Dating on FB is not an easy gig, I'm not afraid to admit. Jumped headfirst into it about 1.5 months ago, conclusions:
1. Dating ads are extremely sensitive to day parting and week parting. As Finch eloquently expressed on his blog, no one is thinking of shagging during breakfast.
2. Even if its easy to hit 50-80% ROI in foreign language markets, the volume on FB is just not there to scale reasonably. Especially with 30+ age restrictions etc. I'm sure those 12,000 Danish guys have seen it all by now, especially when there are about 2 offers that are even relevant there.
This is the classic debate of: do you want to take a piece of a small, possibly less competitive pie (50k FB demo @ 6$ payout = 300k pie) or your own small piece of a huge pie (Dream Marriage I'm looking at you)? In the venture capital world, they always pick the latter. This all depends on what you consider an acceptable daily scalability ceiling.
3. Image is very important, yes. But so is the headline. On top of that, the promise/benefit in the headline should probably be in line with the promise/benefit in the ad copy. 25 characters for headline and 130 for ad copy pushes your Ogilvy-meter to the max.
4. Get some CPM going (15k at least) for each ad variation, and make conclusions. Higher CTR DOES NOT MEAN HIGHER Conversion Rate. Don't mess this up. Curiosity clicks are bad. Convert ads to CPC to get steady traffic on FB. CPM gets throttled hard after a couple days. Track everything religiously, use prosper202.
5. If you want to spend 50$ in the next 4 hours, your daily budget should probably be at least 50 * (24 / 4), if you know what I mean.
6. Targeting a 6 figure demo isn't going to work. Maybe we missed those happy days. High 4 figure, low 5 figure demo with keyword/workplace targeting works. I'm still learning about how to properly scale. There's the approach of getting one city (Toronto) profitable, then replicating the campaign nation wide with the gleeful assumption that good ads are universal. Or there's the porting of entire campaigns to various English speaking countries, or even translation to other languages. Or there's the possibility of getting ads for all 24 year olds profitable, then copying it for 24 +/- 5. I'll probably lose some money learning this properly.
This is just in my personal experience in the past month and a half. It's easy to lose a lot, but if you don't spend and test for yourself, you won't learn jack. Not even after 100 pages of an FB ads thread, which is 90% complaints about the FB system/approval/interface.
If I can't break 100% ROI on FB dating, I'm going to need to diversify into other verticals. Dating just seems to be such a good fit for what I perceive to be the mental state of the user while they are browsing a social network.
