Leveraging AI-Generated versus Human-Created Marketing Content for Enhanced Qualified Lead Generation in SaaS Companies: A Comparative Study

Abstract

In times of digital transformation, marketing departments increasingly face the challenge of producing large volumes of high-quality content to generate leads and support customer journeys. The emergence of generative artificial intelligence tools promises to address this pressure by automating content creation. However, in practice, it remains unclear whether AI-generated content can effectively substitute human-created content, particularly in B2B SaaS environments. This thesis addresses this gap by examining the comparative effectiveness of AI-generated versus human-created content in driving qualified lead generation. While prior studies have explored AI in marketing or compared content formats, little empirical research exists that tests performance differences in a real-world B2B setting using controlled A/B testing and live campaign data. This study aims to provide both scientific clarity and practical relevance. The central research question is: “How does AI-generated marketing content compare to human-created content in terms of engagement, conversion rates, and lead quality within the B2B SaaS industry?” A sub-question explores whether content format and platform influence performance differences. To answer these questions, a field experiment was conducted in collaboration with pixx.io, a German SaaS provider. A/B tests were performed across four digital marketing formats, comparing AI- and humangenerated content using metrics such as click-through rates, engagement, form submissions, and where possible, lead quality. A standardized briefing and prompt structure ensured content comparability. The analysis also accounted for the role of prompt engineering and platform-specific delivery constraints. The results show that AI-generated content performs comparably to human content in standardized formats, with no significant differences in most KPIs. In contrast, human-created content clearly outperformed AI in engagementdriven formats such as Meta ads, particularly in CTR and interaction metrics. The thesis contributes to research by empirically validating the potential and boundaries of AI in content marketing and by framing AI not as a replacement but as a co-creative tool. For practice, the study offers a structured comparison, actionable insights for content strategy and a workflow model that balances AI efficiency and human creativity. Limitations include short campaign durations and restricted SQL tracking in some channels. Future research should extend the testing window, explore more granular audience segments and examine longterm brand effects of AI-generated content. In summary, this thesis shows that AI can reliably support content-driven lead generation, especially in scalable, low-complexity formats. However, human input remains essential for persuasive, emotionally resonant communication, calling for a hybrid content strategy that leverages the strengths of both.

Date of Award	2025
Original language	English
Supervisor	Michael Amann-Langeder (Supervisor)