Mobile Follow Along - busting out of the daily grind (31)
Hey guys, first post and first follow along. First of all a bit about me.
I'm a really lucky guy from Australia, recently married to my beautiful wife and a little girl on the way in May. I have a stable job in online marketing / traffic acquisition for a large media company earning low $xxx,xxx per year, but absolutely hate the work environment and am desperate to spend my days working on my own success rather than sitting in a cubicle listening to some middle-aged schmuck telling me to "call Google and let them know that if we're not happy we're going to take our business to Yahoo" (like they give a shit).
I started back in 2003 in mobile media (remember logos and ringtones) and have worked on and off in affiliate marketing for the better part of 10 years, but never with any serious intent. My company roll has focussed heavily on generating desktop software and app downloads.
Leaving a job while paying for wife/child/rent tends to add urgency and focus, so here I am ready to work like a motherfucker 
Starting Budget
$5,000
First Goal
Generate NET $300 / day to replace income and allow me to work on this full-time.
Vertical
Mobile downloads of course! Not just because mobile is my passion, but I see huge growth potential.
Traffic Source
I've signed up to a few networks and have studied their available traffic and am leaning towards starting with Decisive in some less-competitive markets. My wife is Colombian, so it seems logical that I should start in Latin America with help from her and her friends for translations and local references.
Offer
I am starting with a generic android install available on several networks that will let me experiment with angles.
Tool kit
- VPS at Hostgator (would appreciate advice on whether this will be sufficient?)
- STM Mobile Tracker
- Trafficlab subscription for designing banners
- Sublime Text and template for landers (Thanks Caurmen)
- Gimp for other stuff
Really excited to finally get started, watch this space
Welcome to the forums carboot and big ups on your first post being a follow along!
My one question would be your hosting - is it managed or unmanaged? I don't think HostGator VPSs are particularly well regarded, and you can generally get unmanaged VPS a lot cheaper cf. the price/performance of those HostGator ones.
Hey Zeno....I figured it would make me accountable if I put something up straight away
It's a managed VPS which I've had for about 6 months and will be using for tracking + landing pages.
I'm more concerned about speed...are there any tests or minimum speed benchmarks that you would recommend (or any other hosts)?
You can always check out Serverbear but this is more for technical VPS benchmarks, i.e. filesystem tests, IO benchmarks, etc. Real world performance depends a lot on actual configuration and connectivity.
I found one HostGator benchmark and it was abysmal. If you have root access to your server you could run things like UnixBench. For real world testing... well, this is harder to do in fair way without having other servers on hand to test against. You can try something like blitz.io to see how the server handles a lot of concurrent connections.
You should invest in, at the very least, an SSD-based VPS. If you want fully managed then many people on the forum use BeyondHosting (there are discounts here too). If you want unmanaged there are tonnes, I have used RamNode, DigitalOcean and Edis.
subscribed!
I'd recommend DigitalOcean for a VPS - $5 a month will get you a basic box, and if you're familiar with server installation you can get a tracker up and running on there in about 10 minutes. Their SSD performance is pretty darn good, which will be important, and you can't beat the price tag.
Quick update: As suspected, my hostgator VPS is causing dramas (can't install memcache) so I am going to switch today. I am looking for a managed VPS, and after a bit of research Beyond Hosting looks good.
The VPS2048 is the highlighted option....would appreciate any input as to whether that will be good enough? Happy to spend more if I need to!
Thanks guys 
Will be plenty. For tracking purposes that kind of VPS should be able to handle ballpark a million clicks a day, ergo you will be absolutely fine.
Before I begin, thanks to Zeno and Caurmen for their help 
First results are in:
Spent: $37.64
Clicks: 758
CPC: $0.05
LP CT's: 38
LP CTR: 5.01%
Leads: 3
Revenue: $5
Net Profit: -$32.64
ROI: -86.7%
A few notes:
I've created 8 separate angles for this generic Android download offer. I've only uploaded creatives for one angle, so I'm probably jumping the gun by posting results.
- I used a statistical significance calculator to cull failing creatives based on a confidence level of 90%. I've decided not to measure conversions at the moment because achieving a high CTR for this offer is critical.
- Each angle is broken into 4 campaigns - wifi/site, wifi/app, network/site, network/app.
- On my traffic source/country, there is way more wifi traffic, and site CTR's are outperforming app CTR across the board by about 3 to 1.
- Banner CTR is too low, but expected without any optimization (and absence of all angles)
- LP CTR is WAY too low...I have optimized the design for the majority of Android handsets using Caurmen's guide (which is awesome) and Browserstack, the CTA is above the fold and loading seems to be good, so this one is a bit puzzling.
To do:
- Upload all creatives to cover all angles.
- Clone campaigns with direct-linking to test against my shitty landing page results.
