A New Video Frontier With Contextual Media

  • by December 30, 2009


By Kevin Schaff


With a new year upon us, many professionals in the digital media, marketing and advertising spaces are looking to see what's ahead for video. The answer is "contextual media." Looking back, the dynamic nature of digital publishing has enabled flexible and contextual delivery of messages and content in a way previously not possible when we were locked into the one-time printing of an ad on a page, or the scheduled broadcast of a television program. Advertising was revolutionized by Google AdWords, allowing marketers to serve messaging in a micro-segmented and automated fashion. On the content side, text such as links and articles have also been served in a dynamic and customized manner, providing related links and topic pages to match consumer interests and creating new ad inventory. Now, video is exploding as a content type and contextualization will come into play here as well, bringing the engaging quality of the medium together with scale and relevance.  Richer metadata, deeper digital archives, and flexible delivery tools will enable delivery of relevant video content around categories such as sports, news, and entertainment in new engaging ways.


The 'Nuts & Bolts' of Contextual media

The Web has made everyone from newspapers to cable providers into online publishers, which increasingly includes video. Contextual media tools enable them to serve video around stories and topic areas to deepen users' experiences and increase dwell times, heading towards a future where "instant documentaries" are possible.

Next-generation contextual media providers use advanced technology to organize video through metadata and enable API "hooks" for searching and serving it while minimizing manual post production work. Along with dynamic delivery, this enables online publishers and media companies to produce new video packages at scale to drive user engagement and ad revenue. Organizing and delivering motion content through metadata production also allows people to jump in and out of a video to the moments they want to see, creating a more engaged experience that fits the way people interact with content on the Web.


What's the Value?

In today's Internet driven world the average news story breaks in minutes, sometimes even seconds. Contextual media tools enable online publishers to quickly put context around breaking news through archives of curated relevant content. For example, say the most horrific category five hurricane in US history hits southern Florida. A site like MSNBC could rapidly assemble archival footage of every category five hurricane to have ever hit the region and within hours produce an engaging online video analysis with the speed and efficiency required to make online content monetization economics work. Simply put, contextual video enables online publishers to put a documentary right next to the breaking news event. Adding this type of contextual video around news can tremendously increase site traffic, visitor dwell time, and extend the depth and lifecycle of a story.

Contextually relevant motion content is a new market that fits in between costly premium content built through post production editing and low-quality user generated or mass-produced video. Many publishers, such as USAToday and New York Times, have already started to see the ROI in adding more contextually relevant video to their sites and I foresee this trend to continue in the years ahead.

The engaging nature of video combined with relevant and contextual content can drive usage metrics that support higher CPMs and sell-through rates. Given today's tough ad market, these new ways to create value are certainly appreciated by digital media, marketing and advertising executives.

5 comments about "A New Video Frontier With Contextual Media".
Check to receive email when comments are posted.
  1. Thom Kennon from Free Radicals, December 30, 2009 at 3:19 p.m.

    Interesting post, and agree there is a trend in there somewhere...

    But it's not just the traditional "media" companies like "newspapers and cable providers" who are now potential masters or instigators of contextual video --- everybody from brands to people to the local community college now come with reams of their own videos.

    Brands - esp. tech, auto & CPG - have terabytes of products demos, MOAs & TVC spots.

    Entertainment properties have, well, a ton of video ranging in size, shape, nature and quality.

    People have their YouTube channels where they collect, upload, tag, enjoy and share all sorts of content.

    Institutions - think universities, think tanks, political parties, NPOS - have their interviews, original content, streaming events - all worthy inputs to contextual mashups.

    The tools & content are all there. What will make this a trend worth watching - or better yet, driving ; > - will be how and who creates dynamic sticky environments that create value and engagement for our end customers, consumers, fans, voters & citizens.

  2. Peter Schankowitz from Joe Digital, Inc., December 30, 2009 at 3:34 p.m.

    Terrific post. The notion of Contextual Media as, perhaps, the cost-effective yet engaging content offering that sits between professional video and the limitations of UGC, is what will be worth watching in 2010. In the end, this boils down to the meeting of two key elements---enriched storytelling and measurable engagement. As always, tell me a great story, and I will stick around.

