the nal product.

Channels -information & knowledge transfer
Conventional on-line text based communication.
Video/audio material. Interactive media, e.g., simulations, which let "touch the model", but not disclose the commercial secrets. Video/audio conferences for the core team to facilitate discussions. Mobile access to increase time spent on the platform.

Cost Structure
Dynamic model of external HR -easy to control costs by controlling amount of external HR. Initial investment in the platform & its maintenance. Internal HR salaries for preparation of primary material & aggregation of the nal product.
Other services, like design, animation, copywrite could be taken cheaply from freelancers in the core team. Compensation for the core team. Empowerment of the core team cuts costs on compensation for 1 st level users.
Lawyers who could help in case of con ict (outsourced or kept as barter exchange result).
Sales & marketing of the platform.

Revenue Streams
Value-in-use pricing -good & complete solutions cost much. Challenge price -price for the nal product. Extras cost extra: featured project listing price, urgent solution, highly complex solution (platform is not capable to do it by itself) fees, feasibility studies fee.
Fine and kept advance payment for dropping the project (instead of requirement to pay full amount). Direct advertising in 1 st level of contests.
Fee for "branded" space. Fee for building external commi ees. Fee for browsing through ideas bank & portfolios Barter market (especially with NGOs) -services provided in exchange of internal knowledge (which is used to build knowledge base for the core team) and advertising of the platform.
Source: created by the authors.