Facebook Ads - Are They Worth the Money?
by Rich Deem

Introduction

As a United States citizen and a Christian, I feel we are compelled to promote not only the gospel of Jesus Christ, but also a Christian worldview as a viable choice among all other alternatives. For this reason, I am involved in politics, mostly at a local level. Being somewhat of an Internet guru, I have done considerable research and experimentation with Internet advertising services. Initially only being involved with Google AdWords, I realized a while back that Facebook has an enormous following and that their advertising services might provide a great way to direct our target audience to our political website.

Scenario

I was working on a local referendum in which a mining company had been granted the rights to destroy Van Tassel Ridge in Azusa (to mine rock) in exchange financial incentives to the city of Azusa. Facebook ads promised to specifically target people in Azusa for the ad campaign. I set a budget of $100 for the last 27 days of the campaign. The first thing I noticed was that the click-through rate was almost exactly going to use the budget. In fact, at the end, it had spent $98.89. I thought it strange, since my Google ads had never been so close to budget.

Checking up on Facebook

I also had Google Analytics installed on the website (nominingexpansion.org) to keep track of the Facebook ad referrals. The table below shows information about the Facebook ad referrals:

# City Visits Pages/Visit Avg. Time on Site % New Visits Bounce Rate
1. Azusa 59 1.58 00:01:04 89.83% 71.19%
2. West Covina 33 1.94 00:01:19 66.67% 66.67%
3. Covina 24 1.38 00:01:29 41.67% 70.83%
4. Los Angeles 14 1.00 00:00:00 42.86% 100.00%
5. Beverly Hills 11 1.18 00:00:04 0.00% 90.91%
6. Baldwin Park 2 2.00 00:00:17 50.00% 50.00%
7. Apollo 1 1.00 00:00:00 100.00% 100.00%
8. Bellevue 1 1.00 00:00:00 100.00% 100.00%
9. Claremont 1 2.00 00:00:30 100.00% 0.00%
10. Corona 1 1.00 00:00:00 100.00% 100.00%
11. Eugene 1 1.00 00:00:00 100.00% 100.00%
12. Glendora 1 2.00 00:02:04 100.00% 0.00%
13. Hacienda Heights 1 1.00 00:00:00 100.00% 100.00%
14. La Puente 1 1.00 00:00:00 100.00% 100.00%
15. Managua 1 2.00 00:00:01 100.00% 0.00%
16. New York 1 5.00 00:04:46 100.00% 0.00%
17. Newhall 1 1.00 00:00:00 100.00% 100.00%
18. Norwalk 1 2.00 00:00:54 100.00% 0.00%
19. Pasadena 1 2.00 00:00:26 100.00% 0.00%
20. San Francisco 1 1.00 00:00:00 100.00% 100.00%
21. Upland 1 2.00 00:00:06 0.00% 0.00%
22. West Hollywood 1 1.00 00:00:00 100.00% 100.00%

Although all the users were supposed to be from Azusa, most came from surrounding areas. Since we were targeting Azusa voters, these referrals would be of no use to us. Although it is possible people were accessing their Facebook accounts at work, it seems unlikely that so many would work in Beverly Hills and Los Angeles. Another concerning thing is the average amount of time spent on the website following a Facebook ad referral. For the 25 people in Los Angeles and Beverly Hills, the average time spent on the site was less than 2 seconds! At least 23 referrals spent 0 seconds on the website. I am not sure how this is possible, but it seems bogus to me. Even for those referrals that probably represented real people, the quality of those referrals was much lower than the site average and much lower than comparable Google ad referrals. For example, average time on the website following a Facebook ad referral was less than half of normal, along with half as many pages/visit. In addition, the bounce rate was nearly double. Clearly, Facebook ad referrals in this instance were of very poor quality.

Conclusion Top of page

No, I won't be using Facebook ads any time in the near future. The are definitely not worth the money and were of very poor quality. Were some of these referrals padded with outside ad showings? What about all the visitors who spent 0 seconds on the website? Maybe Facebook can answer these questions.



http://godandscience.org/general/facebook_ads.html
Last Modified February 7, 2011

 

Rich's Blog