International Dating on Facebook (26)
International Dating on Facebook
The purpose of this case study was two-fold:
1) Demonstrate the process of getting ad copy translated for international offers
2) Compare CPC vs CPM bidding on Facebook
Part 1 - Setup
This was a direct link Be2.com offer for Argentina traffic.
Demo:
Females in Argentina, 30-45 exact match, single, interested in men - 58k users
Males in Argentina, 30-45 exact match, single, interested in women - 126k users
Offers:
Three offers were tested from two networks –
NeverBlue 11461 - $0.83 payout
Eagle Web Assets 140918 - $0.47 payout
Eagle Web Assets 140917 - $0.47 payout
Important Note - This is not the necessarily the best campaign to be running (dating in Argentina). While cheap traffic (after optimization) can compensate for the low payout, beginning caps for the offers are 75-100, so you have to deal with that. I ran this campaign previously and was slightly above break-even after 600+ conversions.
This is the LP users were sent to:
I’ve always direct linked. I use “noIPfraud” cloaker to avoid any issues with regional re-directs. On the “fake” page, I still send them to the Be2.com offer, but to a universally accessible page.
CPVLab Setup:
For those interested, here is a 3-minute screencast on how I set up in CPVLab.
http://www.screencast.com/t/kKtsOeA2A00N
Ad Copy:
Two ad copies were used:
Local Singles? Find local singles with Be2. Search profiles for free. It's easy. Just click here.
In the first ad, the appropriate text (girlfriend/boyfriend) was used per the demo.
The text was translated using www.onehourtranslation.com. Watch the video below to see me go through the process step-by-step [15 min].
http://www.screencast.com/t/diTCMe7L2q
Images:
5 images were used for each demo. The final ads are shown below.
(Note, both ad copies were used for each image. Only one is shown for brevity.)
Building Links:
For those who don’t use an uploader tool (I use Spyder’s and the new XY7 AdOptimizer), here is a screencast on how to use Excel to build tracking links.
http://www.screencast.com/t/yc9WXrqH
Here is the excel formula I used:
="http://yourtracker.com/base.php?c=999&key=abcdefghijklmnop&keyword="&F4&"&adg="&A4&"&adt="&B4&"&adb="&C4&"&ada="&D4&"&adi="&E4
For both CPC and CPM, and for both the Male and Female demographic, I used a bid of $0.15.
In the past, I had really good success with CPM in Argentina. My best ads were running at 2.5 cents per click. The image below shows the results from my PREVIOUS campaign in Argentina. ($0.05 per 1,000 impressions, and 2 clicks per 1,000 impressions [0.2% CTR] = $0.025 per click!)
Results continued in next post below . . .
Part 2 - Results
The graph below shows the offer conversion rate. While the NB offer is slightly higher, I would, for the most part call these pretty even.
But since the NB offer paid almost 2x as much as the EWA offers, there was a HUGE difference in the earnings per click, as shown below.
In this case, the NB offer is the clear winner. In fact, during the campaign run, this was clear after each of the offers had about 10 conversions, so it was switched to 100% for the remainder of the campaign.
A $0.05 EPC isn’t much, but if you can get a $0.02 CPC and 1,000 clicks, you can make a quick $300 profit.
So, how’d I do?
Here is the offer %ROI data.
Crash . . . And . . . Burn.
It should be obvious I didn’t get $0.02 clicks. Let’s take a look.
The following table shows the data for the both the CPC and CPM bidding broken out by Male and Female. (There were, of course, differences between images. But nothing exciting enough to post.)
The big surprise for me is found in the CTR column. Notice the HUGE difference in CTR between CPM and CPC.
Now, I expected CTR to be a lot less than the 0.2% I showed at the start of this case study, since I used safe images I knew would get approved.
But I was blown out of the water seeing how BAD the CPM did. It was CPM bidding that got me 4,000 clicks for less than $100 when I ran this traffic before.