Great work so far! I would recommend testing angle first. In my opinion, angles >> banner. You can make dozens of banners easily and come up with new ideas in a heartbeat but 'angles' don't spawn so readily.
If the country has A LOT more WiFi traffic then focus on that - it is where your volume is. If site CTR >> app CTR then also focus on site to begin with, provided this app CTR isn't skewed by the traffic predominantly coming from a few apps.
Definitely split test direct and via your LP. How fast is your load speed? How is the lander hosted now? Can you measure the load speed from the country you are running the campaign in?
I wouldn't panic too much with the lander CTR just yet since there has been no optimisation, but you may want to also create a second lander to test against.
Not a lot to report today due to a 202 setup error not allowing me to track conversions 
I uploaded several banner variations for each of my 8 angles and have already culled a couple that are seeing particularly bad CTR's (confirmed with 90% confidence test).
I can already see a pretty big difference in CTR's between angles. The one which I was certain would perform worst is actually the best....in other words, never assume anything!
More to come tomorrow
Yesterday's results:
Spent: $23.47
Clicks: 360
CPC: $0.07
Leads: 1
Revenue: $1.45
Net Profit: -$22.02
ROI: -93.8%
I am testing angles with simple creatives and without LP to try to narrow-down which to run with. This is causing a disconnect between the banner and the offer page.
Once I settle on an angle, I am going to re-do my LP that was created originally and a couple of variations, and start building out banner variations to drive down CPC.
Offer was paused, so I'm starting over with an app download in one of my LATAM markets using Decisive.
Based on some suggestions, I ran the offer DL'ing with supplied banners first to test for conversions. I broke this into 4 campaigns: wifi/site, wifi/app, network/site, network/app
Overall Results:
Spent: $32.52
Clicks: 901
CPC: 0.04
Leads: 3
Revenue: $1.30
EPC: $0.001
Net Profit: -$31.22
ROI: -96%
Not fantastic, but at least some conversions. Upon closer inspection, I noticed that all 3 conversions came from network/app.
Network/App Results:
Spent: $5.28
Clicks: 64
CPC: $0.08
Leads: 3
Revenue: $1.30
EPC: $0.02
Net Profit: $3.98
ROI: -75%
Some observations:
- Network traffic is more expensive but seems to convert better.
- Network traffic is much lower volume than Wifi on Decisive in LATAM - potential future issue.
- Decisive live stats are good, but the lag is shitting me to tears. Thinking about changing.
Some questions:
- Should I cut the other 3 and run with Network/App already?
Next step:
- Take the default banner design and modify text with different angles.
So here's what has happened in the last month since I posted.
1. Wife gave birth a few weeks earlier than expected...proud father of a baby girl
2. Productivity absolutely destroyed from dealing with daily tasks that come with being a new father.
3. Burned about $1.5k testing campaigns. A few promising results, but no time/energy to invest.
4. Pressure from my day-job building, increasing my motivation to make this work.
Here are some general findings:
- I probably have 2-3 solid productive work hours per day, and need to work as efficiently as possible.
- I'm sticking with mobile, but have frustrations with some of the traffic sources.
- For whatever reason, mobile game downloads seem to stimulate my brain the most, keeping me awake when I would otherwise drift off and get nothing done. I've been DL'ing 100% of the time with games, but want to start testing LP's.
Questions for those familiar with mobile gaming:
1. For banners I'm finding that visuals make the biggest difference as opposed to written angles. Is this just me or generally accepted?
2. For an overall campaign with 0 conversions, I need to make a decision when to cut and move on. Is there a general rule of thumb for this?
3. When I begin testing a landing pages, should I focus first on copy or design?
Back to the grind
Originally Posted by carboot
1. Wife gave birth a few weeks earlier than expected...proud father of a baby girl
2. Productivity absolutely destroyed from dealing with daily tasks that come with being a new father.
3. Burned about $1.5k testing campaigns. A few promising results, but no time/energy to invest.
4. Pressure from my day-job building, increasing my motivation to make this work.
Here are some general findings:
- I probably have 2-3 solid productive work hours per day, and need to work as efficiently as possible.
- I'm sticking with mobile, but have frustrations with some of the traffic sources.
- For whatever reason, mobile game downloads seem to stimulate my brain the most, keeping me awake when I would otherwise drift off and get nothing done. I've been DL'ing 100% of the time with games, but want to start testing LP's.
Questions for those familiar with mobile gaming:
1. For banners I'm finding that visuals make the biggest difference as opposed to written angles. Is this just me or generally accepted?
2. For an overall campaign with 0 conversions, I need to make a decision when to cut and move on. Is there a general rule of thumb for this?
