Bitmovin’s AI Contextual Advertising optimises ad placement based on machine learning to deliver hyper-personalised ads for audiences based on the content they are watching.
Bitmovin’s new AI Contextual Advertising combines the features and functionality of the Bitmovin Playback system and Bitmovin’s Encoding to deliver hyper-personalised advertisements for audiences based on the content they are viewing. Bitmovin’s AI Contextual Advertising optimises ad placement for users based on a machine learning model that extracts the relevant characteristics of every video scene and analyses viewer engagement.
The assumption here is that viewers are likelier to connect with ads related to content they are already watching, featuring products that they are attracted to on some level, resulting in more ad-generated revenue. Bitmovin’s AI Contextual Advertising can make ads feel more relevant, enhancing both engagement and the overall viewing experience. For example, if a viewer is watching a show set in a luxury hotel, effective advertisements might be from hotel brands, cruise ship holidays or health and beauty.
Cookieless Attribution
Bitmovin considers that one of the main advantages of AI Contextual Advertising is that it is based on cookieless workflows and delivers relevant ads to viewers without relying on personalised tracking. Instead of cookies, cookieless attribution focuses on collecting data anonymously. Examples of such methods are device fingerprinting, which identifies every visitor through a specific set of data that their browsers pass on to a website, or probabilistic modelling to estimate the likelihood that a particular user will interact with an ad.
Bitmovin’s VOD and live Encoding software uses AI to analyse video and audio material, and uses an OpenAI-based model to extract metadata from the content that will help focus advertising placement on that content. Each extracted piece of content metadata is assigned a timestamp to map that information to the specific position in the content it belongs to. The metadata describes the content in sufficient detail to allow this kind of targeted ad placement, but it can also describe the scene more broadly.
Content-Aware Ad Serving
The Bitmovin Player, a feature of the Playback system, then takes this metadata and uses Bitmovin’s API to request relevant ads from a content-aware ad server that, in return, replies with contextual ads. Using a conversion heatmap generated by the Player, the ad's position is optimised to increase the chance of a conversion. This heatmap is mapped to categories based on the Interactive Advertising Bureau’s taxonomy, along with an AI analysis of historical data and the current user environment.
IAB categories are standard classifications that publishers and brands use to categorise web content, aiming to improve targeting. The Content Taxonomy serves as a consistent method of organising this content, differentiating sport from news or arts material. The IAB Taxonomy also helps prevent the use of content categorisation signals to generate sensitive data points about personal traits such as race, politics or religion.
Conversion Rates
“Lack of personalisation in advertising remains a major challenge for brands.” said Bitmovin Co-Founder & CEO Stefan Lederer, “AI can help tailor advertising experiences more accurately for the viewer by analysing the content they are watching, which can help pinpoint what mood they are in and place ads that they will be more receptive to in that moment.”
Furthermore, Bitmovin AI Contextual Advertising & Prediction measures conversion rates at different points of the video, combined with information about the user’s viewing environment, to place the ad when the viewer is most likely to convert. Contextual metadata can also be used to ensure brand safety and avoid placing adverts with content incompatible with the brand. For example, it would avoid showing adverts for airlines during ad breaks for content that features plane crashes.
The fact that Bitmovin’s AI Contextual Advertising integrates two existing products removes the need for third-party integrations, making it simpler for users to integrate. It is compatible with Server-Side Ad Insertion (SSAI), Client-Side Ad Insertion, and Server-Guided Ad Insertion (SGAI). www.bitmovin.com
