Mobile follow along: Julien needs your help (20)
Ok, here's my first follow along.
I've read the guide from The Angry Russian and I wanted to use my experience as a media buyer in hope to find a profitable campaign again on this traffic source I don't know.
It's been 7 months I haven't find something profitable and I am a bit depressed because of that. I'm fucking disappointed, I was thinking I was on my way to success, as I'm going to explain, but everything is going wrong again.
I really need your help fellow STMers 
STEP 1
17 days spent at the hospital for the birth of my twins:
http://stmforum.com/forum/showthread...f-work-200-ROI
For the first time of my life, I have never worked a single hour during those times.
I didn't launch any campaign, I have paused my job as a consultant for a month.
Instead, I've deeply studied the guide from Angry Russian, I've made some summaries, some checklists for the best practices to use and developped a framework to work smarter (I can share this one but it will be long).
I've also studied carefully your follow along threads for the contest, and I've also studied all the slides posted on this thread:
http://stmforum.com/forum/showthread...overing-Mobile
(and it represents a fucking huge amount of slides, but I had my time and I wanted to take the mobile train by being prepared)
STEP 2
From May 4th to May 8th
I finally launched my first campaigns on Jumptap.
Developped some connections with my rep, and all of that.
The guy seemed a bit distant at first but now I think we're good to work closer, I have been persistent.
He's ok to set up specific CPC and everything.
I've picked up the offer my AM at Neverblue have recommended to me (dating _ my goal now is only to work on markets highly scalable and in highly demand).
I trust her and she always brought wise recommendations to me so I have started with just this offer to get my wet feet.
I have to mention that this offer has the best rate in terms of EPC at NB, although I perfectly know what works for other can be a disaster for me, and the reverse is also true.
So, I've literally followed the recommendation from Angry Russian:
- 3 campaigns : Iphone, Featured phone, Android
- Blocked wifi traffic
- Blocked app traffic
- Blocked tablets
- Set up my tracking properly (prosper 202)
- Created a bunch of different ads, not reinvented the wheel, I used what has proven to work.
Did some spying (Facebook Ads, Google Images etc.) to get proven images, proven headlines, adcopy and call to action.
I've started with 6 ads per formats. My goal was to 1) testing mobile and 2) collecting datas
- Kept all categories
I've let run things for 4 days.
Since the very beginning, this campaign showed some promises, I was very excited (I'm not anymore now, as we'll see later) :
May 4:
Spent: $19
Revenue: $3
Profits: $-16
ROI: -84,2%
May 5:
Spent: $18,82
Revenue: $12
Profits: $-6,82
ROI: -56,8%
May 6:
Spent: $37,63
Revenue: $36
Profits: $-1,63
ROI: -4,5%
May 7:
Spent: $74,31
Revenue: $81
Profits: $6,69
ROI: 8,3%
May 8:
Spent: $55,96
Revenue: $54
Profits: $-1,96
ROI: -3,6%
At this point, my conclusion was:
Well, it's my very first campaign on mobile and lucky me: I'm nearly breaking even and I'm using Prosper202 that shows a huge click discrepancy. I've also collected the data I wanted. I trust Neverblue so let's direct link with NB's link directly.
- Now I have data
I've identified 1 device, and 1 carrier that seem to be very profitable.
Moreover, these 2 channels have a very high CTR (5%!)
- I know what is my best ad
- I know the offer converts
- I know there is click discrepancy if I use my own tracking solution
So my assumption was:
==> The best is about to come, I just have to :
- Pause the broad testing campaigns
- Launch campaigns on the device and on the carrier I've identified
- Bidding CPM thanks to my high CTR (which was a good idea since I paid $0,01 my clicks, but now I'm not sure CPM is still good for this campaign??)
- Keep the best creative
I don't want to repeat the mistakes I did before: so if it can work, I'll scale it as hard as I can, as soon as possible.
STEP 3
BOOM it worked like a charm!
I was so happy to have success again that I was about to cry!
May 9:
Spent: $126,73
Revenue: $357
Profits: $230,27
ROI: 64,5%
Very good day!
As I said, I don't want to make the same mistakes I did in the past: just being happy with that and letting things running.
So my thinking and my framework was:
- I've found 2 profitable channels (1 device and 1 carrier)
It represents 3 campaigns.
- I'll now use the pausing/unpausing ads trick and keep uploading fresh ads several times a week
- I'll optimize dayparting
- I'll scale in other countries
- I'll duplicate my winning campaings on new ad networks.
- I'll launch new very broad campaings again in order to test a lot of creatives on a high volume, even if it's not profitable. Just to identify the best headline and the best call to action to use. Then I'll take the winning creatives and upload them in the winning campaigns. With the best headline and best CTA, I'll just change later things like: pictures of the girl, borders, colors, fonts…
But… it's time to get back to reality.
Before I had the time to work on this plan, here's what happened in the very 3 campaigns that brought me a 64,5% ROI in just 3 days:
As you can see, CTR is diving!
And I didn't change anything on this 3 campaigns.
According to you, what happened?
Do you think the people I'm targeting are already exhausted by my ads?
Is banner blindness that fast on mobile that a CPM campaign don't last more than 1 day???
Do you think it can have so much differences from one day to one other? I mean, I can understand there is huge difference between Monday and Saturday, but not between Wednesday, Thursday and Friday…
So I'm done for now, I let you give me some advice or suggestion if you have some ideas it could be very nice and very appreciated to share them with me fellas
It depends on your targeting, when its that tight yes its possible that banner blindness sets in.
Here is some educated guesses
US pop... 300mill
People with web enabled phones 200mill
Men in above 100 mill
Men in above looking to date over 18 maybe 20mill
Met above looking to date who use the device you target and the carrier you selected and at the same time visiting site/app within Jumptaps network... you get the picture
When carving out buckets try to carve them out so you're still targeting a lot of people.
The methodology I listed was not to help you scale but to help you make profit on mobile and believe that it does work. The key now is to find bigger pockets of traffic. Instead of digging into phones/carrier right away track JT_SITE where I found a lot of performance issues arise.
Best of luck.
Thank you for your answer, it really makes sense.
I have always tracked publishers in media buying but with mobile, I think I've been distracted by the technology.
I focused on the devices and carriers, which does not make a lot of sense unless I want to identify platforms with problems, I'll be smarter on the next try
So my next move:
- Launching a new broad test ($100-$200), with prosper 202 for a better tracking, even if I'll lose clicks during this test.
- Tracking JT_SITE to find out some good sources.
By the way, I'm not sure to understand well the difference between JT_SITE and JT_PUBLISHER.
Can anyone explain that to me?
I'll keep this thread updated
Great start.
Have you tried dropping p202 and just using the direct Neverblue link? This should lessen the discrepancies if you're still seeing them.
A single publisher can have many sites, not all of which will be profitable for you, so best to track site instead, or both.
Hi nick, thank you, ok it's perfectly clear now.
So I will track:
c1=JT_CAMPAIGN
c2=JT_HANDSET
c3=JT_OPERATOR
C4=JT_SITE
t202kw=JT_ADBUNDLE
I found out that JT_KEYWORD is useless for me because I'm targeting all categories for now.
Maybe I'll use it later.
Yes I used the direct NB link. Their reporting is great and I trust them, so every campaign with specific settings will be launched with their direct link 
Also, I found out that there's no reason to hide anything, even with p202, they have all the data (carriers, devices, OS...).
My framework is the following (maybe it's not the best thing to do, I don't know) :
1/ test with p202 in order to track easily with c1, c2, c3 and c4
2/ when I identify a good combination with the C parameters, duplicate the winning campaign with a NB direct link
if you just use commas when you make your variable for neverblue its not that hard to separate out...just add some column headers to the csv and it will be very easy to attack with pivot tables. easier than duplicating the campaign imo and now youre not pissing your clicks away (prosper)
Hey codeface, I see... thank you I'll try this.
This way?
-http://fgkmv.com/click/?s=XXXXXX&c=XXXXXX&subid=JT_CAMPAIGN,JT_HANDSET,JT _OPERATOR,JT_SITE,JT_ADBUNDLE
Isn't it a bit long? Do you think everything will pass correctly on Neverblue's side?
EDIT:
I'm going to watch at this thread too:
http://stmforum.com/forum/showthread...ing-on-the-Fly!
Yup. That works fine on Neverblue's end.
Here's some news:
My ads were finally approved, it was a bit long.
Now I'm very comfortable with tracking, direct linking and pivot tables.
I'm focused on sites instead of brand/model/carriers now.
You were totally right guys, there are huge differences between sites.
The best combination between my banners and the sites I'm displayed in, gives me a 2.50% CR.
My best EPC is $0.08.
This conversion rate sounds a bit low to me. I don't know what to expect.
If you were me, what would you do? Try other ads in order to improve this CR? Move to another offer?
Move offers or bid down
Ok so here's my first experience with this offer:
Spent: $848,86
Revenue: $822
Profits: $-26,86
ROI: -3,2%
Total Clicks: 11 787 (according to NB)
Total Conversions: 274
CR: 2,34%
For my next try, I think I'll use a rotator and I'll rotate more dating offers to find one that converts better, and then I'll use the direct link again and I will whitelist or blacklist sites with my rep.
My second experience was with Cellflirt, which only accepts ATT traffic.
3 days running with Jumptap, on Feature phone :
- 570 impressions
- 0 clicks
= FAIL!
ATT traffic is inexistent with feature phone
My next experience will be with MobAff offers.
Hmm, it's seems weird to me as well.
Let's me reformulate:
"I was unable to have more than 570 impressions and 0 click on Jumptap for 3 days, with targeting feature phones and ATT as carrier, with a $100 daily budget and a $0.106 CPC."
that is some of the most competitive traffic on JT. You need to bid much higher
Ok, thanks man I'll try this.
After a chat with Marty from NB, I think he's absolutely right: I should be more persistent with my initial campaign.
I should not give up yet, and spend more time testing new creatives and traffic sources with Jumptap.
So my plan is to make way more creatives and launch:
- one broad campaign on Feature phones
- one broad campaign on Android phones
- one targeted campaign on the carrier I had success with
- one targeted campaign on the device I had success with
I'll give up only after testing 50 creatives.
If I can't have success with an evergreen niche like dating, with an offer that has proven to convert, and CPC as low as $0.10, I'll have the same problem no matter what traffic source or offer I choose.
I can understand that mobile is a different traffic source than the ones I'm used to, but if I want to make my road to the full-time heaven, I'll have to focus and work with confidence (framework + persistence).
So, let's focus and make this thing work!
Ok so I found out how to find profitable sites.
If the reading of my data is correct, I found out 2 sites quite interesting.
I did the math on my pivot table to keep only the combinations with an EPC higher than my CPC.
I'll target those 2 sites in separate campaigns.
1 campaign = 1 site on 1 OS
(for example camp1 = site XXXXXXXXXXXXX on Android, camp2 = site XXXXXXXXXXXXX on Feature phone etc.)
This way, I can easily target carriers and devices.
Then, I can cut the fat accordingly later.
I asked my rep to create these campaigns for me, I hope he will be ok with this
because he already made me understand that he won't do a lot of things for me
until I spend at least $250 a day.
Spending this budget doesn't mind, but I couldn't do it on a traffic source I didn't know at all until now.
Also, I asked him for a custom CPC rate.
So let's see what this strategy will bring.
I realize I made a big mistake to not tracking JT_SITE since the beginning.
I can't track properly 150+ previous leads.
I made these leads for almost "nothing". This represents a $500 mistake
This is a great follow along! Keep up the good work Julien! Keep the ball rolling.
Hey Julien, I know I"m digging up old threads, but just wondering if your still doing Mobile and if anything ever worked out for you.
Home > Mobile > Follow-along Campaigns