<h2>ABSTRACT</h2>
<p>We study how transparency—the observability of interim performance measures—affects innovation incentives. In each of two periods, an agent selects either a new or well-established idea and exerts effort to implement that idea into products or services. Transparency promotes idea exploration by allowing idea switching in response to negative signals. However, this can become counterproductive by wasting the knowledge acquired through implementation effort and reducing effort incentives if the output relies heavily on implementation effort rather than idea quality. Paradoxically, transparency can be further counterproductive if (1) the acquired knowledge becomes less idea-specific, (2) the variance in idea quality increases, or (3) the interim performance measure becomes more precise.</p>