First experiment with Facebook Ads

In September/October, I’d posted a couple of Facebook flyers to the Raleigh/Durham and the surrounding College networks. These were sold on a “total impressions” basis, and I spent about $25-$30 for a few days worth of 2500 or 5000 impressions in the community. One of these flyers ran over the weekend of Nov. 2-3 and invited folks to visit the branch’s booth at the Raleigh International Festival where we were recruiting volunteers for an Interpreters Directory. We also mentioned the project in the local paper’s community site, and saw a couple of dozen hits from Facebook and fewer from the paper. During the weekend we had a couple of volunteers over the web, but we didn’t ask “how did you hear about us.”

In early November, Facebook discontinued these flyers and substituted ads, giving different kinds of controls on selecting the population and paying for the results. Late Friday, Nov. 9, I created an ad to drive folks to the volunteer form. I said I was willing to pay $5.00/day and $.25/click. The ad was targeted at those 21 and older in Apex, NC, Cary, NC, Chapel Hill, NC, Durham, NC, Garner, NC, Raleigh, NC, and Wake Forest, NC – approximately 100,000 Facebook members.

Things started well on Saturday morning — 7970 impressions and a couple of clicks with an average cost of $.14. But then things just stopped.

On Tuesday, I increased the max payment per click to $2.00 and saw an immediate surge in both views and clicks. The campaign ran through 11/17 with the following stats. We paid $5/day for the clicks we got starting Tuesday.

Date
DOW
Imp
Clicks
Avg CPC ($)
Avg CPM ($)
Imp/Clk
11/17/2007
Sat
3926 4 1.25 1.27 982
11/16/2007
Fri
5820 4 1.25 0.86 1455
11/15/2007
Thu
5910 3 1.67 0.85 1970
11/14/2007
Wed
10803 7 0.71 0.46 1543
11/13/2007
Tue
17856 9 0.56 0.28 1984
11/12/2007
Mon (HOL)
0 0 0 0  
11/11/2007
Sun
0 0 0 0  
11/10/2007
Sat
7970 2 0.14 0.04 3985

We did get three volunteers that may be attributed to the campaign — one on 11/13, two on 11/14.

I’m not sure we’d do this again, but I’m interested in insights others might have.

  • Any ideas on why the cost/click rose over the week?
  • Does anyone else have cost/click data? While Facebook says it “adjusts” the amount based on demand, is there any documentation on that algorithm?
  • Are our impressions/click (Facebook reports the CTR% which rose over the week) in line with what others see? There’s very little that can be configured in a user-written ad. But could the ad itself have been “snazzier” to get a better result here? Or did the impressions/click and cost/click interact so that there wasn’t much we could tinker with given our budget?

I guess the main lesson I learned was to include a “how did you hear about us” in the form. I’ll also be more careful about tracking Facebook traffic to the page in question.
Comments welcome!

Leave a Reply

Your email address will not be published.