So you want to get rolling with FB, but but you’re overwhelmed by the number of offers… so what do you do? I’m going to type as fast as my mind thinks about how I would approach this problem.
Step1: Read the FB Terms (google it)
Step2: Decide what kind of verticals would be OK to run according to the Terms.
Step3: Pick a vertical… for this example say I pick education.
Step4: Check to see what networks are strong in education. Ask your AM what are good education offers
This is important many AMs love to dump new offers on you. Tell them to go fuck themselves and ask them about the offers (advertisers) that have been around for a while and have consistent EPC’s month to month. Why should you waste your hard earned money testing new offers that get capped, get pulled, get scrubbed, or flat out dont convert. Nothing makes me more angry then getting profitable and then finding out I can’t run the offer because the advertiser doesnt have their shit together.
For this example I’m going to go with one of my favorites Scholarships 4 Moms Why is this offer one of my faves? Because it meets 2 criteria to be successful on social. It has an angle, and it has a "free" element to it.
Let me explain the difference between social and say search. With search someone is actively looking for a solution to a problem. They are looking for your ads, you just have to tell them why you’re better than the guys on the page with you.
Social on the other hand is "interruptive" marketing. People on Facebook don’t give a shit about your offer. They are there to stalk ex’s and read "look how awesome I am" status updates. Your ad has to interrupt them and say "HEY BITCH LOOK AT ME IM FUCKING AWESOME… learn more"
Step5: Back to the research and planning. One of my favorites is Google Insight (google it). If I check "scholarships" in the US I can see that the niche is cyclical. Since July is the troth (<– is that how you spell it?) I know now is a good time to get in and catch it on the upswing. Next I need to know my demo’s for which I use 3 tools since each one has its own bias.
Google Ad Planner
I put in the offer url and to no surprise it skews female with kids. The age thing is all over the place so I cant trust that but a few other interesting skews. Browsing location home. Poorer. Some College. African American. Georgia and south/midwest.
Step6: Think about how you can target using FB to hit your demo. Age obviously, gender of course, single moms = single. What else might they have in their profiles? "my daughter/son" to hit up Afr. Amer. "Tyler Perry Movies". We can also use some census data and see which states/cities have the highest proportion of Afr. Americans and/or Single Moms. Another thing thats overlooked is their employers. Check largest employers in the US (google it)… maybe a little "McDonalds" or "Walmart" action. Anyway, start writing your notes down and check in FB how many people in those buckets.
Step7: Plan your attack. Since we’re doing social we have to think about it from a mindset of the Facebook user. They are probably sitting at home with their baby dreaming about what life could have been if they just used BC. Most likely they are single and trolling FB for guys to pull into their web of deceit. We need to stop them in their tracks with a eye catching ad.
The image is the most important part of your ad. Theres a great guide on this forum on how to find images. The key is they have to stick out to the user you’re targeting, but also somehow tie in to your offer. In this case its easy… BABIES, ugly babies, crying babies, babies with dads, babies with moms, babies with money, etc
Get yourself a nice set of images. Use photoshop and get your dimensions straight 110×80 and make sure they look CRISP. Don’t even worry about split testing the headline yet as the images are sooo crucial and we want to minimize cost and variables to test. For the first headline Keep It Simple Stupid
"Are You a Mom?" or "Mom’s Only" "Mom Get $10k" or "Mommies Wanted" <– all your trying to do is call out your audience. The body copy is where you can write what the offer is about to appease the FB reviewers, but most people who click don’t read that ish anyway. "Food stamps? Who needs them mom’s make BANK after getting their degree. Here’s $10k to help" (sarcasm make it spicy but also appease FB ad approval)
Step8: Testing. Some may disagree with me on this, but I think because of ad decay its better to go small 10k to 50k users and then scale out. So in this case I would micro segment my list to say: Fem Single "Playing With My Daughter" 20 – 35 Georgia
Load up your images and let ‘er rip. In this case I know the offer gets between .30 to .50 epc because thats what my AM told me so I’ll prob bid in that range to get some data. Great you got your top 5 images.
Now take your top 2 image and combine it with a dozen or so headlines/copy to test. Since your micro segment has already seen your images you wont get fair data so its time to make a new micro segment. I’d prob do the same targeting except this time use Alabama instead of Georgia.
Run it again and now you have your top Headline + Image combo.
NOTE: keep all other variables same… time of day, day of week, offer, etc
Step9. So now you know "about" what kind of CTR’s you can expect. What kind of EPCs you can expect. You can now scale. Take the same targeting and slowly scale your it. Add more states, add more related terms, etc. If your performance is way off the bench line then you’re probably too broad and need to zero in on your core group more. Ideally you would have a few campaigns running where your only job is to continue to "refresh" the ad images due to ad decay. If the campaign fizzles out put it on pause. Come back 2 weeks later it will profit again for a little while due to the sheer number of new users.
Step10. Now you know your offer and know whats possible. You can now approach other traffic source because you KNOW YOUR METRICS. For example you can go do a media buy by finding sites that have the demo you found to be most profitable. Since you know your EPC you can figure out what kind of CTR and at what CPM you would need to hit exact ROI goals.
PREAMPTIVE Q AND A:
I know people will bring up a few questions so I’ll answer them now.
Doesn’t performance vary by demo etc? Yes, but in general offers tend to fall within a "range" of EPC and as long as you’re sending quality traffic it will usually fall within that range.
Should you use a lander? that is up to you. Ask your AM and maybe once you’re rolling in the green you can add a quick split test and decide for yourself.
Have I ran this campaign? Yes, in fact there was a PDF the advertiser sent about what isn’t allowed which featured most of my ads. Moral… don’t use celebrities.