3. When I begin testing a landing pages, should I focus first on copy or design?
Back to the grind
As for your questions:
1. Gamers are visual creatures, so yes the visuals are going to play a major part. Telling gamers how cool a game is doesn't work as well as showing them.
2. Hmm, hard to say but for a $2 game app install I probably wouldn't ditch more than $100 on it... After testing multiple angles/demos/banner of course. Some games need different angles/banners/audiences but if it just performs poorly across the board it's often because of the game or other mitigating factors. If the game absorbs 100's of clicks and goes nowhere it's a good sign to focus on something else.
3. For games? I'd say design. Copy is important in all cases and for angling but gamers are visual creatures as I said... So you want decent ad and lander imagery/design. That being said, ugly often converts so you could try the opposite! But I don't think gamers often need to be comprehensively sold on the idea of a free game, why they should install it, why it is so amazing, etc. They know the drill. They need the desire, not the long winded pitch.
Congratulations on the birth of your daughter!
Both from what I'd do and from the research I've been doing for the Mobile Cookbook (coming soon), I'd recommend cutting a sub-campaign (one of your four app/site/wifi/mobile combos if you do that) at about 10x payout if it hasn't had any conversions. If you're running across app, site, wifi and mobile as a single campaign, I'd let it get to at least $50 spend before making the call.
OK so I'm back in the swing of things and have decided to launch a fresh campaign for this follow along.
This is an Android Antivirus campaign from F5 paying $3.00 (I'm sure many of you know which one I'm talking about).
Big thanks to Maynzie for all of his help
As stated in my last post, my passion is for game app installs, but for this follow-along I wanted to choose an offer with proven success to see how far I can push it and of course to learn as much as possible.
I am testing with 2 landing pages and 22 banners of varying designs. These were setup very quickly and without much thought. Potential for better things across the board.
I will be sticking to Finch's suggestion of testing in the following order:
1. Offer
2. Angle
3. LP Design
4. LP Content
5. Banners
These are the results from 3 days of testing with no cuts or adjustments:
wifi/site Results:
Spent: $15.78
Clicks: 197
CPC: $0.08
Leads: 1
Revenue: $3.00
EPC: $0.015
Net Profit: -$12.78
ROI: -81%
wifi/app Results:
Spent: $53.30
Clicks: 1,690
CPC: $0.03
Leads: 8
Revenue: $24.00
EPC: $0.014
Net Profit: -$29.30
ROI: -55%
network/site Results:
Spent: $13.85
Clicks: 277
CPC: $0.05
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$13.85
ROI: -100%
network/app Results:
Spent: $19.54
Clicks: 737
CPC: $0.03
Leads: 2
Revenue: $6.00
EPC: < $0.01
Net Profit: -$13.54
ROI: -69%
Some observations:
1. Offer seems to be converting well.
2. Day 1 was profitable for wifi/app.
3. wifi/app is providing much more volume on my traffic source and cheapest cpc's.
4. My banners consist of a variety of angles....some winners are starting to emerge.
5. Potential to improve many of my banners using animation an bring cpc down.
To do:
1. Using results from banner CR's and CTR's, choose some winning angles and apply these to both LP's and continue testing.
2. Run some stat. significance tests on banners to see if any are ready to cut.
More soon...
I am interested in knowing which age group performs better for antivirus app download campaigns. Question to mobile guys: Do you see similar ROI graph ( stats above ) for pin submits . Or 1 kind of traffic from wifi/app performs better always ?
I've moved on to testing angles. In order to do this an keep budget under control, I have set these up under wifi/app, which was the best performing combo so far, and with a single design which was also the best so far.
Results so far:
Angle A:
Spent: $4.19
Clicks: 150
CPC: $0.03
Leads: 1
Revenue: $3.00
EPC: $0.02
Net Profit: -$1.19
ROI: -28%
Angle B:
Spent: $7.18
Clicks: 224
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.18
ROI: -100%
Angle C:
Spent: $7.23
Clicks: 246
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.23
ROI: -100%
Angle D:
Spent: $7.14
Clicks: 249
CPC: $0.03
Leads: 1
Revenue: $3.00
EPC: $0.01
Net Profit: -$4.14
ROI: -58%
Angle E:
Spent: $7.24
Clicks: 238
CPC: $0.03
Leads: 1
Revenue: $3.00
EPC: $0.01
Net Profit: -$4.24
ROI: -59%
Angle F:
Spent: $7.26
Clicks: 233
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.26
ROI: -100%
Angle G:
Spent: $7.25
Clicks: 228
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.25
ROI: -100%
Angle H:
Spent: $7.18
Clicks: 270
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.18
ROI: -100%
Angle I:
Spent: $7.20
Clicks: 240
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.20
ROI: -100%
Angle J:
Spent: $7.24
Clicks: 219
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.24
ROI: -100%
Angle K:
Spent: $7.26
Clicks: 226
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.26
ROI: -100%
Angle L:
Spent: $7.26
Clicks: 197
CPC: $0.04
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.26
ROI: -100%
Angle M:
Spent: $7.24
Clicks: 265
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.24
ROI: -100%
Angle N:
Spent: $7.27
Clicks: 239
CPC: $0.03
Leads: 1
Revenue: $3.00
EPC: $0.01
Net Profit: -$4.27
ROI: -59%
Angle O:
Spent: $7.27
Clicks: 221
CPC: $0.03
Leads: 0
Revenue: $0.00
EPC: $0.00
Net Profit: -$7.27
ROI: -100%
Some points and questions:
