The problem with social media advertising. Hint: It’s ‘social’ before it’s ‘media’.
Context is the basis for relevance in third-party monetization. For relevance to be high, the message to the user needs to be in the user’s context. In traditional media, the context was static based on the content surrounding it (magazine articles, TV show, etc.) which was taken as a proxy for the expected demographic and interests of the user.
Dynamic context is based on three parameters that define user choices at any point:
The first significant move to dynamic context was led by Google when it started monetizing intent (AdWords). The second big move was led by the emergence of GPS-enabled mobile which added location and time to the mix. In recent times, the social graph and the interest graph have been generated to, again, flesh out the context of the user based on the sum of her online activities while valuing the predominance of the current activity.
Here’s a quick definition of Social Media to set the context for this discussion. Social Media is essentially:
Social: Interactions among participants
Media: Monetization of a big audience
The traditional media model grabs eyeballs and sells them to advertisers. There are no interactions between users in traditional media, so this model serves it well. The eyeballs serve as great context for advertisers. Social media, on the other hand, has more than eyeballs. The actual interactions are the new context created.
The problem with Facebook’s advertising is that it is still trying to monetize the traditional media way: by monetizing eyeballs. Monetizing eyeballs is fine when your users are unconnected individual silos and interacting only with content but when users are connected and constantly interacting in an ever-changing context, monetization should target interactions. Interactions define the real-time context of the user.
To Facebook’s credit, it’s making advertising more ‘personalized’ by making it more granular than on any other media form, but it still is out of context on a page of conversations where the content of the conversation may have nothing to do with the ad.
Let’s classify platform interactions into two broad categories:
The two may overlap as well. e.g. in the case of a marketplace like eBay or a Freemium model like Flickr, both the above are manifested in the same interaction.
Also, a platform may have multiple instances of each interaction type. e.g. Spotify has multiple monetization interactions: Advertising and Freemium.
In case of Facebook, the monetization interaction (Advertiser – User) is targeting the static Facebook profile. The usage interaction (User – User), on the other hand, is often completely unrelated to the profile and is in the context of the dynamic status update. The content of the status update has little or nothing to do with the profile.
Unlike Facebook, LinkedIn’s monetization interaction (Recruiter – User) is in the context of its usage interaction (User-User). The usage interaction involved two users connecting around the resume that is created on the profile. And, well, the monetization interaction involves a recruiter connecting around the same resume or a user asking for enhanced access to another based on the resume/profile (via InMail).
1) It’s already creepy ‘big brother’ image would raise further privacy alarms
2) Analyzing context from semantics remains an unsolved problem and runs the risk of yielding false positives which would only result in increased noise in the system
There’s a lot of talk about how discovery is moving from search to social i.e. from the query to the status update. However, the problem is that the platform doesn’t own as much information about the status update as it does about the query.
Facebook is still targeting eyeballs. There is no overarching problem with monetizing social media except that the advertising should be targeting interactions, not eyeballs. Companies that target eyeballs are still stuck in the traditional media world. That model worked well for traditional media where users were unconnected silos but in the context of social media, the context of connected users is dynamic and defined by interactions.
Get to where the users are instead of waiting for them to come to where you are.
Feature creep is the surest way to kill your startup.