  3. Jeff Mayerczak from CELLMANIA, December 31, 2009 at 9:26 a.m.

    So who provides the platforms and services that allow the Publishers to compete in the Contextual Media world?


  4. Pinaki Saha from Me!Box Media Inc., January 1, 2010 at 12:03 p.m.


    Wonderful post.. and all the points that you highlighted are so timely!

    Contextual media is here to stay. All the above points converge with our case studies and focus group feedback. Yes, we have a platform that actually provides all the above and right now we are getting real time analytics from usage that are amazing... I will share in a moment.

    Dwell time (that Kevin mentioned) is the fundamental driver to everything! With dwindling CPM rates and Netflix ads appearing in a Mayo Clinic site is actually diluting recalls and call to actions. In a 80-20 demographics, the hyper users are indeed seeking out more from the video file. When we entered the market in a pilot form, we targeted brand videos first... in order to test user interest and aspiration to learn more about the product being showcased. We found that along with e-commerce possibility, there is a huge opportunity for building loyalty/followership, viral awareness of products/services and consumer advocacy and evangelism when the right information is delivered with the right context in a video. However, one item of caution - the platforms need to be very careful at what point contextual info becomes annoyance due to intrusion in the media experience - too many hotspots, on media roll overs, annotations and other interactivities. In early adoption, too much is not always too well received.

    You may play around with one of our pilots in the ASUS video site. Go to and click on 'View Video'. (our old site is being modified right now to showcase our latest offerings)

    I would like to share some real figures that we are getting from 3 other pilots we are running in CPG domain:

    1. Out of 100 video views, 45-55% are engaging in clicks to dig deeper into the contextual information associated with the video

    2. Of those clicking to engage with the media, there is a 450% click ratio

    3. And the average dwell time for a 1 min video for a user is 14 minutes! and that's what threw us off.... So there is definitely a tremendous interest and behavioral tendency to stay with your media if you can provide the right contextual information at the right position and time

    4. Long tail is resurrected with a much higher half life. Our pilot producers have embedded old videos from their libraries at correlated points in the new media, which has generated nearly 30% view back conversion

    Thom, you have just mentioned a significant section of our target groups. Actually, we are in the process of wrapping some content for a citizen journalism project and one of our new beta feature allows your audience to submit content/URL on a timeline event that you open up for contribution - so this builds advocacy and user groups around your property category.

    Please feel free to ping me if you would like to learn more about the case studies and what we have found during our 6 months data gathering from the brand market and brand properties.

    Wish you all a very happy and prosperous 2010!


  5. Pierre Wolff from Livefyre Inc., January 4, 2010 at 4:12 p.m.

    Totally agree w/Thom's comments below. However, I'd go further to say that there is one assumption made in this piece which merits more attention.

    The metadata issue has not been solved. It needs to made more consistent and normalized to have a chance at creating a compelling user experience. Off-the-cuff tagging done by editors and publishers today is very inconsistent. Whether it be how it's applied by different publishers or editors, or how it's applied differently based on the content property. This is a devilish problem that stands in the way of contextual media. Just as an example, there are at least seven different spellings in tags on CNN for "Qadafi" (or Gaddafi). Yes, this might be an extreme example, but there are plenty of less extreme examples like this.

    This matter gets exacerbated when desirable opportunities for content syndication between publishers can take place to enhance the user experience. For example, one could imagine CNN providing BabyCenter with relevant news items on newborns or baby formula. Where different publishers maintain different tagging or classification standards, it becomes difficult for the sender to communicate what is being sent over so the receiver can manage it appropriately, and for the receiver to request what they want from the sender. Mapping taxonomies and tagging vocabularies, becomes critical to making this process work smoothly.

    Note, none of the existing automated classification technologies pick up nuances in the content very well. It's a difficult prob that continues to require manual intervention (by editors) to do it well. Given the amount of video that each publisher produces annually, a manual solution is actually not much of a burden. For UGC, the problem is a bit more intractable.

    Publishers now need the tools to resolve these issues. However, to date they been more focused on just getting video up on their site and managing this for basic TV model-like ad sales, to demand this. I expect the next generation of publisher innovations to be more focused on this contextual media oppty and for the video platforms to enhance their services by providing these tools.

Next story loading loading..