# From Words to Context: Otto Media Explains Why Advertisers Must Move Beyond Keyword Dependence


Over the past decade, keyword blocking has been treated as one of the safest and most widely adopted tools in brand-safety strategies. Yet as programmatic buying has expanded at an exponential pace, user-generated content has come to dominate mainstream media traffic, and AI-generated material has spread rapidly, keyword blocking has begun to resemble a hidden media cost rather than a genuine form of brand protection. In 2025, IAS conducted experiments across four major US media outlets and found that as much as 72% of content deemed "fully suitable for advertising" was incorrectly classified as unsafe, while 70% of impressions were unnecessarily blocked. At an assumed $6 eCPM, this translates into more than $14m in monetisable inventory disappearing without trace.

As this pattern spreads across global markets, advertisers, media platforms and creator networks are confronting the same question. In an era where content volume far exceeds human review capacity, defining context by individual words can no longer accommodate the complexity of the current media environment. For Otto Media, which operates large-scale creator ecosystems and distribution networks, the implications are particularly acute. Misclassification not only undermines campaign efficiency but also erodes the value of vast pools of high-quality mid- and long-tail content.

## Why Keyword Strategies Break Down: Modern Media Requires Semantic Judgement, Not Word Matching

Keyword blocking can function when traffic volumes are modest. In a multilingual, cross-cultural and hybrid-content internet, however, its weaknesses are quickly exposed. A single word often carries multiple meanings. "Shot" can refer to violence, a medical breakthrough, a basketball attempt or a film production. Traditional blocking systems are unable to distinguish between these semantic contexts. IAS research shows that such errors generate several structural consequences: media waste, missed exposure in high-quality news environments, delayed post-campaign optimisation and declining publisher revenue.

This problem is particularly pronounced in Southeast Asia. In Indonesia, where content frequently blends semantics, dialects and English expressions, keyword-based systems produce far higher misclassification rates than in Western markets. Otto Media has observed in its localised advertising operations that much of the blocked inventory is not unsafe at all, but simply misunderstood due to a lack of contextual comprehension. This is especially evident in collaborations with MCN creators, where a significant share of "positive content addressing sensitive social issues" is mislabelled as negative, causing brands to miss valuable creative opportunities.

## Contextual Positioning: From "Risk Avoidance" to "Suitability Assessment", a Structural Shift in Advertising

Contextual positioning has emerged as a shared direction for global advertisers and media platforms not because it is more sophisticated, but because it aligns more closely with how content actually functions. In contextual intelligence systems widely adopted across Asia-Pacific markets, content is analysed through NLP models that assess narrative structure, sentiment, tone and theme, rather than merely identifying isolated words. This approach allows systems to distinguish between "reporting on a tragic event" and "explaining scientific progress", and to recognise whether creators address social issues in a neutral or constructive manner.

Contextual positioning also aligns with three long-term industry trends. First, as third-party cookies disappear, contextual signals are becoming a more stable form of targeting. Second, brands are shifting away from the pursuit of absolute safety towards "brand suitability" aligned with their values. Third, advertisers are increasingly aware of the invisible budget leakage created by keyword blocking.

Within Otto Medias content ecosystem, contextual understanding is applied not only to ad delivery but also to creator matching. Different categories of creators are paired with brands through contextual analysis, enabling more precise collaboration layers without reliance on manual screening. Context, in this sense, is not merely an advertising technology, but a way for platforms to understand and organise creator ecosystems.

## From Loss to Growth: Why the Industry Must Move Beyond Keywords

The greatest danger of keyword blocking lies in its ability to make waste "invisible". Advertisers believe campaigns are running smoothly, while blocked impressions never appear in reports, leading to persistent misjudgements about media performance. Over-filtered inventory shrinks supply, bidding becomes more expensive, learning models are forced to slow, and the monetisation capacity of high-quality publishers is weakened. The result is a negative feedback loop that affects the entire industry.

By contrast, the economic advantages of contextual positioning are becoming increasingly clear. Advertisers gain more effective impressions under controlled risk conditions. Media platforms monetise content more fully. Creators no longer see their work "hidden" due to keyword misclassification. Brands achieve broader and more scalable reach. For a bidirectional platform like Otto Media, spanning both content and advertising, this means preserving value within the ecosystem rather than filtering it away through blunt and outdated tools.