The self-service ad model may not be the secret to Facebook's future fortunes, but it presents marketers with some largely untapped opportunities for reaching the most precisely targeted audiences
online.
If you talk to Facebook's users about advertising, you'll hear a number of criticisms. Some say it's brash or irrelevant. Many others don't notice it at all. I'd expect many
consumers wouldn't even think of the best ads as advertising, such as those ads for TV show or movie premieres on the homepage where you RSVP if you're going. Some of the worst problems with the
site's advertising have been minimized, such as those in my musings last summer on Facebook's
'Jewhavioral' targeting and overly personal
weight loss ads.
Forrester's Marketing Forum last month provided me with an excuse to run another ad trial, as I demoed the platform to an attendee during a break and wound up creating a
live campaign. I was covering the event as a blogger (read the roundup), so I had
something to offer. Here are a few things I learned in the process. You can view screen shots from the campaign on
Flickr or SlideShare.
Facebook self-service ads remain as easy as ever to create.
While I can't say my ad copy was perfectly written, putting up something passable took all of a few minutes. There have been few changes since the service launched. The
targeting options such as Keywords and Workplaces allow precise ways to reach consumers volunteering this information. I had the campaign target people working at Forrester Research, which brought up
about 400 people in the United States. Most advertisers won't want to cast such narrow nets, but the option's there. Creating similar ads is also a cinch. I used this feature
to create five versions, all of which are shown on Flickr and SlideShare. These include ads targeting Forrester's competitors, and one targeting Forrester employees in the Netherlands.
Expect low CPMs. Recommended bids ranged from 30 cents to 46 cents. I set my bids significantly higher since I was targeting fewer than 1,500 people through the various versions of the ads. I tried
entering various other keywords and targeting options and couldn't find anything higher than 50 cents. More precise targeting does not lead to higher recommended bids. The
actual CPM after a week was 60 cents across the campaign. If Facebook is able to reach its $5 billion valuation, it's not coming from advertisers like me. The performance was
underwhelming, with a 0.25% click-through rate. That could be the fault of the ads. One ad targeted to Forrester employees where I used a better image reached a 0.60% CTR. I also know at least one of
the clicks came from the analyst Jeremiah Owyang, who was featured in that ad. Moral: people will click ads with their own picture in it. But do that too much and you're probably going to creep out
everyone who sees it. Fortunately that targeting's not offered by Facebook directly, though some application ad networks can pull in profile pictures. After going through the process as an
advertiser, I'm reminded of how relevant the advertising can be for consumers. Advertisers know a lot about me from the site, and they can infer a lot more. Fans of "30 Rock," Christopher Guest
movies, and Jonathan Safran Foer books living in New York probably would welcome hearing about restaurants in the city's theater district. The cost of testing these ads is negligible; for around $15,
I amassed over 25,000 impressions, which is plenty of information should I choose to use it for another campaign.
Self-service ads won't be the best option for running a major branding
campaign on Facebook. The engagement ads on the homepage serve that purpose. With this kind of targeting, though, I'll fork over that quarter per click when there's a reason to reach a segment of
Facebook's audience.