By Keith Arrowsmith, Global Product Marketing Director, Onetag
As marketers navigate the online world, ever mindful of the risk of their ads appearing in poor-quality environments, ensuring brand-safe advertising alongside relevant and high-quality journalism has never been so important.
Most browsers are already cookieless, with Chrome to follow, so the old ways of targeting based on the final click are yielding to different methods that allow us to reach the right consumers in the right environment and take them through the marketing funnel. And at the forefront of these is a return to the proven approach of contextual advertising and its more powerful equivalent, semantic data targeting.
Contextual targeting is a long-established driver of advertiser engagement and conversion, and also helps drive brand preference and consideration, encouraging audiences who aren’t immediately in-market to click and convert. The crumbling cookie is bringing advertisers back to this approach, and the latest semantic targeting technology has greatly expanded its richness and scale.
As IAB Europe’s latest survey reminds us, today’s contextual targeting incorporates advanced statistical methods, machine learning and semantic analysis to provide “surgically precise content classification instantly through programmatic pipes”. Today, almost 64% of the survey’s respondents are evaluating contextual solutions.
What makes contextual programmatic advertising such a popular method is its ability to serve impressions according to both site classification and keywords. That means it reaches interested audiences in relevant quality environments in a privacy-friendly way, while delivering high personalisation. Going a step further, semantic data harnesses machine learning to enable greater scale and deeper relevance for brands.
Another great advantage of this targeting technology is that it significantly increases the scale of traditional contextual advertising beyond the limited scope of domain lists. Campaigns connect with relevant consumers wherever they are consuming relevant content topics across the premium open internet, including general news, entertainment and lifestyle sites. This new approach makes advertiser budgets go further and work harder, helping to ensure a diverse and well-funded premium open internet for all.
Moving beyond keywords to topics
For advertisers considering implementing semantic data in their media plans, the benefits are clear: greater accuracy beyond keywords, higher specificity beyond categories, universality beyond cookies, and availability across all DSPs, formats and devices. But how does it work?
Semantic technology uses machine learning to understand the meaning of each page of open internet content, to ensure greater relevance for advertisers for whatever content topics they require.
Onetag semantic data is built on a proprietary knowledge graph leveraging over 100 million data points, enabling the platform to propose larger contextual targeting recommendations based on the relevance of related premium web content.
Our algorithm only analyses the core contextual content of webpages, disregarding unrelated editorial information and thus streamlining the process. Algorithmic scoring happens for both prominence (text appearing more prominently on web pages) and relevance (the relationship of words to their related entities). This scoring system ranks the overall relevance of web pages and informs the search results that appear in Semantic Topics.
Understanding Semantic Topics
Semantic Topics is built into the Onetag media curation platform UI with a search box similar to that of Google. Curators can simply input the keywords for the content they want to target, as well as leveraging related topic suggestions generated by the semantic engine.
Searchable content can be categorised into three categories for the optimum blend of semantic advertiser targeting:
- broad search terms relating to the consumer lifestyle or environment in which products might be consumed, e.g. “theatre” or “film”
- medium terms referring to a product category, e.g. “alcoholic beverages” or “beer”
- narrow terms homing in on the brand or competitor brand
Using generative AI planning and Semantic Topics, we help customers identify highly targeted placements across the premium open internet. We then extend this search by combining our knowledge graph and Relevant Topics tool to identify relevant visited content interests beyond the core topics.
As discussed, the power of new semantic data technology is driving great results. At Onetag, applying Semantic Topics to always-on display and video campaigns has been shown to drive increases in average CTR by up to 500% across a range of industries from fashion to telcos and everything in between.
As a cookieless, privacy-friendly, brand-safe way to reach interested audiences in quality environments, it’s all to play for as semantic targeting fuels the re-emergence of contextual advertising.
Original published on: LinkedIn
