Posts Tagged ‘Advertising’

Contextual Ads Based Off Social Network Profile: Twitter and Facebook


Things are moving very quickly now, in fact I was pleased to learn about these contextual ads from my new friend Cory O’Brien in SF yesterday.

In my latest report “The Future of the Social Web” we pointed that in the near future we’ll start to see web pages dynamically created based on user profile ID in social networks. Essentially, your corporate, media, or ecommerce site could provide contextual media, content, and advertisement based on users’ info before they login.

[In the Future, The Era of Social Context Will Serve Personalized Content, Media, and Ads to Users based on their Social Networking Information]

Here’s an early example of a contextualized advertising campaign from VW (by agency Crispin Porter + Bogusky, read more about the modernist campaign) that’s intended to help prospects find the right model based on their preferences. Please note this isn’t yet the full entry into the era of social functionality as users have to manually submit their login info or user name (like this Twitter example below) and just examines recent tweets. When the era of social context matures, it will look beyond just profile info, but also behavioral data, friend data, location, and content analysis of explicit and implicit data.

Type 1: Context Ad based off of opt-in Twitter profile.

Above Ad: Enter your Twitter name to see a product recommendation

Type 2: Contextual Ad based off of Facebook profile.
Corey also pointed me in the direction of a second “Meet the VWs” Facebook app that asks users to opt in to analyze their profile and then recommends products based off simple profile info. Read the pros and cons from the smart folks at the Future of Ads of this Facebook advertising effort.

This Facebook App scanned my Facebook Profile to Suggest two products
Above image: Facebook recommended these products to me on the VW fan page

Future Expectations:
Expect social context to impact not just ads, but many websites in the future. Also, expect the accuracy to increase as social and behavioral data starts to merge.

Facebook, Twitter, have a tremendous amount of explicit and more importantly, implicit data that could serve up information about users, yet we should expect years of refinement for these engines to truly be accurate. Interestingly, the Twitter ad suggested I’d like the Jetta, yet the Facebook app suggested a Rabbit and Beetle, which I find funny as I’d never drive a Beetle, that’s really not me at all.

In the future, these ads, media, or recommendations should be more intelligent and also find friends with similar cars, or people with similar traits to me that I don’t know and suggest products. As user ID start to federate and connect with other such as Open ID and Facebook Connect, we should expect a higher degree of accuracy.

Then, users may choose to opt-in to expose parts of their identity as they surf the web on trusted sites to receive a contextual experience. For example, I may trust Amazon, eBay and Google search to expose my identity in exchange for a more personalized experience.

We should also expect a rash of privacy concerns and user backlashes to happen, even if they opt in, we’re just scratching the surface here. I have so much more to write on the topic of social context, but it’s 3am and I need to go back to bed, so I’ll save it for a future blog post.

Key Takeaways

  • The above ads are simple experiments of how context can be served up through social data
  • Expect this contextual content not to be limited to just ads, but also on media sites, ecommerce, corporate sites, and TV
  • Expect digital content to be contextual –even without express content of the users
  • During the early years, expect privacy concerns to overwhelm brands, causing them to rethink this approach
  • Although it will take years to perfect, expect context to increase CTR, and therefore the cost of ads
  • What did the Twitter ad and Facebook page recommend to you? Were they accurate recommendations?

    Breakdown: Digg Allows Community To Choose Advertisements


    Left: This screenshot, provided by Digg shows how an EA Sims ad is embedded in main body editorial as sponsored, as well as in upper right bug.

    Digg Launches Community Voted Advertising: “Digg Ads”
    Digg, who was formerly partnered with Microsoft for advertising, announced that they will be launching a new type of advertising unit that allows Digg members to vote up (digg) or down (bury) ads that appear in the editorial stream. As a result, the ads that are voted up will cost less to the advertiser.  Nodding to the power of the community isn’t new for Digg, in the past the Digg community actually has more control than the management team, so turning over the advertising power to them strategically makes sense.   After engaging in a discussion on Twitter last night about this, I gave it a good night’s think, here’s my take:

    Social Ads Not New:

    Requirements for Success:

    • Homogeneous : Community ranked ads will likely work better in homogeneous communities where there’s a common interest or demographic, rather than a large broad community where consensus won’t be found.  In the case of Digg, I’d make a guess from watching the community that it’s a lot of Gen X and Y males that are technology optimistic, and liberal.  Having spent time at the live Diggnation event (the super fanboys, photo by Brian Solis) it could be a representative sample.
    • Engaged: Community ranked ads make sense for the Digg community as they are already highly engaged in voting for stories, as well as the very active comment (over 100 comments is norm per article on front page).  
    • Transparency: Dislcosing in the editorial stream that the ad is sponsored.


    • Gaming: Expect gaming of the site, not from marketers, but from fan boys, perhaps those that love Apple products will bury Microsoft ads.  Since you must have a registered ID to vote on items in Digg, the chances of the advertiser influencing the ad price will be limited.
    • Unusual engagement: Expect that most users are more likely to bury ads, not engage with them and promote them.  However, if a user buries an ad they don’t like, this cost per action is still an engagement, which is higher than not paying attention to them at all.

    Forward Thinking:

    • If this works, Digg or it’s partners could replicate this product and extend to other sites.   Interactive advertisers like Federated Media would do well to open discussions with Pluck and Kickapps, who have a strong media focus.
    • For brands and advertisers this is a great way to find out why an ad may not work for a particular community, rather than make guesses based on CTR performance.  Advertisers that analyze or even engage in the dialog may benefit their next generation effort.
    • In the most ideal sense, community preferred ads become information and content –not invasive content.

    Takeaway: If Anyone Can Pull This Off, It Will Be The Digg Community 
    Digg is a very unique case study, and if these ads work here, it will be hard to replicate on other communities, a unique mix of a very engaged community that is somewhat homogeneous will be required to make this work.  Let’s see how it unfold.