Organizations are turning more and more to crowdsourcing to harness the power of online communities for the purpose of ideation tasks. These organizations are also building online crowd ideation platforms to collect knowledge contributions from a large number of individuals (Afuah & Tucci, 2012). Online crowd-ideation has become an essential method for companies to utilize to harness the collective wisdom of the crowds on an online platform. Examples of this are the IdeaStorm funded by Dell, and the ideation website called My Starbucks Idea funded by Starbucks to look for solutions to various business issues. Crowdsourcing has been used in various fields, most notably in the business world, where it has been applied in areas such as advertising, product design, customer service, and the online product community.

However, online crowd-ideation is inefficient when faced with numerous contributors. For example, only 277 ideas have been utilized by Starbucks from a huge pool of 150,000 ideas. Contributors always provide much overlapping or superficial content for a crowd-ideation task. First, most submissions are redundant without accessing or processing others’ ideas. Given the high redundancy and loose structure of existing ideas, reusing them is difficult and results in inefficient crowd-ideation. Second, much content is superficial. Ideation is a complex cognitive task for users. Within limited cognitive capacity, idea contributors may waste too much effort processing unnecessary information during ideation tasks and lose an opportunity to absorb stimulus when proposing ideas. To solve these critical problems of overlapped and superficial content, this thesis designed two studies to help users access the existing ideas with shared collective memory (Study 1) and make the task easier by increasing the efficiency of occupied cognitive resources (Study 2) to answer two questions.

Transactive memory systems have been acknowledged in the information systems literature for improving knowledge processes and performance for co-located teams. Studies have indicated a shared awareness among group members of those who know what happened (Moreland, 1999). However, overlapped content without systematic construction in online crowd-ideation causes insufficient collaboration. Its potential is far from fully realized because of the unorganized and redundant idea pools. Thus, building a transactive memory system to utilize online knowledge and promote contributors’ effectively sharing (i.e., transacting) their knowledge becomes essential. Study 1 advanced a research model to extend the transactive memory system that triggers contributors’ external knowledge processing to improve knowledge self-efficacy at the individual level and collaborative mindfulness at the collective level. This thesis explores the role of the transactive memory system in enabling knowledge reuse and creation among contributors in the context of crowd-ideation. Two theory-driven features (visualization and interaction feature) were designed to fill this gap based on the role of transactive memory.

Cognitive load theory explains how to use our cognitive capacity to construct schemata. It has been applied broadly in learning and offers various design principles to promote learning efficiency and effectiveness based on causal factors, such as task difficulty and interaction between task and learner. More generally, cognitive load theory is regarded as an instructional design theory that reflects our “cognitive architecture,” or the way individuals process information. In the realm of crowd-ideation, the idea contributors are volunteers without tangible benefits. When faced with numerous ideas posted on the crowd-ideation platform, the critical challenge is to help users access others’ ideas more easily without unnecessary cognitive load. Study 2 extended the cognitive load theory into the goal-free cognitive task in the context of crowd-ideation. Moreover, Study 2 delineates three types of cognitive load (intrinsic, extraneous, and germane) in goal-free cognitive tasks and advances three theory-driven features (visualization, interaction, and selection) that can be incorporated into crowd-ideation platforms to bolster contributors’ efficiency in cognitive load allocation.

This thesis conducted a series of online experiments to validate our hypotheses. First, our design features (visualization and interaction) showed their significant effect on inducing accessibility to the transactive memory system. Second, the influence of the transactive memory system has been verified on collaboration performance (i.e., collaborative mindfulness and knowledge self-efficacy) through knowledge reuse and creation. Third, the goal-free cognitive task exhibits many differences compared with the goal-fixed task, and a ranking measurement has been used as an appropriate proxy to represent the cognitive load allocation scheme. Fourth, the design features (visualization, interaction, and selection) can help users alter the cognitive load allocation scheme during contributors’ ideation tasks.

Findings from this thesis can contribute to both research and practice. First, a deeper understanding of how crowd-ideation platforms can be designed by constructing a transactive memory system to promote collaboration among contributors has been provided. Second, the effect of transactive memory systems on collaboration in the context of crowd-ideation has been investigated by including the role of knowledge reuse and creation. Third, this thesis investigates the manifestation of cognitive load allocation for the goal-free cognitive task. Moreover, it extends the cognitive load theory by considering cognitive load allocation, not merely cognitive types, which improves our understanding of cognitive load theory. Fourth, this thesis confirms the effectiveness of design features in facilitating efficient cognitive load allocation and provides a potential research stream on information presentation design with cognitive load allocation, which indicates its significant development potential. Finally, this thesis offers much instructive advice to practitioners to utilize existing knowledge in the crowd-ideation context. The theory-driven features also provide principles to guide platform design, even benefiting from our protocol website.