New Facebook camapaign walk through (11)
Well it's been a few years since I actually setup a Facebook campaign myself from scratch. But I own a new company called Seshday.com that is a private sales site for action sports gear. And we're testing some signup funnels with Facebook advertising. Things moving to slow for my taste as always, I decided to setup a campaign myself to test.
And since it's fresh in my mind I thought I'd walk through my process. ( please keep in mind this isn't for you come rip my campaign off, but to see how I go through the process of building any new campaign out. I own Seshday so I have the upper hand anyways :P )
First thing I had to work out was what demographic we wanted to hit. Because, action sports fans are 80% male and younger based on common knowledge. I knew I needed to appeal to a 20 something male demographic.
Now I needed to go gather pictures of things that speak to our demographic, and illustrate what we do as a company.
Because it's an action sports site I laid out a list of what the top selling items that largest spectrum of people like:
watches
shoes
headphone
shorts
shirts
skateboards
surfboards
Then of course things all young males like:
hot chicks
So with these ideas in mind I first went and grabbed around 50-75 images relating to both:
hot chicks in sunglasses
hot chicks surfing
hot chicks with everything else.
The hot chick images from STM that Riley put up
+ lots of straight product shots for the different items
I was on a Mac at the time so I opened each image in iphoto and cropped each until it looks about the size of a Facebook ad. Then hit "enhance" which makes the images pop a little more through brightness and contrast.
Now I had all my images to setup my ads with. I then came up with a single headline and body to start with. My belief is you attract clicks by the images and then actually get the leads because the image has something to do with the offer + the body and headline really sell the action.
I went with a simple headline:
78% off, surf, skate, snow
Then the body was:
78% off on shoes,
skateboards, hats,
etc, etc
Sign up now
Then I was off to the races, last night ( Saturday ) I setup around 80-100 ads. I direct linked using the a4dtracker.com link which I'd setup the campaign in A4D. I'm a simple lazy fellow so I used no tracking and the subid's in A4D to track what worked.
I setup the subid's like this:
c1 = traffic source
c2 = image
c3 = for future copy changes to headlines and body I use a 1-1 type system for what headline and body I use.
Then I can go login to A4D and find out what ads are actually driving leads and not just clicks.
Low and behold all my ads got approved at 2am in the morning. I had set my campaign budget to 50$ a day just to get some initial test data.
Some oversight of my outcome: What I'm doing is I just want to see what direction I want to head from this test. Once I have some preliminary data I'm going to expand out the campaign in that direction that looks like it has potential by testing more images, headlines and body's relating the the 1-2 winning ads.
Now before I tell you how each ad did based on the images mentioned above what did you think did best?
Stop and really make a decision so you can be wrong or right 
Wait for it....
Not a single one of the hot chick ads got a click. Not even 1.
The shoe ads were the hands down winners. On a 4 million set of people we were running around .175 ctr with shoe ads to males 21-29. Who would have guessed.
And on top of that I had a bunch of bright colored shoes I thought would draw attention to the image. Some nice fluorescent ones as well. Nope none of those did that well. The white shoe was the winner. Again I'm baffled, that just doesn't follow any of the things I thought were going to work.
Also a few of the other random ads got similar ctr's and clicks but their conversion rate to leads was around 3% where the shoe ads were around 40-50%.
From here I've gone to test 100 more shoe images. Close-ups, far away, two shoes together, tons of styles, colors etc. Those ads haven't been approved yet.
My point of this story is just start with everything you can think of targeting a large demo and seeing if you can even get clicks or a good CTR. I've been in this business for around 6-7 years now and you think I'd be able to guess a winner a lot of the time. Well I'm wrong much more then I'm right. And it all comes down to testing, pruning, optimizing and more testing.
I've spent around 100-200$ at this point to get quite a bit of very good data. I know a bunch of people ask "what should I spend on a campaign before I bail on it". In my opinion there's no fixed % or amount you should spend. Effectively these leads are worth 1-2$ each. I've spent 100 x's what a lead is worth to get the data I have. But that's how I go about things. If I had only spent a little money on a few ads I probably wouldn't know what I know now. I've got my click cost down to 26 cents on my shoe ads and leads coming in around 60-90 cents. Only way you can effectively get where you want to go time and time again is this same strategy. You might get very lucky from time to time and find a winner by chance right from the start, but 95% of the time that's not going to happen.
Hope this helps some of you with psychology.
P.S. I didn't give my images for a reason, this is a live campaign I'm running.
P.P.S I think I can get my lead cost down to 30 cents or less based on optimizing the landing page, ads, demographics. This business is all about data collection and learning.
P.P.P.S. My best CPA acquisition cost was a few years ago on a media buy. We started at a $250 cpa cost for an offer that paid $42. 2 weeks later we were down to a $20 cpa cost and running 1200 sales a day making $22 profit per sale. Now if I would have quit after I spent 10x's what the cpa was I never would have made this money. If you know something is going to work build it and iterate. That's how you make real money in this business.
Sweet write up Jason. How did you like building a campaign after so much time has passed?
What makes great marketers is there ability to be patient and test tons of variations. I sometimes get tired of testing a campaign and dump it too quick, and I shouldn't have. There was a previous thread with how a company tested 27k variations.
I guessed a picture with a chick would do the best, but after I thought about it for a while, the demographics for your website are adventurous guys who are more interested in gear, especially shoes these days. Everyone always has to have a fresh pair of kicks.
I personally test 10-20x the payout to gather data. Data is King.
How much was spent on the offer at $250 cpa before profitable? And why did you start the cpa at $250?
awesome post.....
I knew shoes worked for women....but never tested for men.....learn something new everyday...
I of course love running campaigns. I don't do it because my time is more value placed other places. But I really enjoy it when I do.
We spent around $15k testing the 250$ CPA campaign. Well that's just where it started with nothing optimized. That's why when someone says I put up a campaign and am losing a little money or break even. You're pretty golden if you ask me. That means there's just a few things that need to be tweaked to become very profitable. Maybe it's just make the eyes bluer in the image. You're always usually 1 or 2 steps away from doubling your money at that point. Only if we knew what those 1-2 steps were. 
Just have to test and eventually you'll find them.
P.S. Also note from a logistics standpoint I make each image I make a new consecutive number. Then when I make the ad I hit "review ad" and I make my ad names in FB like this 1-4 for the image number and the headline number. That way it makes it easy to find which one is working when I have 100's.
Awesome post!
Originally Posted by digitalninja
Your site is awesome Smaxor. Could you elaborate on how the site works? Where do you find your inventory and are you handling all the shipping etc? I've been looking to get into something more long term and dare I say "legitimate".
We have buyers that deal with manufacturers direct. We get get closeout product lists from them then buy bulk. It's a 4 person operation right now not including me doing marketing.
1 buyer, 1 guy that runs everything and manages the tech stuff, 2 guys that double as cs/fulfillment.
Funny, I don't know if this is your campaign or not but I'm seeing ALOT of shoe creatives right now in my right bar... so either it's you or someone has already robbed ya 
Is this site open as an offer to publishers?
Thanks for the post Smaxor, especially the comments on optimizing campaigns as I believe I often give up to easily. LOL at the offer requests.
Home > Paid Traffic Sources > Facebook & Instagram