What happens when you use different ad objectives [test]?

Do you love testing Facebook ads?

Since it was voted as the favourite tip among you guys last week, I planned a series of Facebook ads tests for the next few months.

In this article, I will be sharing my results when I ran ads to test the performance between different ad objectives.

For some time, I have been hearing many Facebook ad managers say that they were finding better website conversion rates when using the post engagement objective instead of the website click objective.

Do note that in these cases, they couldn’t run website conversion campaigns because they do not yet have a record of conversions that Facebook can use to optimise.

So being the curious cat that I always am, I decided to run two ad campaigns, one with the post engagement objective and the other with the website click objective.

Test Design

As you already know, I created two campaigns with different ad objectives. I made use of naming conventions to distinguish both campaigns:

  • Jason How - J Social - Blog post - Post engagement
  • Jason How - J Social - Blog post - Website clicks

Within each campaign, I created two identical ad sets, one targeting my website visitors and the other targeting competitors’ fans. Again, I followed Jon Loomer’s naming conventions to distinguish between the four ad sets:

  • Website visitors - Post engagement
  • Competitors’ fans - Post engagement
  • Website visitors - Website clicks
  • Competitors’ fans - Website clicks

Within each ad set, I created one identical ad. Why only one? Because the point of this test is to see if there will be differences in ad performance between two different ad objectives, so it won’t matter whether I have 1 or 10 different ads.

The post engagement ad is on the left and the website click ad is on the right. Because I couldn’t promote a link post for engagement, I adapted the ad creative to insert a link in the text copy and added an image as well.

Use of two conversion pixels

Since the main goal of both campaigns was to drive email conversions, I placed a conversion pixel on the “thank you” page after someone signs up on my website.

But I also wanted to know how many page views were driven by these four ads, so I decided to create a Key Web Page View conversion pixel and placed it on that page only. I did this using the Facebook conversion pixel plugin on WordPress.

The test results

After spending $46.19, here are the results so far.

1. Email signups

  • Website click ad (competitors’ fans): 0
  • Website click ad (website visitors non-fans): 0
  • Post engagement ads (competitors’ fans): 2
  • Post engagement ads (website visitors non-fans): 2

The results do not look like much at this moment. But considering that website click ads should drive more website traffic and hence more conversions, it surprises me that post engagement ads are driving the conversions. Does this mean that post engagement ads are driving more relevant traffic to my website?

2. Website clicks VS Key Web Page Views

  • Website click ad (competitors’ fans): 15 VS 11
  • Website click ad (website visitors non-fans): 13 VS 9
  • Post engagement ads (competitors’ fans): 2 VS 4
  • Post engagement ads (website visitors non-fans): 5 VS 7

As expected, website click ads did drive relatively more website traffic than post engagement ads. But did you notice that the key web page views for website click ads were lower than their website clicks?

On the contrary, do you see that the key web page views for post engagement ads are higher than their website clicks?

I can think of three reasons why this is so.

First, someone could have clicked on my ad and closed it even before the conversion pixel managed to load and fire.

Second, someone who has read my post could have returned again, adding to the key page view count.

Or both things could have happened. For example, someone could have clicked on my post engagement ad and closed it before the conversion pixel loaded and fired. At the same time, one or more people who clicked on the ad came back to read it.

So this result suggests two things:

  1. Website click ads are driving more clicks because Facebook optimises them by showing my ads to people who click on lots of links.
  2. In contrast, Facebook is showing my ad to more people who like, comment, share, and click on links with post engagement ads.

This deduction seems to support the claim that Facebook is driving more relevant traffic to my website with post engagement ads - people who are genuinely interested in consuming content and not just clicking on links.

But of course, I’m aware that this is just a small sample.

To better compare the results, I need a bigger sample size. Preferably somewhere in the region of 10,000 people reached with more than 100 email signups. So…

Why am I sharing the results now?

As someone who frequently fiddle around with my clients’ ads, I couldn’t sit still when I saw the subpar ad performance. As you probably noticed, the email sign-up rate (email sign-up count/reach) was a lot lower than 1%.

I wanted to change things around, use a way more visually attractive image and write more compelling copy now that I have a better idea of who the article can benefit.

That is why I’m going to restart this test, spend $200 to $300 on it and see if my premature deductions are right.

If you’re interested to see my findings, then enter your email address in the sign-up form that appears after this post. I don’t want to spam you with stuff you don’t want to see, so you will only receive my findings if you let me know that you want it. If you sign up using the form below, you will not receive any other emails from me except the new findings I just mentioned.


What about you?

Are you seeing better off-Facebook conversions when you use different ad objectives? Maybe you ran a couple of ad sets with different ad objectives but never took notice. If you vaguely remember doing so, go to your Facebook ad report and take a look now. Do you see different results between the two campaigns?

Let me know in the comments below - I’m excited to hear about your results!

  • http://www.antoniocalero.com/ Antonio Calero

    Good stuff Jason, very informative. We should talk biz someday…