Bust Out the Baby Oil, Another Adult Follow Along! (31)
Hi all,
I've decided to try my hand in the adult dating space, specifically mobile. I have a little ad arbitrage experience and some limited affiliate experience. Overall, I'd say I'm an advanced amateur. What I lack in Affiliate Experience I make up for in programming skills, tracking, and understanding analytics.
The Setup
- STM mobile tracker setup on Amazon Web Services
- Traffic Junky
- F5 (and hopefully a few others as I go)
- CloudFront CDN for all static assets (I'm trying to remove speed & technology issues from the start)
My Initial Goal
- Break even consistently, while learning how to provide quality leads
- Gather as much data as possible, and learn the traffic in/outs
- Make a few friends
- Rotate offers from other networks
- Get a comprehensive idea of what currently works banner/lander wise.
My Longterm Goal
- $100 day, from there scaling out
- Move to multiple traffic sources
- Create, test, and push new creatives/angles.
- Keep my fapping under 3 times a day
Initially I've used the spy tool to rip banners and landers, though I don't plan on sticking with that strategy for too long (refer to initial goal of making a few friends).
Target Country: Germany (I'll be serving all landers through a .de domain)
Budget: 10k initially
My initial plan is to start with 6 banners, 2 landers, and an offer recommended by my AM. I'll be redirecting prepaid carriers and, at behest of other posts, under 25 traffic. I'm following most of the advice on the forum for tracking and testing. I'll use statistical significance as an indicator of what is/isn't working.
This is where I have a question! I'm initially starting my campaigns in Germany. The only draw back is that I can't use WHOIS on a .de domain. Does anyone have a suggestion, other than getting a prepaid card/burner phone/email and signing up with some false info. I'd rather my personal info not be floating around on the internets!
If you see anything that looks idiotic, feel free to let me know. Otherwise, I'll post my daily stats and hopefully make a few friends in the process.
Thanks!
Other then that man looking forward to this start up!
Looks good - interested to see how you get on! Would agree with Maynzie on testing multiple offers and a few more banners.
Domains: I'd use a generic .com or .info domain. As you say, getting the .de is not worth the pain of not having a whois redirect.
Domain names don't make that much difference, and particularly internationally people are very used to .com as a domain.
Glad to hear the top-level domain isn't a deal breaker. I'm gonna take the advice and add more banners and offers to my setup. I appreciate the feedback!
Originally Posted by thebrent
great setup! subscribed!
The setup looks good so far, keep us updated!
Just a few updates!
I decided to run desktop traffic, because it seemed to present better options for offers. One is a SOI and the other is a DOI, do you guys recommend running the same landers for the offers? Or, would it make more sense to stick to a single type of offer?
I was going to rotate 2 landers, but I'm not sure if this is a good idea seeing as how the offers are somewhat different. Any recommendations?
Depending on your budget i'd rotate 2 landers with both offers.
Lander 1 w/ offer 1 & 2
Lander 2 w/ offer 1 & 2
The more you split test the more you'll find a more profitable campaign
Better late than never! Sorry about the delay, but at long last I am up and running. I will post Day One stats after my first 24 hours running.
Day One:
9-18-2013
Spent: $0.75
Revenue: $0.00
ROI: -100%
There is nothing too exciting to report. I'm still getting used to the placement/bidding system with Traffic Junky. Getting ads approved takes quite a bit of time, does anyone know if they have a specific schedule for approving ads?
Within the next few days I should be able to ramp up to $100+ a day.
sometimes they will approve ads in the morning when they open up and sometimes throughout the day, but I see ads consistantly being approved between 5:30 and 6:30EST
Day Two:
9-19-2013
Spent: $1.94
Revenue: $0.00
ROI: -100%
Another slow day for spend, but I'm learning the bidding system and that I need to create a TON of ads at once, otherwise I'm going to be waiting awhile to get more approved. I'm confident that over the next week things will be up and chugging nicely!
Day Three:
9-20-2013
Spent: $9.40
Revenue: $0.00
P/L: -$9.40
ROI: -100%
Day Four:
9-21-2013
Spent: $15.18
Revenue: $0.00
P/L: -$15.18
ROI: -100%
I'm still working up to the $100+ a day spending mark. The only issue I'm having is that I don't trust my tracking. I've noticed that my traffic source will show 100+ clicks, but Prosper only shows 20 "Real Clicks", and maybe 40-50 "Show all Clicks". I don't see any errors in my server logs, and when I click on my ads through a proxy they seem to redirect to the correct location. To check, I added google analytics to my landing pages and google is showing roughly the same stats as Prosper.
Is this a common thing with Traffic Junky, or do the clicks shown in their analytics dashboard usually match up with your tracking server?
Also, do you have an impression count that you use to decide if a banner is a "dud"?
What hosting are you using? It's possible you might be getting drop-out from slow loading.
It is probably worth checking your lp loading times with tools.pingdom.com or Uptrends, just to check nothing is screwy.
In my experience, TJ stats always line up pretty accurately with my tracker besides a couple of clicks here and there, so I'd definitely check your server / tracking setup.
As for the impression count question, check out this thread: http://stmforum.com/forum/showthread...ats-Calculator
Day Five:
12-22-2013
Spent: $15.14
Revenue: $11.92
P/L: -3.22
ROI: -21%
Day Six:
12-23-2013
Spent: $14.04
Revenue: $4.48
P/L: -9.56
ROI: -68%
Day Seven:
12-24-2013
Spent: $6.46
Revenue: $2.24
P/L: -4.22
ROI: -65%
Day Eight:
12-25-2013
Spent: $16.95
Revenue: $5.20
P/L: -11.75
ROI: -69%
After using Pingdom for the last few days, it seems that my landers are serving, on average, less than 300 milliseconds with almost no downtime. Unfortunately, I'm still seeing a large discrepancy in clicks, upwards of 60%. I'm going to reinstall ubuntu and launch an Ec2 instance in the EU to see if that clears things up.
I outsourced my tracking server, but I host my landers/assets with Amazon. Seeing as my prosper stats are lining up with my G analytics stats, for the landers, something is getting lost between ad click and serving the lander.
Once I'm comfortable with my setup and things are tracking right, I'm sure this follow along will get "slightly" more exciting.
Hope everyone had a good Holiday!
Thanks.
Hmm - is your tracker also on an EC2 instance or is it hosted elsewhere? Unless you're running high I/O, EC2 ain't great for tracking - that might explain the discrepancy.
Have you tried running a stress-test on your tracking, as opposed to your LPs? Here's a tutorial on doing that - I'd definitely recommend it in this case.
I was away for the Holidays and had little time to debug this and get everything going, so my main goal is to figure out why my tracking blows and move forward!
I serve all of my landing pages through Amazon (Ec2 - medium instance, CloudFront - cdn), but for the tracking server I decided to use Beyond Hosting and Prosper202. I'm still seeing a 40% discrepancy between clicks on TJ and uniques on G-analytics/prosper. I don't * THINK * this is a BH problem, mainly because Google Analytics and Tracking202 show the same numbers.
I've changed the script that rotates my landing pages, so that may do the trick. I'm also going to stress test the tracking server.
Does anyone know if TJ charges per unique click, or for any clicks (ie: repeat clicks)?
Day Nine:
12-26-2013
Spent: $13.77
Revenue: $2.24
P/L: -$11.53
ROI: -83%
Day Ten:
12-27-2013
Spent: $4.20
Revenue: $0.00
P/L: -$4.20
ROI: -100%
Day Eleven:
12-28-2013
Spent: $21.18
Revenue: $17.84
P/L: -$3.34
ROI: -15%
Hmm, OK, this is quite strange.
Have you contacted TJ to ask them about the discrepancy?
My only other suggestion would be to check whether Cloudfront is getting in the way - are your landing pages PHP or HTML? It seems unlikely that CloudFront's causing a problem, but it's worth ruling it out by running a page that's not being served by them. You should be able to run via direct.yourdomain.com to bypass them, I think.
Day 12:
12-29-2013
Spent: $22.50
Revenue: $2.24
P/L: -$20.26
ROI: -90.4%
Day 13:
12-30-2013
Spent: $16.92
Revenue: $2.24
P/L: -$14.68
ROI: -86%
Day 14:
12-31-2013
Spent: $0.00
Revenue: $0.00
P/L: $0.00
ROI: 0%
Day 15:
1-1-2014
Spent: $6.73
Revenue: $0.00
P/L: -$6.73
ROI: -100%
I finally figured out what was causing the user drop off between Traffic Junky and my Landers. I did a bunch of speed tests for my landers, and the first request across multiple different test sites showed the first request took, on average, 5 seconds. This was due to the Prosper JavaScript tracking pixel. Even though it was before the </body> tag there were some rendering issues, so I'm assuming the users dropped off before it was able to track. Either way, I removed the prosper pixel and left Google Analytics and everything started lining up between the TJ stats and Users hitting my landing page.
In light of this, I decided to stop being a lazy fuck, purchase cpvlabs, and install it on a server I can manage. Things should get slightly more interesting from here.
Well done! Sounds like it was a bit of a nightmare to figure this one out.
Watch out - in my experience CPVLab isn't much faster than Prosper. It's all going to be in the server stats and setup. Keep a close eye on your tracking speeds!
You are going to have issues with cpvlabs also, when you want to do certain things and not having the source to figure out why.
Prosper is fast enough, you might want to find out why it was so slow (Hosting or DB)
Day 16:
1-2-2014
Spent: $7.23
Revenue: $0.00
P/L: -$7.23
ROI: -100%
Day 17:
1-3-2014
Spent: $3.78
Revenue: $0.00
P/L: -$3.78
ROI: -100%
Day 18:
1-4-2014
Spent: $16.15
Revenue: $0.00
P/L: -$16.15
ROI: -100%
Day 19:
1-5-2014
Spent: $15.95
Revenue: $0.00
P/L: -$15.95
ROI: -100%
Day 20:
1-6-2014
Spent: $24.53
Revenue: $4.48
P/L: -$20.05
ROI: -81%
Day 21:
1-7-2014
Spent: $16.80
Revenue: $0.00
P/L: -$16.80
ROI: -100%
Day 22:
1-8-2014
Spent: $11.00
Revenue: $2.24
P/L: -$8.76
ROI: -79%
Feeling pretty decent about my CPVLAB setup. After filtering some of the bot IP's the stats are lining up pretty well. I've put together a few excel spreadsheets so I can put my data together in a meaningful way. Now I'm going to start bidding more positions and attempt to up my daily spend.
I definitely understand the importance of introducing new angles, because everything in my GEO is exactly alike.
Question: How many offers is too many, when rotating offers to find the best one?
No such thing as too many offers provided they'll all work with the rest of your funnel - but the more offers you test the more data you'll need to decide which one to run...
Day 23:
1-9-2014
Spent: $9.91
Revenue: $0.00
P/L: -$9.91
ROI: -100%
Day 24:
1-10-2014
Spent: $12.47
Revenue: $0.00
P/L: -$12.47
ROI: -100%
Day 25:
1-11-2014
Spent: $21.77
Revenue: $2.24
P/L: -$19.53
ROI: -89%
Day 26:
1-12-2014
Spent: $14.09
Revenue: $0.00
P/L: -$14.09
ROI: -100%
Day 27:
1-13-2014
Spent: $10.24
Revenue: $2.24
P/L: -$8.00
ROI: -78%
Day 28:
1-14-2014
Spent: $15.44
Revenue: $5.20
P/L: -$10.24
ROI: -66%
Day 29:
1-15-2014
Spent: $31.47
Revenue: $0.00
P/L: -$31.47
ROI: -100$
Day 30:
1-16-2014
Spent: $55.08
Revenue: $11.92
P/L: -$43.16
ROI: -77%
Day 31:
1-17-2014
Spent: $52.39
Revenue: $5.20
P/L: -$47.19
ROI: -90%
Slowly starting to ramp up my spending and everything looks good so far! All the analytics are on point and I'm learning how to bid competitively on the traffic platform I'm using. From here it's a matter of finding a banner with a good ctr and an offer that converts well!
The landers I've made have been performing marginally, but I haven't spent enough to really have any worth while data!
Question, is it possible to negatively bid against yourself? IE, if I have two separate campaigns on Traffic Junky and each campaign is bidding for the same spot on the same site, will this have a negative effect?
Thanks
Originally Posted by thebrent
Day 32:
1-18-2014
Spent: $34.68
Revenue: $6.72
P/L: -$27.96
ROI: -80.62%
Day 33:
1-19-2014
Spent: $60.47
Revenue: $18.39
P/L: -$42.08
ROI: -69.59%
I'm currently ramping my spending up and rotating offers, so everything is going good! So far, it seems that my angles aren't doing too well, but I'm still learning how to put together a conversion flow.
I've been reading as much as I can about writing ad copy, any book recommendations?
Day 34:
1-20-2014
Spent: $56.07
Revenue: $4.57
P/L: -$51.50
ROI: -91.85%
Nothing great to report, just collecting data and working on creatives/ad copy! My main goal is to find the right offer/s and then move forward from there. Thanks for following along!
Home > Paid Traffic Sources > Adult Traffic (NSFW)