Why did you write such an extensive post? Because I want more thanks than Mr. Green.
If you like this post and want more then give thanks and lets send Mr. Green back to the 5th spot where he belongs.
Great Post and almost makes me wanna try FB Ads
You are one awesome angry russian. Let me give you your first thanks and I’m sure there are many more to come. I’m brainstorming a list of offers to promote on facebook and this came just right in time.
Superb! I love the smell of process in the morning.
Great post dude! Chicks do love the babies.
I have a question for you that you didn’t seem to answer, and for that reason I refuse to thank your post (well until you answer sufficiently). You mentioned you targeted places like Alabama & Georgia, did you mention that the targets in your ad copy or images? If so, how? If not, why not?
You have (1) pending thanks.
P.S. I am still kicking ass.
You can go that route, but for the mico segments I just really care what images perform. Plus FB doesn’t like when you call out users by state and age since that was beaten to death by the auto insurance crowd back in the day.
And yeah yeah maybe I need to get a cartoon avatar too?
Fair point. It probably wouldn’t increase the CTRs by too much but I’ve always wanted to mess around using the targeting in the images. E.g using state icons, flags, monuments, popular joints etc.
Not possible dude, only happy people are allowed cartoon avatars.
Mr.Green vs The Angry Russia is well worth the STM subscription!
Fantastic post. The STM forum is getting better and better!!
Amazing post. Much respect man!
I’m sure much of this could be used for POF as well?
Thanks a lot
Great post man and this should get me move my lazzy ass and research a vertical/offer/ and get the ads up and running and maybe make some money.
Russian be killin it. Mr green def needs to step his shit up…
Hmmm… ads disapproved:
"The content of this ad does not accurately reflect the product or service advertised."
Any advice on how to get photos showing babies past the nasty reviewers? Any tips in general for this problem would be appreciated, I’m finding less and less of my "stand out" angles are getting through to be approved.
You need to give us examples and more info. Also for the Scholarship offer if theres an exit pop I know FB doesn’t like that. Easy solution… bait and switch OR just use your own lander.
Well, I’ve created a campaign with ugly babies :-). Ad text is relevant to the scholarship offer and there’s no exit pop up so I’m assuming it’s the images. I guess this is more a general question to get your thoughts on how to get photos with the "WTF factor" approved if they may not be 100% related to the offer.
Babies are 100% related to the offer. You’re targetting moms so babies make sense. Maybe find pictures of ugly moms with babies. Point is sometimes ad reviewers are cunts and its not a "fair" system. The same ad may get denied 10 times ina row and on the 11th get approved.
1 thing to note is FB doesn’t want "offensive" ads so make sure your babies are ugly but still cute not straight up deformed or scary.
There is a thread on STM on how to find good pictures. Read through that and find more pics.
One last tip. When you submit 30 ads, make the first 15 SUPER compliant and then sneak 1 or 2 that are pushing it through, then finish up with more SUPER compliant ads.
Thank you, good advice! +1 thanks coming your way (is Mr Green still shooting flying sheep??)
Any tips on resubmitting ads/campaigns? Do you just submit again everything exactly the same or do you change ad names etc?
I’d change a small element to like in the body or the title just in case they decide to review your whole campaign.
This post should be a sticky!
I’m only just trying to get to grips with FB at the moment, so it’s extremely useful to have a process. Rinse and repeat right.
The point you made about tested your campaign on just a segment of your demo, by targeting individual areas/cities so that your ads don’t get played out too quickly is a great one, and something I tried to do on a campaign I launched yesterday (I’m gonna start a separate thread with the info, so I can get some help). Problem was, even though I was bidding at the minimum suggested bid (even raising it later), I just didn’t get much traffic. It was UK traffic, and I restricted viewing to London only (+40km) which gave me a possible reach of 25k. I left it running for a few hours but only hit around 1.3k users and got 13 clicks!
As I said, I’ll stick the stats into a separate thread so I can get some feedback, but I’d be interested on what kind of traffic to expect when you have a reach of only 10-50k users as mentioned in your post.
Great most by Angry Russian (: Thanks from me is on the way… Wish you good luck for beating the Mr. Green in thanks war
Fantastic post! This is what makes private forum likes this much valuable than anything else, including the monthly subscription fee!
Thanks a lot Angry Russian!
I have just started to work on Facebook and I have a few questions hope that you can advice:
To be effectively and statistically test all the images equivalently, assuming I have a demo of 50k and less, is there any number of impressions vs clicks that we need to wait for until we can pick the top 5 images and work on it?
I have launched a campaign with 11 images and after running for 2days+, the highest impression an image get is about 70,000+ and another one is 1,800+ with CTR ranging from 0% to 0.085%. At this point yes there are some leads generated so I am not having 100% losses yet but due to the still high CPC ($0.22 per click) that’s why I am in the red.
I have now stopped all the images that get less than 0.08% CTR (thanks to EWA guide) and created 4 variations with border, odd shape, etc and continue for 2nd round testing.
My question is… Should I do it earlier, i.e. as long as some ads received like 8,000 impressions or not?
This seems like the logical way to work on next but I have launched more variations of images instead and continue to improve the CTR. I am hoping by aiming for higher CTR I could get a lower CPC. I will let this runs for 1-2 days and see if this is okie before stopping it though.
I have a question here… my demo is about 50k and now after 2 days+ running the campaign, it has reached about 25k of demo, still 50% yet to expose to my ads. Is this ‘non-reachable-yet’ audience size is good enough for another round of testing? I afraid this left behind FB users are those not so active users or even fake profiles, is it possible? If so, 50% non-active members & fake profiles is it too much?
Hey Guys, any one up for some advices here?
Thanks in advance!