Abstract
This research explores the dynamic nature of in-feed advertisement avoidance, moving beyond static personality-based models. By distinguishing between "skipping" and "blocking" behaviors, the study identifies how a user's cognitive state—convergent versus divergent processing—influences their reaction to ads. The authors propose a multi-state transition model to track these behaviors and introduce two novel retargeting strategies: Guided Attribution for blocked ads and Prominence for skipped ads, aimed at optimizing platform engagement and ad efficacy.
Introduction: The Challenge of In-Feed Ad Avoidance
In-feed native advertising, which integrates ads seamlessly into social media and news scrolls, has become a dominant digital marketing format. However, users have developed sophisticated avoidance mechanisms. Traditional research often treats ad avoidance as a static trait or a simple negative reaction. This paper challenges that view, arguing that avoidance is a dynamic process driven by the user's real-time "information processing mode."
The core problem for platforms is the "black box" of user behavior: why does a user scroll past one ad but actively block another? By linking these actions to cognitive states—specifically whether a user is deeply focused on a goal (convergent processing) or browsing for variety (divergent processing)—platforms can better predict and respond to avoidance.
Theoretical Framework
The researchers propose that blocking and skipping are cognitively distinct actions triggered by different psychological needs.
As illustrated in Figure 1, the framework identifies two primary paths:
- Convergent State: Users are goal-oriented. When an ad feels intrusive, it threatens their sense of autonomy, leading to high-effort blocking behaviors.
- Divergent State: Users are seeking variety. If an ad fails to provide sufficient psychological arousal, they engage in low-effort, habitual skipping.
Methodology: Modeling Dynamic Transitions
To capture the fluid nature of user behavior, the study employs a Multi-state Frailty Model. This mathematical framework tracks how users transition between different states of interaction during a browsing session.
Figure 2 details these transitions. State 1 represents active browsing, while States 2 and 3 represent skipping and blocking, respectively. The model accounts for "residual effects," where previous interactions influence the probability of future avoidance or session termination (State 4). This approach allows platforms to move beyond binary "click vs. no-click" metrics to a more nuanced understanding of user flow.
Key Insights and Retargeting Strategies
The research suggests that avoidance signals are not merely failures but data points for optimization:
- The Signal in the "X": Blocking an ad is a high-quality signal of attention. The authors propose Guided Attribution, using the specific reason for a block (e.g., "not interested" vs. "too frequent") to refine the timing and content of the next exposure.
- Breaking the Habit: Skipping is often a learned, low-arousal habit. To counter this, the Prominence Strategy suggests making subsequent ads more salient—using concrete elements for focused users or high-arousal, abstract elements for variety-seekers—to interrupt the habitual scroll.
- The Timing Window: Retargeting follows an inverted U-curve. Re-engaging too quickly feels like harassment, while waiting too long loses the "residual effect" of the initial exposure.
Conclusion
By aligning ad delivery with the user's dynamic cognitive state, platforms can mitigate the negative impacts of ad avoidance. Matching an ad's "concreteness" to the user's processing mode significantly improves recall and attitude, even after an initial avoidance event. Future work will look toward applying these dynamics to rapid-scroll formats like short-form video and balancing prominence strategies with long-term user well-being.
Summary updated