We study a variant of the voter model to optimize the influence of social media bots in a network where node’s interest in a certain topic or product is changing. In this model, bots possess finite lifetimes, and their influence intensifies with prolonged activity. Using tools from martingales and optimization theory, we develop a framework to determine the optimal temporal allocation of bots to maximize influence. Our findings provide insights for applications in marketing and political campaigns. For instance, Our results reveal a counterintuitive allocation strategy: when bots’ influence accumulates rapidly during their activity period, the optimal resource distribution shifts toward temporal uniformity rather than concentrated bursts.