In fact, if you do the math (and you should!), if the CTR for the CPM campaign was the same as on the CPC campaign, I would have had a 10% ROI on the Male demo.
The only thing I can conclude is that the CPM impressions were give poor placement leading to the low CTR. And, if you are using CPM, you HAVE to have a high CTR to make it profitable.
Overall, this could be a great campaign for someone on a low budget to run. You get data quick, and the financial risk is low. If you're willing to spend a couple of bucks translating ad copy, there's no reason you can't go for it!
The only thing you need to do is get some higher CTR images than the ones I used (easy!), use CPC, and you'll be making a profit literally on day 1.
Key Learnings –
1) Don’t be afraid to run foreign language copy!
2) Test offers from different networks!
3) Test bidding models!
4) Everything is always changing. Don’t assume. Test.
Awesome case study man!
It would be interesting to see the results on this with a higher CPM bid (or with several different CPM bids). It's possible that your CPM bid was only being served to trash traffic because the international dating offers are being pushed more now.
That conversion rate seems quite low, is it normally so bad for Be2 offers?
@m0thm4n - What's baffling is that I ran almost exactly what I did before, which was only a few weeks ago, same bids, close to the same copy. And I was getting 0.02-0.03 CPM before. On Be2 offers (for me) it's been in the 5-15% CR range. So, it was about what I expected. It varies a lot by country.
awesome info, it definitely varies huge by country. When you do hit close to profit though, or are trickling profit it's time to hit up these advertisers direct or find networks more directed to these latin-american offers so you can take advantage of the better payouts!
Good stuff... I've noticed lately duplicating high CTR ads or even switching currently successful ads to CPM has been resulting in CTRs that just plummet. It's weird as I don't like to come up with theories and such, but seems as if FB is serving lower position/quality inventory when running cpm right now. With how high my bids have needed to be as of late to get a good level of impressions on CPC, it would make sense that cpm is getting lower priority.
I actually ran the exact campaign like you 3 weeks ago. I was getting 0.15 ctr for my image, and I direct linked the be2 offer as well.
The result? I didn't make a profit just like you do due to the very poor conversion rate. I guess it was around 2-3%.
By the way, I ran the campaign in both CPC and CPM. Look at my screenshot, the exact same case result like you. Running in CPC gets a higher ctr than running in CPM.
This is an ad that has been making me good money across few countries, and I set up both in CPC and CPM.
Anyway I stopped it for not bringing me positive ROI. Maybe I should not be lazy and get a landing page set up to presell the visitors... I have to admit that I've been lazy lately. 
Awesome case study! Keep up the good work and keep testing!
By the way inversion, I notice you are using CPVLab to track FB ads, how do you actually upload the correct click cost to CPVLab? I am looking forward to learn it as I just switched over to CPVlab from p202. Hope you can give me some help on this. Thanks in advance!
... time to negotiate a higher payout. This honestly made me sick to my stomach, I'm absolutely embarrassed. It's especially upsetting considering I just wrote that massive guide.
this ruined my night.
bit too much vino last night Mr Eagle..???
Originally Posted by kokofai
If anyone is interested, I can show you how I track my costs. Essentially, my daily process is this:
1) download FB data from last day
2) use Excel 2010 pivot table to summarize clicks and spend per campaign (by campaign I mean at a higher level, not at the FB level, so you need to combine all of the FB campaigns associated with the higher-level campaign; this is super easy and fast if you use "slicers")
3) paste the click and cost data into another excel worksheet that tracks the data for the campaign by day
4) put in the payout data from the network report
5) calculate CPC and EPC. mark the row GREEN if EPC > CPC. mark the row RED otherwise. and record any actions that need to be taken against that campaign.
It sounds complicated, but the whole process takes me less than 10 minutes and gives me much better "control" of what's going on.
Originally Posted by Ryan Eagle
I've had some good runs with other EWA dating offers, but Be2 offers have been soft through your network. Argentina was probably the worst example.
