Does artificial intelligence always contribute to marketing performance?

Purpose Although the benefits of artificial intelligence (AI) in marketing are widely acknowledged, it remains unclear how AI reshape marketing capabilities to influence performance – particularly whether AI always contributes to marketing performance. Design/methodology/approach Data were collected using a questionnaire. A structural equation modeling approach was used to analyze 261 valid responses. Findings By integrating resource-based view and capability theory and the data–information–knowledge–wisdom framework, this research concludes that cognitive analytics capability (CAC) significantly enhances AI-powered two-way (AI-TCC) and one-way communication capabilities (AI-OCC), which boost marketing effectiveness (MEC) and marketing efficiency (MEI), thereby enhancing performance. Crucially, moderation effects reveal that customer needs uncertainty weakens the contributions of AI-TCC to both MEC and MEI as well as the contributions of AI-OCC to MEC. Additionally, the heterogeneity of customer needs weakens the contributions of both AI-TCC and AI-OCC to MEI. Practical implications Organizations should recognize the importance of cultivating diverse functional marketing capabilities following AI integration. The results provide guidelines for managers’ decision-making in using matched communication approaches when facing various customers. Originality/value This study uncovers the mechanisms through which AI affects marketing performance and examines the conditions under which AI may fail to enhance it. By doing so, it directly expands the research on AI and marketing performance and further enriches the theoretical development of the AI marketing communication domain.