Facebook’s f8 announcements could have been designed exactly to address the worries we recently aired about how the social network works for brands. Facebook is striving to be effective for brands in the way in the way Google is, by being useful without being intrusive. Jon Prideaux of Finextra hits the nail on the head. Google has solves the problem of intention for advertisers in a very practical way. When you search you look for the answer to a question and there are ways in which commercial services can add value.
Social networks have a dilemma in formulating an offer to brands. Social networks are about direct communication between individuals. How can a brand add value to this without appearing obnoxious? A brand has to wait unobtrusively with practical assistance, taking cues from the thematic context (shared links and topics). It risks being irrelevant, or misinterpreting the context at every stage. I sat next to a woman yesterday who was offended by being offered weight-loss treatments on Facebook. She has never expressed any interest in the topic. The ad is most likely being served based on a rather crude demographic cue – she is single, early thirties.
Considering the richness of the data Facebook is sitting on, serving weight-loss adds based on the age and sex of the user is not very sophisticated or – as we see from the low click-through rates – fruitful. The challenge that Facebook has set itself is that since f8 2010 its user settings now prohibit it from using more sophisticated methods (such as the tastes of the consumer) to serve ads UNLESS THE USER HAS OPTED IN TO AN APPLICATION. To spell this out, the key value Facebook could offer to brands is now closed to it.
Mark Zuckerberg’s f8 2011 announcements make it clear that Facebook too is coming to believe that its value lies in the hands of those (like Zynga or Spotify) who can use the network as a platform for their gaming or entertainment-based applications, offer entertainment to user in return for their data and then use the data to serve far more targeted advertising based on collaborative filtering (people who like X also like Y). The network (Facebook itself) could serve as a CRM repository for an ecosystem of application providers.
Facebook’s job is to keep those application providers on its platform (rather than moving off to iOS or Android) by adding so much value with its CRM data that it is worth content providers paying the “tax” to use the platform. It is this thinking which sits behind the controversial idea of “frictionless” (i.e. constant, involuntary) sharing.
If Facebook can persuade users that this trade-off (privacy for useful suggestions about cool stuff) is worth making, then this would be a good outcome for the business. When I worked at Amazon.com the collaborative filtering links post-purchase were by far the best performing on the website. The approach can work.
There are two challenges: one is the users themselves. Amazon’s collaborative filtering is anonymised, discreet. Facebook are proposing something in your face. Mary likes The Saturdays, Tom likes Blazin’ Squad. Not surprisingly this feels a bit intrusive to many users.
The second challenge is that serving ads using collaborative filtering techniques requires an ecosystem of “bought-in” vendors who are happy to share their data in order to benefit from the increased understanding it offers. Furthermore the big issue for Facebook is that this is not currently a business opportunity open to it. Only a successful application provider with really compelling content – an HBO or Warner Brothers – can take advantage of this, or an advertising network built on the back of a variety of such vendors.
When Google was in the same position in 2000 it benefited by from the work of Bill Gross, who brilliantly invented paid search at Overture and inspired the web’s most successful current business model. Facebook lacks a Bill Gross. In the meantime, and at the risk of sounding like a broken record, you would have to say that Facebook still lacks a business proposition to make it investable at anything like the mooted $75bn.