Someone needs to get pecked in the arse for that.
Awesome case study, inversion, much appreciated. I like how you revealed every little detail, even how you got the ads tranlsated!
I'm wondering though, you said you ran the same campaign (same ads, i.e. even the same images?) on FB before and made a profit just a few weeks before this case study. Any theories of why you couldn't get the cheap $.02 clicks anymore?
P.S.: Would you say that you could get around that excel wizardry by simply using P202 instead of CPVlab?
@alex_b - I did the case study fully expecting the CPC to do worse. I really don't have a theory at this point. I'm thinking about doing some additional testing (different accounts, etc.)
I've never used P202. I should probably give it a try.
Originally Posted by inversion
For a single Be2 offer, I would usually split test between EWA and Adsimilis and the end result would always be, Be2 in adsimilis ALWAYS convert better than EWA.
Ryan, I really have no idea why would that be, but being a publisher, a wise choice is always stick to the one that makes me more money. To be honest with you.
* Be2 are not the only offers, I split tested other offers like as well but the end result are the same. I concluded to myself that I will stop running dating with EWA.
Just something to keep in mind, CPM can sometimes bring you worse converting traffic. I've done quite a few tests, and results always vary
awesome casestudy.
re the cpm vs cpc - great findings as well.
its in line with what I read a while back from facebook - CPM is more from brand based campaigns where its more about crating brand recognition and awareness, where as cpc is more for action based campaigns.
as you suggest - CPM based bids are likely to get the secondary placements.
It would be interesting to test CPM bid price and how it changes traffic quality.
CPM vs CPC. Always split test. I run a dating offer. And for targetting men CPM is the best and for women CPC...Next time it can be the other way arround.
EWA vs NB: I split tested the campaign above between EWA and NB. Indeed NB converted better...much better. But worried me was the fact that I had some conversions on EWA...the Pixel fired...but nothing to see in the reports...
I don't want to bitch about a few $, so I didn't report it, also due to the fact NB converted better anyway.
I ran Be2 Mexico and Argentina on EWA... had the exact same results in terms of not being able to make it profitable. It is fucking hard. One thing I've noticed, that goes along with what other people have found, is that it seems the "bid CPM on intl traffic" does NOT work. I've tested it time and time again, and I've run .6% CTR camps in Mexico, and CPM does NOT work. Facebook DOES prioritize traffic to CPC over CPM. The differences in my CTR on the exact same image are like night and day. Right out the gates, I'll get a sub .07 CTR on a CPM bid, and .8% CTR on a CPC bid. Those numbers aren't exaggerated either.
By the way, did I mention that my conversion rate was fucking awesome?
When it is highlighted in yellow like that, it means above the network average, correct?
Given the fluctuations in CPC that you're inevitably going to see through the day, it is REALLY hard to make this work.
Originally Posted by grieve
By the way, did I mention that my conversion rate was fucking awesome?
When it is highlighted in yellow like that, it means above the network average, correct?
Given the fluctuations in CPC that you're inevitably going to see through the day, it is REALLY hard to make this work.
24% is insane... THEN maybe I can see this offer working. What's your issue at the moment with scaling? CTR?
expanding on the demo, more precision targeting and FB disapproving my last couple of hundred ads 
my CTR is solid, I'm trying to play w/bidding to get some more views
My other consideration is I've seen day-parting has impact on conversions but I know if you pause and resume campaign it affects the CPC so right now I'm letting it run
Here's a tip -
Are you doing "Create similar ads" for a submission? If so, stop. Create every single ad manually, from scratch. It is a pain in the ass, but I saw my denials get drastically reduced. Give it a shot 
nope I use spydr tool to upload ads.. but I did notice that when I was using keybroker social I wasn't getting as many disapprovals...
I've being doing some campaigns in latam and I know for sure that cpc is much better than cpm... check
Home > Facebook & Instagram >