- Sep 24, 2008
- Reaction score
Before we begin, I just want to say something for the people that may not know, as I know there are some out there.
Media Buying – Basically means buying traffic on websites and making a profit on the revenue. So it's pretty much paying for advertising. People know this already, just some have not linked it to the term “Media Buying”
People may also not be aware just how valuable paid advertising can be over your typical “free traffic generation methods” which can be slow at first and time consuming. What I love about Media Buying is you can test things quickly with a small budget, find winning campaigns and quickly scale up.
You just can't get that with free traffic gen methods.
And specifically, I'm in the business of Adult Media Buying, so not Facebook, or Google Ads, but buying advertising on all the big porn sites
This case study is based on my follow along journey that I did last year, you can find the day by day break down here: https://www.blackhatworld.com/seo/f...it-days-and-beyond-with-paid-traffic.1367454/
I'm doing this case study of it, since I said I would release it as a Christmas present to everyone, but did not get around to it for Christmas. So here it is now lol.
As I mentioned in the follow along, I was running USA, Mobile Header placement on Pornhub.com
Running JerkMate DOI PPL Offer with around $50/ Day cap to start.
But I didn't reveal the exact banners, the angles and the offer pages used, or the targeting I used to dig more profits out. So this is what I'm revealing in this case study and it's purpose, to try and help newbies out more to get a better understanding of the psychology behind media buying and what works and why etc etc.
The Offer: JerkMate DOI PPL
The Payout: $3.50
The Placement: PornHub Mobile Header USA on TrafficJunky.com
The Spend Cap While Testing: $50/ Day
The Offer Pages:
So, as I mentioned in the full breakdown thread, I'm only testing 5 banners and 2 offer/tour pages... this gives us 10 possible combinations to test VS testing 10 banners and 5 offer pages at once, giving us 50 possible combinations and this is just far too much.
There is something called statistical significant data, and basically it means, you should send enough clicks to EACH POSSIBLE COMBINATION so that 1 or 2 extra leads won't cause a huge jump in your metrics.
So for example, if I send 10 clicks to a banner > offer page combination and get 1 lead, that's a 10% CVR number. If I happen to get another random lead, in those 10 clicks, my CVR for this combination will now jump to 20%
But if I send 100 clicks to the same banner > offer combination and get 1-2 leads, that's a more reasonable 1-2% CVR, and then another lead will only increase the CVR to 3%.... a more realistic number across MORE clicks. Giving us more reliable data to make tweaking decisions later on.
Having a banner combination with only 10 clicks and 1 lead, with a CVR of 10% is too random and insignificant, because over the longer term with more clicks, the CVR will come down and it might turn out that while this combination looked good on day 1 with 10 clicks..... after 3 days running and 100 clicks, it may still only have 1 lead, with a 1% CVR and not be a winner.
So to send ENOUGH clicks to each 50 possible combinations when we have 10 banners and 5 offers is too expensive, unrealistic and not the right approach.
So the pros launch quickly and tightly, with just 5 banners and 2 offer pages, giving us 10 possible combinations to work with, sending enough clicks to each, for significant data and is much cheaper as well. You don't have to spend as much as in the previous example.
Hope that made sense
I ran this campaign for 3 days straight, I recommend when launching a brand new campaign to cap the spend to around $20-$50 / day - in this example I capped it at $50.
You should run a new campaign for 3 days untouched because in this game there are random off days, where stuff can just convert like shit, and if you paused the campaign on day 2, day 3 might do really well and you didn't know because you paused it too early. And this also allows us to get enough significant data across all our 10 combinations so we can make proper tweaking adjustments later based off what the data is telling us.
The Results After First 3 Days:
Hey, So, it's been a while since I've done a good thread/follow along. Haven't actually posted much here since 2013! Guess I just got really busy learning about media buying and then finding success and just kept hammering out campaign after campaign. I'm at the point now where I want to...
As you can see, after 3 days I spent $175.42 for a revenue of $154
So loosing a little bit of money, but as you can see, the losses came from some under performing banners.
So at this stage, I cut banners 03, 05 and 01 and just left running 02 and 04.
I kept running like this for a few more days and increased the spend a little bit, to around $300/ day.
The Results After Day 9:
Nice profits try to scale up more op. Yep, ramping up right now, increased cap and bids sightly :) Will update results tomorrow!
You can see that now banner 04 is coming out as the winner and I also had a look at my OS targeting.
This is the next step in tweaking your campaign. You must dig into your data and find pokets of profits, this is how you can create winning campaigns. You will find a lot of unprofitable traffic that you're paying for at the traffic network and the trick is to dig into your data and filter out the shit traffic as best you can.
We can see that Android traffic is coming out on top, iOS is basically breaking even and we have some highly unprofitable Windows desktop traffic coming in. In fact it is eating into profits quite a lot.
So next the trick was to setup a new campaign that is just targeting Android OS and nothing else, so no iOS or Windows traffic leaks in. And with banner 04 only.
So at this point, let's talk about the creative that's working and why I think its working...
I actually created this banner from scratch. Based on my history with running on PornHub, with it's dark black/orange colour scheme I know that hot pink or bright colours can work well on the page. Because it's completely different and stands out a bit more on the page. That's the first thing.
2nd) I'm calling out to the user directly... kind of calling him out. Guys are on PH ready to jerk off, but in this banner I'm calling him out and saying "hey wait, before you jerk off, try this"
It also triggers curiosity as well, so you get a bit of a boost in CTR because playing on the curiosity factor, guys are more wanting to click on it.
3rd) The "call to action" is a "go private" button... so this indicates Cams, giving them a bit of a hint as to where it's going to take them.
4th) It features a couple of popular pornstars... lesbian angle... "what are they going to do on cam?"
I didn't mention this in the main follow along, but I ended up cutting Tour 2 after a while... you could already see Tour 1 coming out on top on Day 3... with Banner 04.
And you can see why..... I put the same 2 pornstars that are on Tour 1 in banner 04.... so there is concurrency here..... the user clicks on the banner expecting to see these 2 pornstars on cam, and are taken to a page where they can "command" them and tell them what to do.
This brings me to the Tour page itself... this one is one of those "command and obey" interactive pages. Where the user gets to tell them to kiss or do other things, and they are pre-recorded actions which are played on screen... then they are guided to join the cam site if they want to see more.
So it pays to create your own banners, don't get lazy. Yes, sure it's fine to rip existing banners when starting out and test, just to get a benchmark, but as you can see, my brand new created banner came out as the winner.
The Results After 22 Days:
May I ask what are you testing? Type of creative on banners or something else? What prevents you from scaling faster? I’ve dialed this campaign in over the last couple of weeks. Started with 5 banners and 2 different offer pages for JerkMate targeting Pornhub Mobile Header in the USA. Then I...
Total Spend: $37,689.99
Total Revenue: $46,385.50
Total Profit: $8,686.15
I found a winner with banner 04 and tour page 1 so I kept scaling as much as I could and kept running until it started to burn out..... that's what you do in this game, when you find a winner, you must try and scale up hard quick and milk it as long as you can.
During days 9 and 22 I continued to make refinements to the targeting of the campaign. By looking at my tracking data and looking for pockets or profits and also traffic that was running negative. This is why its VITAL to have a tracking solution like Voluum for example, so you can track everything.
I made adjustments to the browser targeting options and keywords/category targeting and things like that. Which helped the campaign to stay in profit for as long as it could.
I hope this helps some newbie media buyers out, and gave you a better understanding of how to launch and run profitable campaigns, what kind of psychology goes on behind creatives and things like that.
I'm happy to answer any questions you might have. Just ask away...