Engagement has received a profound impact from contextually-related ads on images. Attention is focused on them. These ads have proven better than standard banner ads. New channels are created for advertisers and monetization is created for publishers.
Images need to be understood. Research over the last five years has shown that images are a powerful way to advertise. Have we always doubted that? Apparently so. Even when we have been bombarded with images our whole lives. Perhaps it's putting the image category, as I'll call it, into the online experience.
Technology has come a long way. There are about 5 trillion images online. While not all of them are suitable for advertising, the sheer volume has got to help the contextually relevant ad campaign. To first categorize images, we must first know what images are all about.
There are three basic image "meanings":
- COMPUTER VISION. For purposes of in-image advertising, programmatic image recognition has so far been geared toward automobile i.d., skin detection and facial recognition.
- SEMANTIC ANALYSIS. Information about photos can be identified through meta data.
- HUMAN CLASSIFICATION. Review and enter information about photos.
In-image ads can be served as display overlays or embedded links.
- DISPLAY OVERLAY. Serve a relevant ad during the time a viewer's attention is focused on the ad.
- EMBEDDED LINKS. Embedded links within images can convey more information about a photo and send people to the marketer's site.
Understand with this kind of advertising, there are creative and targeting strategies to consider. Concentrate on engaging the target audience. This will make the experience more consumer relevant.