1. I won't be cutting any angles until I've spent 3x payout without a conversion.
2. 3 out of the 4 angles are along similar lines. The 4th is completely different.
3. Traffic is quite slow, so I'm going to have to test with some different bid points.
Any general suggestions?
Like you said you haven't completed your test yet, keep the traffic coming, hopefully some more data will come through in a positive fashion. Not to much really else to say here man lets fingers crossed some more leads come through mate need to keep the spend up!
Hey mate, are you promoting what your AMs would call a 'hot' offer? Or did you just pick it up yourself?
redrummr: yeah it's a recommended offer.
I've finished that round of testing, and have paused all angles except for 4:
Angle A:
Spent: $6.50
Clicks: 219
CPC: $0.03
Leads: 1
Revenue: $3.00
EPC: $0.014
Net Profit: -$3.50
ROI: -54%
Angle D:
Spent: $9.44
Clicks: 326
CPC: $0.03
Leads: 1
Revenue: $6.00
EPC: $0.009
Net Profit: -$6.44
ROI: -68%
Angle E:
Spent: $9.54
Clicks: 322
CPC: $0.03
Leads: 2
Revenue: $3.00
EPC: $0.019
Net Profit: -$3.54
ROI: -37%
Angle N:
Spent: $9.57
Clicks: 311
CPC: $0.03
Leads: 1
Revenue: $3.00
EPC: $0.01
Net Profit: -$6.57
ROI: -69%
Looking into the data further, I found the following:
LP CTR
A: 16.3%
D: 11.7%
E: 8.2%
N: 7.4%
Offer CR
A: 2.2%
D: 3.0%
E: 8.3%
N: 4.8%
Some thoughts:
1. Angle E had an offer CR that I can work with.
2. Angle E LP had half the CTR of angle A. If I can improve CTR here, I should be much closer to profitability.
To do:
1. LP design optimization. What LP CTR should I aim for?
Onward
LP CTR: I'd recommend reading Finch's post on the subject. http://stmforum.com/forum/showthread...y-Don-t-Matter
Angle E definitely looks like the most promising.
Be aware, though, that you're dealing in really quite small sample sizes here - $9 spend on a $1.50 payout isn't really enough for a thorough test. Let 'em get to $15 spend at absolute minimum before making any more decisions.
- spent a couple of days building some masterpiece landing page designs to test
- switched on traffic early this evening
- went for a walk
- ordered a pizza
- watched a movie
- logged on to check stats.....heaps of clicks and 0 click throughs
- yelled what the fuck and ran around crying
- checked my landing page url: "You have reached a domain that is pending ICANN verification...We estimate it may take 24 to 48 hours for the website to come back online"
better that than shitty landing page stats I guess
Ugh, no fun. Sympathy.
Take a break, do some research or prep some more campaigns, and come back after your domain's finished being screwey
I was going to post an update with stats today, but it's not really worth it seeing as though traffic is so low and only 2 conversions.
So instead I'm going to ask a question about bidding.
Currently I'm doing the following:
1. Manual CPM
2. Choosing a CPM based on gut
3. Realise that traffic is too low
4. Clone campaign with multiple bid levels
Does this seem like a reasonable strategy, or is there a better way that I can go about things? I can't use smart CPM because after trying a million things I can't get conversion tracking working in Decisive.
:/
Use the planner as a guide for a sensible CPM for that country and go from there.
Very inspirational follow along Carboot thnx! I'll be starting with my first mobile campaign soon, this gives me a good idea what to expect in the beginning. I am sure if you keep it up you'll start to see the green soon!
btw, something I haven't seen here, did you try a split test with a direct link?
Thanks Dennisp, I would highly recommend starting a follow along.
I haven't done any split testing with DL yet, but am about to. Will provide updates soon
I am in the same boat as both you and Dennis up there! Very inspirational / informative follow along. I will definitely be posting a follow along with my first ever mobile campaign next week. I am still in the reading / information gathering stage but I'm really eager to get started. Looking forward to the rest! And of course... STACK THAT MONEY.
Home > Mobile > Follow-along Campaigns