Facebook Ad Testing

2008 October 19
by Barry

Now there’s something I completely forgot about, Damien Mulleys Facebook Ad thingy, I actually stumbled across Gordon Murray’s blog, he had also done some testing on Facebook. Luckily I did set aside a folder of assets, notes and stats.

So I went slightly over budget (yeah I forgot to pause the campaign) and ended up spending $117.56, but for purposes of sharing the love, here’s the key points

  • CPC better than CPM
  • Don’t trust suggested bid/spend
  • Relevant/Targetted works best
  • The time taken by Facebook to approve Irish Ads is annoying
  • Only useful when you have something relevant to offer after click

First of all I knew I would get high click through rates, as I was advertising my Dublin Movie Club (the facebook group not the website)

Here’s a quick run down, details might be lacking as I’m just going from my rough notes of 3 months ago.

1. Ad Setup
As I was advertising a Group which is on Facebook, adding the ‘Social Actions’ really helped the viral aspect of the Ad. If someone on your friend list has joined whatever you are promoting – a small endorsement appears beside the ad showing that they’ve joined. Nifty eh?


I set most of my ads up on a CPC basis, starting with the minumum amount, and raising by $0.02c when required to trigger enough impressions, this ensured lowest possible cost per click. If I had of used CPM on my most successful ad, I would have paid around around 120% more.

Facebook recommended around $0.95c for clicks, while I started at $0.05c and worked up, in CPM ads it’s a similar case, start low and work up until you hit a point where you are getting low cost clicks.

Unlike Adwords, I’m not sure if there is enough published details to take into account ad performance or even anything similar to the Adwords quality score and ranking algorithm, but it’s a bit of a shot in the dark.

2. Ad Content
I played around with various types of ads, not nearly as much as I wanted, when using other ad systems like Google Adwords, I would create variations of an ad with single differences to determine which was more successful. Frankly the turnaround on getting ads approved was almost a day, and I wasn’t prepared to do that only to find an ad or variation had been rejected.

It’s worth noting that whoever approves the ads are not working on GMT or European time, that’s something to bear in mind if you ever wanted to create a quick reaction ad to something topical. Create your ad and wait the 12+ hours for it to go live, hmmm – maybe with facebook operations opening in Dublin this will improve.

The group were heading to 2 movies during the test period, both which were visually strong in image ads, and hey it also helps if one of them is ‘The Dark Knight’

3. Ad Targetting
With no effective method of targetting regionally, which would obviously help my movie group based in Dublin, I toyed with ads targetted to Dublin colleges, companies, movie related interests, but in the end – I decided to leave it running generally as the cost was low and I was tired making ad variations :)

4. Got clicks, and then what?
The primary goal of the ad campaign was to get people to join the facebook group (which in turn helps CTR with the social actions I mentioned earlier), and the secondary goal of getting them to signup to the website.

In terms of measuring group member joins, Facebook insights doesn’t show this, but from the 438 clicks I got during that period, 93 visitors were referred from Facebook, of which 21% (or 20 people) converted into new signups.

On my facebook group ‘landing page‘ I don’t hugely promote the website link, merely provide it, so this kind of stat is very encouraging. Imagine if this was a product or service, which was relevant and we were seeing 20 new signups/purchases even after an unoptimised landing page which doesn’t really promote the conversion :)

5. Conclusion
The lack of reporting from Facebook is being addressed with new campaigns reporting on social clicks and impressions. The time taken to get an ad approved hasn’t changed for me yet.

Facebook is really good at delivering low cost traffic, mix it up with a good social precence and relevant targetting and it is something which works really well.

Here’s some shots of the other ads used.

I'd LOVE if you shared this:
Talk to me on Twitter!

I'm @barryhand on Twitter.
Talk to me, that's how I find people to follow :)
Click here to find me on Twitter.

View Comments leave one →
  1. October 22, 2008

    Good review Barry, I too found the ad verification a bit slow, and when I started off my test in August it was a lot slower. I did add a few more ads in Sept/Oct. and the delay was shorter and they did have a feature that had a suggested text (normally I forgot punctuation) and you could click to accept that and your ad would go live without another round of approval.

    I had a product that I targeted at Irish students, in the end I had 195 clicks for my $100 , Avg. CPM ($) was $0.38 and I had about 13 conversions netting $260 which I was quite happy with.

Trackbacks and Pingbacks

  1. Damien Mulley » Blog Archive » Fluffy Links - Wednesday 22nd October 2008

Leave a Reply

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS

blog comments powered by Disqus