This was the year of Facebook: Mark Zuckerberg was named person of the year by Time, The Social Network movie was a huge hit and is probably the odds-on favorite to win the Oscar for Best Picture, the site passed 500 million users, and rumors have it that the company also made more than $2 billion in revenue. Not too shabby. Amazingly, however, building a great advertising platform is noticeably absent from Facebook’s impressive list of accomplishments.
Facebook’s user interface (UI) for advertising is sub-standard to say the least. Despite the incredibly rich demographic and psychographic data available to advertisers, running tests on different segments of data is simply not scalable in the UI. On top of that, Facebook’s tracking technology was misleading (treating both view-throughs and click-throughs equally) and was recently discontinued altogether. This lack of scalability has resulted in two consequences: first, many advertisers have tried and failed to make Facebook PPC work, resulting in low CPCs and a lot of negative press from advertiser blogs, pundits, etc. Second, numerous 3rd party software companies have integrated with the Facebook API and created tools to make it easier for advertisers to scale on Facebook. Most major PPC management platforms (Marin, Kenshoo, SearchForce, etc) have included Facebook management as part of their workflow automation.
The big question I have about Facebook advertising is whether we need to look at it from a demand creation (branding, top of funnel) or demand fulfillment (direct response, bottom of funnel) perspective. Obviously, if you are marketing to someone who lists their marital status as “engaged” and is a fan of “wedding rings”, you can take a direct response approach to this consumer, much like you would if someone typed “buy wedding rings” into Google. And if you want a quick way to build buzz around a new movie, you can reach millions of 18-34 year olds who like “movies” in a matter of days on Facebook (indeed, Facebook is now the largest server of display ads online, so for many demographics, it’s becoming the mass medium of the Internet).
These are two extremes, however; the bulk of advertisers – and of impressions on Facebook – fall somewhere in the middle. If you take the wedding ring example, the vast majority of Facebook users will not spend the time to express interest in “wedding rings.” When you buy an ad directed to all users who are “engaged”, should you assume that this ad will have the same ROI as your AdWords campaign, or should you think of this as “awareness” that will lead to a conversion down the road?
The answer to this question most likely comes down to proper attribution – if you can actually track a consumer from a click (or view) on Facebook to an eventual purchase days later through, say, a PPC click, then you can give Facebook the right amount of credit. Unfortunately, 99% of online advertisers at the moment have no way to track at this level of granularity (we’ll discuss this more later on).
What I’ve seen in 2010 is that advertisers that try to hold Facebook to the same standard as a PPC campaign will be sorely disappointed – it is definitely a hybrid of “creation” and “fulfillment” marketing. In 2011, I think that smart advertisers will change both their thinking and their measurement of Facebook and find a way to make Facebook work as a meaningful complement to existing PPC campaigns. Here are a few ways I think this will happen:
a. Proxy Metrics. A proxy metric is a “soft” metric, i.e., one that shows a likelihood of a positive outcome, but is not the final positive outcome. For example, a user who clicks on a link to “get directions to store” could be measured as a proxy metric for user intent to purchase something in a store, whereas a shopping cart conversion is a hard metric. Because I think Facebook is a hybrid medium, you need to combine proxy and hard metrics to judge performance. So even if the conversion rate to a purchase is 50% of your average conversion rate on Google AdWords, if you also see that Facebook users have a much stronger tendency to sign up for emails, refer friends, bookmark pages, click the “contact us” button, and so on, you may still discover that Facebook is incredibly powerful for your business. Note that almost all of these metrics can be easily tracked through Google Analytics.
b. Attribution: Online measurement tools will improve in 2011, and part of that improvement will come from better attribution. If advertisers can determine how Facebook impacts downstream conversions, the result may be more budget spent on Facebook ads.
c. 3rd Party Tools: We’ll see more and better Facebook tools in 2011. So far the best one I’ve seen is Alchemy out of the UK, but I’m pretty confident that the paid search campaign management companies will be catching up quickly to the stand-alone Facebook tools (or they’ll just acquire them). These tools will enable advertisers to quickly and efficiently test hundreds of cross-sections of user data – e.g., massive combinations of geo, age, psychographics, etc – as well as many different images and ad text, and identify the sweet spots amongst the millions of users out there. Unless Facebook suddenly comes out with a better user interface (I’m not betting on it), investing in one of these tools will be a must for anyone who wants to spend more than a couple of thousand dollars a month on Facebook.