How Facebook Can Solve Its Advertising Problem

Publication: The Mail Room Blog


mary meeker internet trends

[Chart from Mary Meeker’s Internet Trends 2012

While perusing (if one can peruse a 112-page PDF) Mary Meeker’s insightful, clear-cut summary of dominant and looming internet trends, one chart in particular caught my eye. It showed the user growth of the mobile video creation and sharing app Viddy between March 31st and May 29th of this year. Until April 11th, Viddy acquired negligible new users. From April 12th through April 21st,  Viddy’s growth was fairly linear, going from around 500,000 monthly active users to about 8 milllion monthly active users. Then, on April 24th, Facebook enabled automatic newsfeed stories whenever a user uploaded or liked a Viddy video. One week later, the service had 25 million users.

The reason this chart stood out to me was that the surge is a result of newsfeed mentions, not display ad impressions (if we are to believe there is indeed a correlation between the newsfeed mentions and the surge in users).  It got me thinking about why we go to Facebook (to see what’s new in our social networks), and how Facebook and its ecosystem might better take advantage of that.

There have been many, many observations about why display advertising is wrong for Facebook: in sum, it doesn’t mesh with user intent. We don’t go to Facebook with intent to buy, but we do go there with an expectation to discover. In general, there are two discovery hubs: the newsfeed and photos. Both hubs provide opportunity for embedded advertising.

The infrastructure for the first hub is already in place, thanks to Open Graph. Put simply, it informs you of your friends’ activities and the app or service through which they occurred. As such, it is inherently viral.

In addition to Viddy, you see it used by Spotify (“Sarah listened to Islands by the xx on Spotify”) and the Washington Post Social Reader, (“Matt read ‘Snooki Looking Scary Thin at Cannes’ “) among others.  Any company, not just Facebook partners, can take advantage of the Open Graph in their apps, and they should—though without automatic social sharing, which (rightfully) tends to trigger a lot of backlash.

The infrastructure for the second hub is not in place, but all of the pieces exist today. You know those early sections in tabloids that consist of paparazzi shots of celebrities with itemized breakdowns of which designers they’re wearing? What I imagine is a dynamic version of that, where users can scroll over their friends’ pictures and videos and get ecommerce and review links to this dress and that bottle of wine and this restaurant and that bicycle. Even if I hate Brad’s gingham button-down, I still might be curious as to where he got it. (In fact, depending upon how I feel about Brad, I might be especially curious.)

To have their goods or services displayed, companies would have to sign up and pay some amount per-click-through or per thousand impressions, same as they do now with display ads.

The technical feasibility of this is, for many items and places, achievable. Image classification is getting really good, and it’s something Facebook, with their homegrown facial recognition software, could easily pick up. As for privacy—I think it would have be something users opt into, but it doesn’t, at least to me, seem to have malignant consequences.

Both of these approaches align with how we primarily use Facebook on our mobile devices: we browse our newsfeed, rifle through photos, watch short videos. In addition to increasing exposure and revenue, both give companies customer deep data. In my (admittedly naïve) opinion, it’s a win-win. What say you?

[Second image via GadgetGuyTech; third image via People StyleWatch]

Comments (



%d bloggers like this: