I thought it would be interesting to develop a case study comparing the type of picture to the click-through-rate of a given ad. This study ran for a few days and I ended up spending just under $100 gathering the data.
I’m working on a handful of other case studies to publish from Facebook, PPV, and potentially Myspace/POF traffic. As I said before, case studies are a great learning tool and I love analyzing the data. Here you go!
Basic Info and Demographics
I created six groups of five ads. The ads were targeted towards single men ages 45-60, using pictures of women. The ads were served over 150,000 impressions, so there is definitely room to gather more data. However, the preliminary results are promising and do fit trends I have seen in the past with my ads.
The Offer
I did this study using the Real Mature Singles offer at EWA Network.
The Picture Types
- Black/White (no color)
- Not Smiling
- Wearing a Hat
- Wearing Glasses
- Looking Away from Camera
- “Myspace” Type Pictures
For the purpose of this study, I am comparing the best performing ad in each campaign, for a total of 6 ads, one from each picture type.
According to my study, men are most interested in clicking on pictures of women with glasses, followed by pictures of women looking away from the camera.
The Ads
Here are the actual ads I ran for this study. I made sure they were very “clean” so they would get approved. With some work, I’m sure the CTR can be improved.
Profitability
This campaign ran slightly under profit during this case study. Again, with additional optimization, I’m fairly certain this can turn into a profitable campaign.
Conclusion
Of course other factors affect the CTR of the ad, but the results in this study were generated keeping literally everything the same aside from the picture. If you have any experience using Facebook Ads, you know how much your ads may fluctuate, so keep that in mind as well.
Let me know if you have any questions about this study or requests about other studies I can try.
Have a great weekend!
Related posts:
- SEO Case Study – Intro (Part 1)
- SEO Case Study – Backlinking Strategy (Part 3)
- SEO Case Study – Factors (Part 2)
{ 30 comments… read them below or add one }
Ha! Thats awesome.. my guess is that the glasses carries a somewhat sexual innuendo if you know what I mean.. but I really didnt think it could make it the winner!
Thanks for the stats!
PS: Did you find that the glasses one brought in the most amount of leads too?
Dino, yes that one did bring in the most leads. It was consistent but also received the most traffic.
One of the pictures I used had a woman with glasses AND a hat – She did very well too!
How about a women with glasses looking away? I love these kind of studies – that put real analytical results to a bunch of numbers.
I’m not in the 45-60 demo but I would think putting attractive somewhat older women in one of the test ads would have had an impact too. It would have at least been interesting to see how it performed. Also, attractive but somewhat plain looking women always catch my eye. They just make the ad seem more real than the 21-year-old super model types do. But who knows?
Nice research..
Thanks for the tips Matt. I’ll keep that in mind and perhaps incorporate that into a future study!
Nice case study Jeremy. So what would you recommend to make this campaign profitable?
Testing additional images with a higher CTR to lower costs would be my goal for this campaign.
Thanks for sharing Jeremy. You should follow ShoeMoney’s tip about cleavage and see how that effects the CTR and conversion rate. Throw an older picture in there also and see how that works on that demographic.
Mike,
Thanks for the tips man. It’s a pain to get cleavage pictures through now. I actually got warned by Facebook awhile ago because my pictures were of women who were “too busty”
great case study, and I like how you broke down the CTR in a graph.
It would also be interesting to do a case study between this and ads targeted to your specific demo of ages 40~60 … like pictures of older woman or mentioning “mature” “senior” in your ad copy and see how they match up
imKazu – Thanks for the comment. So many different variables can be tested. Maybe I’ll put together something with all the suggestions here and see what works best!
Is there some magic to getting dating ads approved on FB? I’ve tried so many times.. every time, i get some stupid generic error like “You have to make it clear that this is a dating ad”
FreshSuperCool, it’s all about being persistent. Don’t use stuff like “Come See Pics” as your titles and you should be fine.
Hey Jeremy – thanks for sharing these results. It would be interesting to do another test at some point with another set of photos and see if the trends stay the same on things likes wearing glasses, looking away from the camera, etc.
Tom,
Thanks for the comment. That’s a great idea, gather more data to determine whether or not it was a fluke or if it is actually a solid trend.
I’ll keep that in mind!
Jeremy,
Awesome post! This is what online marketing is all about! Careful test construction and great analytics. This is the type of framework that really creates winners. Well done!
Best,
Ian
Thanks for the comment Ian. I definitely agree. It’s very hard to be successful without being careful and using solid analytics.
What a great site. Great tips, thanks
Thanks Tom, glad you appreciate the tips!
I tested a ton of ads last week with this post in mind – both POF and FB. Suffice to say they really are making it difficult to get ‘cleavage’ pics through. All my US, CAN, AUS targeted ads were spat back at me but, oddly, every UK targeted ad was approved!
On another point… as I look for no-cost advice
… should I scale up a high CTR pic by upping the campaign budget or the CPM bid?
Hey Chris, glad I gave you some inspiration!
I would only scale a high-CTR pic if it is making you money — no point in spending more money just because you are not paying much per click. If you have the budget, try to get as much traffic as you can while remaining profitable. Set an ROI you would like to achieve and fluctuate your bids based on that — If you start missing the mark, make additional adjustments.
Good luck!
Thanks Jeremy.
I’m happy with the ROI on these ads I just can’t seem to generate any decent volume of traffic. I seem to be stalling on a CTR of 0.2 with, say, 50,000 impression.
The other day I had an ad with 150,000 imp, 0.850 CTR but only 5000 uniques. I’m wondering if I’m under-bidding on the clicks?
Should I be bidding $0.50 instead of $0.25 to get more volume?
Still very new with this stuff.
Test it out man. You may get more volume with a higher bid or you might be getting as much as you can right now.
Not much I can say other than just start playing with the numbers.
Just ran into the site a few days ago and signed up without hesitation after an hour of reading. Really great quality stuff on here man.
I’ve started advertising on POF and I’ve yet to build a profitable campaign. Even though it’s only been like 4-6 days since I started my first one lol. But I’m just wondering how much you’ll let an ad copy run before you stop it and move onto making another one? And do you just throw the picture of the bad ad away or use it with another copy again?
Hey Michael, glad you like it here!
Usually I’ll spend 1-2x payout on an ad before I trash it, but sometimes sooner if the CTR is terrible.
If I think the picture might do well I’ll give it a handful of tries, otherwise I will get rid of it.
That’s another thing I love, you answer us! Lol. Most other blogs the webmaster doesn’t care to answer.
Ok well looks like I’m trashing 3/4 of my ads. But how do you tell which particular ads in that campaign the conversions come from?
And is there a general amount of impressions that you will let it run at least before being certain you want to trash it?
Thanks for the replies!
No problem. People won’t keep commenting unless there’s some kind of discussion.
To track conversions you need to be running some tracking system, whether that’s manual or not.
I let it run for 1-2x spend usually, the impression count varies depending on bid type, bid cost, and CTR.
If you need additional help, just contact me and I’ll do what I can.
Concise and written well, thanks much for the post
Glad you like it Marisa, thanks for stopping by!
There are way too many variables to consider.
Contrast, positioning, type, etc.
You’ve still produced an excellent study. Keep up these posts!
I agree about the variables, but you have to start somewhere. Thanks for the comment and the encouragement!