Kelly Maher

Facebook ad experiment

First, I want to thank everyone who saw my ad and decided to like my Facebook page! The rest of this post will probably be only of interest to those who are exploring whether or not to run ads for their pages 🙂

I decided to run an ad as an experiment after hearing Mandi from Smexy Books talk about running a similar experiment at the Frederick Book Festival last weekend. She got the idea from Smart Bitch Sarah at the panel they were on together at RT (which I also attended). Mandi stated her two day, $10/day experiment generated about an additional 75 likes for her page. I asked her if she had done any additional targeting of the ad, and she said she went with the defaults. With that information, I chose to start the ad after I got home from the day job on Friday.

The purpose of this experiment was to see how many likes a basic Facebook ad would generate. My budget was $20, and I went with the default setting of $10/day so it would run for two days. Since the RT convention, I’ve been more deliberate about building my presence on social media sites. I tend to prefer Twitter over Facebook for a number of reasons, partially because alter ego has a large Twitter presence and it fits my general communication style better. What this means is that even though I’ve had a Facebook page for probably two years now, I had four likes when I started the campaign. One of those likes was myself, which I did right before I started the campaign. Turns out that even when you’re the admin for a FB page, you aren’t automatically set up to receive updates from it in your news feed. :headdesk:

The ad campaign started around 5 pm Eastern time Friday night. The only change I made to the default settings was to include the following countries in the target range: Canada, United Kingdom, Ireland, Australia, New Zealand, and Germany. The English-speaking countries are probably no brainers, but why Germany? First, Germany likes their erotica. Second, Lust at First Bite has been translated into German. By midnight Eastern, I had 60 likes, 53 generated from the ad, and 3,171 impressions. Now, I believe Facebook’s timing is Pacific-based, so their final report is a bit different from mine. They have it as 62 likes generated from the ad on Friday.

I closed the ad campaign at midnight Eastern Saturday night. The cost of the campaign was $19.96 with the whole daily cap of $10 spent on Friday. Per Facebook’s stats, an additional 18 likes were generated on Saturday for a total of 80 likes from the ad. There were 4,736 impressions with a campaign reach (unique people who saw the ad) of 4,479.

Stats from the Facebook report:

  • Cost per click: $0.15
  • Cost per 1000 impressions: $4.21
  • Campaign reach: 4,479
  • Likes generated on 5/24/13: 62
  • Likes generated on 5/25/13: 18

Facebook split the campaign into three ads. The most successful generated 79 likes for the cost of $0.22/like ($16.83 total) had the following parameters:

“This ad targets 146,866,540 people:

  • who live in one of the countries: United States, Ireland, New Zealand, Canada, United Kingdom, Australia or Germany
  • who are not already connected to Kelly Maher

The other two ads somehow generated two additional likes (one each), though one of them doesn’t seem to count as the total generated from the campaign was 80, not 81.

As of the writing of this post, I’m at 89 likes for my page. Overall, I’m pleased with the results from this experiment as my goal was only to see how many likes an ad campaign would generate. I find it interesting that the bulk of the generated likes came in on Friday night. If I were to run another campaign, I will probably restrict it to a one-day run, and on a Friday night. I may experiment with other nights, but I think the shortened time frame of when I started the ad (5 pm Eastern) caused FB to show the ad more frequently in order to use up my daily allowance. I hope this is helpful for anyone considering running FB ads!

Leave a Reply

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>