Knowledge workers operate in relatively complex settings, where they must often use their judgment and learning capabilities to complete knowledge-intensive tasks. This makes them less suited for the productivity analysis techniques traditionally used in operations management. In my dissertation, I use controlled lab experiments to explore various ways of improving knowledge worker productivity. In the first part of the dissertation, I study how the design features of an enterprise social media platform, a popular form of communication technology that knowledge workers can use to seek help, affects a knowledge worker’s helping behavior and productivity. I also elicit the behavioral mechanisms driving effective design features. I find that features that draw on behavioral mechanisms in the form of descriptive social helping norms and reciprocity result in minimal, and in some cases detrimental, effects on help with no effect on performance. In contrast, I find that design features that use goal setting motivation and symbolic rewards to leverage an employee’s intrinsic motivation to “be more helpful” can be a surprisingly effective way to promote helping behavior and improve performance. In the second part of my dissertation, I study how varying constraints on the usefulness of a knowledge worker’s creative output affects their performance on an originality-focused creative task. I show that low usefulness constraints can be an effective way to improve employee performance by encouraging employees to factor in more usefulness in their creative output. In contrast, moderate and high constraints can result in poor employee performance by respectively causing individuals to create output that is creatively poor or output that is not sufficiently useful. Interestingly, I show that, in such cases, employee performance can be improved by “artificially” lowering the usefulness constraint set or by changing the task goal to emphasize usefulness. In the third part of my dissertation, I study how varying a knowledge worker’s ability and freedom to switch between a creative and a repetitive task affects their overall performance. I show that forcing employees to switch tasks can improve creative task performance, but can also lower repetitive task performance, resulting in similar overall performance. Interestingly, I find that giving employees the discretion to switch can result in infrequent task switching, lowering overall performance by lowering their repetitive task performance without increasing their creative task performance. I show that, in such cases, behavioral nudges can be a surprisingly effective way to improve employee performance by encouraging them to voluntarily switch tasks more often.