Facebook CPC – Take Two

When I wrote the last blog post, I didn’t expect it to quite have the reach it did, it seems like many people have been burned by Facebook advertising and it resonated with them quite well. There were a few misconceptions and other things I saw, and perhaps I unfairly targeted my wrath on Facebook (which I don’t have an ax to grind against, and really want it to work!), where really I think this phenomenon is happening with any provider selling mobile ads.

Some helpful folks on HackerNews and Twitter pointed out that you can in fact only target Desktop (and all sort of other detailed placements) using the Facebook Power Editor extension or the Facebook API (obviously!).

I should note that the defaults of AdWords are also sub-optimal, and it’s really easy to lose money on any ad network if you don’t know what you are doing. Many took me to task for being a Facebook ads noob and not optimizing my campaign for Desktop since my site was not mobile optimized.

Of course those people missed the point.

I didn’t want to target mobile but didn’t see any option to not do so, and in fact my ad was quite targeted in every aspect except platform, which Facebook makes non-obvious to do in default editor.
The point of my blog post was that the clicks seemed accidental, and many were not even making it to my site, so even if my site was mobile optimized a large portion of clicks never saw it. The poor conversions were merely what led me to do further investigation.
So, once I figured out how to target Desktop only, I ran a new test, and the results were much different! Every click is accounted for on the server and Google Analytics.

The pages per session, time on site, and signup metrics are on par or BETTER than Google AdWords.

This is where some will say “ah hah, told you! You weren’t mobile optimized!”. But I say, this in fact proves my point, or at least we’re both right in different ways. It’s way more difficult to accidentally click an ad on Desktop than it is on mobile, and you can see the behavior of the ad clicks reflects that, it does not seem accidental at all.

I’m a little tired of burning money on Facebook, but I suppose a third test would be to run a purely mobile campaign, say for installing the BudgetSimple app and compare, but that’s a little more difficult to track in the same way.

Something else that is interesting is that the prices per click are MUCH more expensive on Desktop only targeting. This may be because there is less inventory (although Facebook says it’s around 50/50 from articles I’ve found, and you’d expect their to be more given greater screen space), or there is higher demand, perhaps there are more advertisers out there that know to only target Desktop.

It could also be that Facebook knows this traffic is of a higher quality and charges more. Like most ad networks the pricing is a little opaque. This experiment in fact took me longer to run because Facebook couldn’t fill my demand until I kept raising my bid to about 3X what I paid when targeting all platforms (You can see average CPC is $1.56 compared to the last one which was only $.58).

So what’s the conclusion? In general, I’d say Facebook ads in general are no worse than other ad networks on a CPC basis, and in fact since I expected similar performance to a Display Network type ad, I was surprised when I instead received metrics more on par with Search CPC.

But any advertising that targets mobile platforms should be viewed suspiciously due to accidental clicks, and Facebook seems worse at charging for misclicks than AdWords (remembering again from my last post that Android clicks on AdWords ads all at least registered hits on my server and analytics). So give Facebook ads a test, but make sure you turn off mobile targeting.