Embedding Images: A Contextual Ad Delivery Model?

Embeddable images through online publishers can be a tactic for extending the reach and brand of publishers, who are the direct recipients of its traffic-building and presence. As you allow sharing, your content network grows, as do link-backs and overall web-imprint of a publisher’s site and branding. More often than not, the image owner and publisher are not the same entities. What’s in it for the image owner?

           UGC aggregators that welcome the outbound embedding of images from their platform typically expunge any metadata from images during the upload process, which might contain useful information on the images’ provenance. Rightful copyright, web links, subject and identification information can help tie-in the image owner to an embed network that they might find themselves party to. When these are absent, only the publisher benefits, marginalizing the very actor that creates the value – the value of end-user to publisher, but more importantly the high-growth network of embeds that the publisher now owns.

Announcing a photography marketplace  powered by Luminate in-image ads ?)
Flickr announcing a photography marketplace (will it be powered by Luminate in-image ads ?)

              While the practice of building up a network of embedded images is no easy task (inventory worth sharing plus traffic), a network as a channel for distribution is well worth the trouble. As a closed system, embeds offer the ability to deliver info to the third-party publisher’s site (rights permitting), namely advertising. In-image advertising has been with us for many years, and recent consolidation with Luminate’s sale to Yahoo! removed a significant provider of in-image ads from the competitive landscape (publisher access to in-image ads were turned off on October 1).  As Yahoo! seeks to launch its own solution and leverage its considerable ad inventory, coupling Yahoo!’s truncated Flickr monetization strategy with it is hard to overlook.

          The In-Image Ad industry isn’t just Luminate, as GumGum is now the clear leader, but both have fostered their own competition. Vibrant Media is modeling their success off of GumGum’s, and like them, their core strength is ad aggregation and publisher acquisition. Popmarker and Imonomy are non-US start-ups seeking a toehold in the US market, but the barriers to entry aren’t only in publisher acquisition. There are a limited number of publishers who have the rights to advertize over the images on their sites – let alone their embedded third party publishers. The real map to monetization lies within the open participation of image owners (or someone who has those rights in aggregate, like Getty) partnered with an in-image ad solution that is above all, relevant.

A GumGum advertisement at work
A GumGum advertisement at work

             Relevancy is achieved though publisher context – and not the context of the image – which is something often overlooked by those within the in-image ad industry. Many place their relevancy metrics on how closely an ad matches the subject of the image (again, something expunged metadata imposes barriers to), but there’s fleeting accuracy. Subjects and the content of an image – programmatically – don’t always convey a narrative, or underlying concept, that it’s trying to convey. When relying upon content of the image solely, the context between image, ad and publisher is lost.

One of many Imonomy in image advertising solutions
One of many Imonomy in image advertising solutions

                  Some are bucking visual recognition applications altogether and going a tried and true route of big (meta)data to contextualize their in-image ads. Netseer has developed an in-image ad solution that matches the ad against the content of the page the image is placed in – an approach that adheres closely to publisher intent, and therefore greater relevancy between ad and image. The in-image ad model, as a click through proposition, lives and dies on relevancy to the end user. The understanding of end user intent and preferences is much more of a big data problem than image tech problem; luckily, publisher sites offer a wellspring of information, and relevancy between the image embedded in the page and the content of the page is already predetermined (and still largely human-driven).

Netseer grabs contextual content to deliver relevant links
Netseer grabs contextual content to deliver relevant links

             As embedded image networks proliferate, it’s incumbent upon the owners of those networks to minimize impediments to sharing while finding new ways to monetize usage. Removing any type of pay wall is critical to adoption, given the type of end user typically involved; also critical is employing a solution that can incentivize continued and expanded use by publishers, while re-numerating image owners and advertisers alike.


Robert Henson will be moderating a panel on Image embedding at the annual PACA conference on Tuesday, October 21. You should register here

Lead Photo by recombiner

Author: Robert Henson

Robert Henson is a stock photography industry professional and analyst with a long executive career in top photo licensing companies. He is also a popular speaker and an insightful innovator. He was previously the VP , Business Development at IMGembed.

Share This!Share on FacebookTweet about this on TwitterShare on LinkedInShare on Google+Share on RedditBuffer this pageEmail this to someone

3 thoughts on “Embedding Images: A Contextual Ad Delivery Model?”

Leave a Reply