We all know about the importance of advertising on facebook if you’re targeting any of its almost 1Bn users. But there was something that was bugging me for a while, I’ve been running a lot of ads on facebook for different clients with different objectives, but I’ve always preferred CPC over CPM, I felt that paying for an action is much more reliable then paying for display.
Now some people might argue that CPM does well for branding, getting the word out and making people notice you. That’s good if you have the money to spend or if your objective is to waste it. But generally I never do CPM until recently, I thought I should take a look at it from a scientific point of view (and by that I mean using excel!).
So I did some research about the average cost of advertising in KSA and the UAE, and found the following numbers on social bakers:
Facebook advertising costs (in USD):
Country | Avg. CPC | Avg. CPM |
UAE | 0.48 | 0.10 |
KSA | 0.40 | 0.9 |
As you can see the prices are close for both UAE and KSA, the rates have recently went up, because when I was doing the math last month the price for CPC was 0.29 USD and 0.06 USD for CPM in KSA, but I don’t think that should change the point I want to reach at the end of this article.
Ok now that we have the prices out of the way, we have to find the average click through rates (CTR) for ads on Facebook and after a quick google search I found a number of sites and forums with similar numbers. But I liked the numbers mentioned in this Quora answer just because the contributor of the answer claims to be a facebook API partner.
Facebook ads Average CTR:
Performance | CTR % |
Optimal | 0.10 – 0.16 |
Above Average | 0.07 – 0.09 |
Average | 0.04 – 0.05 |
Below Average | 0.02 – 0.03 |
Poor | 0.01 |
So lets take the following scenario, I want to carry out an advertising campaign on facebook with a daily budget of 10.00 USD, what would preform better?
Assumptions:
I went with the higher costs of the UAE and the CTR of average performing ads of 0.05%.
Avg. CPC | Avg. CPM | Avg. CTR | Daily Budget USD |
0.48 | 0.10 | 0.05% | 10.00 |
lets see how many clicks and impressions we would get with each, starting with CPC:
CPC Numbers:
No. of Clicks | No. of Actions |
20.833 | 6.042 |
Before I can move forward I have to explain how I reached these numbers, since we have the budget we can work backwards to calculate the number of clicks, and since its CPC the number of impressions is not very important so we don’t have to calculate it.
No. of clicks = Daily budget / Cost Per Click -> 10.00/0.48 = 20.833 clicks
You will also notice that I added a second column called “No. of actions” this is a number to track conversions, for this example I gave it a value of 30%. Basically from every campaign you run, not a 100% of the people who click on the ad actually convert to customers or do whatever you want them to do (like a page, register for a newsletter, fill a survey, etc…). So if 30% of the number of clicks actually carry out the required action it will brings us to a total of almost 6 actions.
CPM Numbers:
No. of Impressions | No. of Clicks | No. of Actions |
100,000 | 50 | 14.5 |
Now for CPM the calculations are a bit longer, because we have to find the number of impressions first. But again we can find out all the information since we have a budget and we’re assuming a CTR of 0.05%:
No. of Impressions = (Daily Budget / Cost Per 1000 Impression) * 1000 -> (10.00/0.1)*1000 = 100,000 Impressions
No. of Clicks = (CTR * No. of Impressions) / 100 -> (0.05*100,000)/100 = 50
Whoa! On paper you can get more than double the number of clicks if you do CPM!
So I planned to run two identical campaigns on facebook each with two ads (total four ads), the only difference between the campaigns was that one was CPC and the other CPM.
I decided to run it for eight days and wanted to see which would preform better, so before running the test I did some math using the same numbers above to have a rough forecast of what to expect. I also went a step further to see how much I would pay for a single action, here are the numbers I got, again starting with CPC:
Campaign forecast for CPC:
No. of Days | Total Cost USD | Total Clicks | Total Actions | Cost/Action USD |
8 | 80 | 166.667 | 48.333 | 1.66 |
Campaign forecast for CPM:
No. of Days | Total Cost USD | Total Clicks | Total Actions | Cost/Action USD |
8 | 80 | 400 | 116 | 0.690 |
An action costs 1.00 USD less with CPM? But how can this be true, were all the campaigns I ran a waste? Could I have done better going with CPM? You can imagine what was going through my mind as I was reviewing the numbers.
Now back to reality, The campaign was targeted to different countries in the MENA, and the action required was for people to fill out the Online, Social Media & E-Commerce usage in the UAE survey. Each campaign had two advertisements as mentioned above, the advertisements were similar in every detail except one linked to the article page on our website and from there users would go to the survey. The other ad was linking directly to the form page which was setup on google docs. Below are screenshots of the ads:
I realize its better to test different ad formats to see which preforms better, however this would have effected the results of what I was trying to do, so I tried to minimize all variables as much as possible.
And here are the results of the campaign (I took out the social clicks and reach):
Facepalm moment!
First of all the numbers don’t match with the forecast, because of the following reasons:
- The actual cost of media was much less than the assumed price: 0.07 vs. 0.48 for CPC and 0.02 vs. 0.10 for CPM
- The CTR was also different between predicted and actual results: 0.08% vs. 0.05 For CPC and 0.01 vs. 0.05 for CPM
- Targeting features on Facebook, both geographical and behavioral
Ok this is by no mean a definitive study to show that CPC is better, especially since facebook serves different ads according to a number of variables, but below are the learnings that I will take from it you’re welcome to give your own thoughts:
- I’m sticking with CPC
- I forecasted a 30% Action rate for conversions, when in reality it was more like 3% for this example. But to be honest a 30% conversion for filling a survey is a little bit optimistic especially users are getting nothing in return. Had it been a competition or a giveaway things would’ve been different
- I managed to get a 0.08% CTR for the CPC campaign without trying to optimize it. I’m pretty sure if I had different formats with different headlines, images and copy I would’ve achieved a higher CTR and a lower cost per click. The best I’ve ever managed to achieve on a CPC campaign was a 0.03USD. But it was a giveaway competition campaign
- Something not in the screenshots above is the “Frequency” or number of times the adĀ appearedĀ to the same user. The average for the CPC campaign was around 3.5 and for CPM was around 10.5. So if you want the same users within your target to see your ad, then go for CPM other than that dont do it
- If you’re paying for advertising, might as well pay for an action. Because the CPC campaign turned out to cost 0.05USD per thousand impression which is 0.03USD higher than what I payed for the actual CPM. But the cost per click for the CPM campaign was 0.18USD which is 0.11USD more expensive then what I paid for the actual CPC
What do you think? Do you have any tips or advice from previous experiences on advertising on Facebook? Then let us know. If you wish to receive tips like this to your inbox signup to receive our newsletter.