You often hear angles,angles,angles and test,test,test right? Experienced affiliates on WarriorForum, StackThatMoney, etc always talk about how important they are. Angles can make or break a campaign, and it is something new affiliates, as well as season ones often forget to test.
Since super affiliates making $1000+ per day always say angles, and testing is the most important thing a media buyer can do. Still, there are many people who just don’t get it… so what do they really mean?
To demonstrate, I did an experiment.
1) I picked a low paying CPI offer (pays 40 cents per install) from my favorite networks ClickDealer & MundoMedia. (TOP TIP: You always want to SPLIT test CPA Networks when runing an offer!!)
2) The CPI offer accepts Android 2.2+, all traffic, and Mexico geo.
3) Next I fired up my spy tool WhatRunsWhere Mobile, and looked into what banners others were running to promote this offer. My narrow down criteria was Sort by Ad Strength, last 14 days.
4) I selected & downloaded the best banner, and made 2 variations of it. (Changed font color, and border color I know works well in LATAM based on past data)
5) Made a new campaign on Voluum – the cloud based tracking solution that we’ve switched to; it is the BEST, most BAD ASS tracker – you don’t need hosting, you don’t need multiple licenses; everything is in one place – in the cloud – and right now its FREE! (so no crazy high monthly fees!)
6) Then I created my control campaign first, using the banner I downloaded from my Spying on WRW, and the 2 variations I made of that banner (total 3 banners). – This was yesterday. The campaign/banners got approved and traffic started flowing yesterday. I begun to optimize when I had a pretty stable CVR going, it was time to see if I can improve this.
7) I brainstormed 10 new angles I can use to promote this app – I came up with angles by searching for what the most common problems of people who’d benefit from using my app I’m promoting might be having.
8) I had the angles translated using a company called Transey, they provide wayyy cheaper translation than 24hour translation all in one place. (I HATE oDESK!)
9) After the translations came in, I sent them over to bannerslanders.com, and ordered 2 banner design, and 10 variations of each (Basically 2 different designs, and 10 variations being the different angles, in plain english – the only thing I changed on my banners was the text, but they looked the same in every 10 campaigns)
11) I logged into my mobile tracker VOLUUM and created 10 new campaigns naming them – Offer-Geo-ANGLE Number (ie KasperskyAntivirus-Mexico-Angle01) – IT IS VERY IMPORTANT TO DO THIS STEP
12) Next I created 10 campaigns on my traffic source, and uploaded 2 banners to each, linking them to the campaign URLs I made in step #11
13) I set a very high CPC bid (25% of offer payout) because I wanted data fast, set my frequency cap at 3/24 per ad (Golden Nugget: anything more will just lower your CR %!) and put set my budget to $20 per. Offer payout is 40 cents, so I like to test each banner with 2-5x offer payout when running CPI!
14) After the campaigns were approved, I waited for the campaigns to run their course and spend my budget per.
15) Now it was time to compare, and find the winners: (CLICK the screenshot bellow to see high resolution!)
-The WRW Banner generated $634.80 revenue and 1587 conversions and performed overall 8.83% CR after running for 24 hours.
-Out of the 10 angles I tested, angle #5 did amazing 13.05% based on 429 clicks! That is HUGE – almost 5% more conversions than the banner I stole from WRW!
-There were a few other great angles as well, above 10% like Angle #6, 10, 2, 13, 7. The others I tested were the same or worse than the WRW one.
As you can see, what the experienced affiliates on forums like STM say is TRUE. It IS all about the angles, and you really have to test, test, test.
If I would’ve just ran the banner I got from spying on WRW I would’ve been floating around break even, but now I am banking hard thanks to this little but extra important test I’ve done. 5% more than the WRW Banner, plus many more other potentials in case my overall winner starts to burn out.
39 thoughts on “The Very Importance of Angles – A Case Study in Mobile Media Buying CPI Offer”
Nice post but I can’t seem to find the translation provider you talked about. Mind posting a link?
Transey is still in beta, it is not yet available publicly. Stayed tuned 🙂
I can only dream to get an app install to +5% CR. Any tips? Is it too depending on the traffic source?
It depends on Angle, Creative, and Placement. If it is a special interest app, ie: PoolPal (swimming pool BBB method calculator) you only want to target people that own swimming pools.. else your conversions will be shit, because only these guys are the ones who need this app. So you’d advertise it on swimming pool forums, in other swimming pool apps, on swimming pool supplies websites, etc.
Hope that helps.
Where did you buy traffic from? Did you place the banners on the same websites where the longest banners were running according to WRR data?
Did you do direct linking?
I have just found your blog and already finished reading some of your great posts. I see you’ve mentioned STM (Stack That Money) in some of your post, could you please shade some lights or a few lines review about STM? Do you recommend STM?
Yes, STM is the ONLY board that has REAL super affiliates on it. No bullshit kool aid, no withholding information required for success because there’s conflict of interest.
are just a few of the well known names who are all members and have been on STM for a long time.
Thanks for the reply. Much appreciated 🙂
You did not mention the landing page. Are you doing direct linking?
It depends on the traffic source and the offer. on PSafe Antivirus for example I use a lander, because the CR goes up with one — without a lander its like 2-3% less.
many thanks for the insights. it seems that most offers need landing page with right angle. I was wondering which kind of offers actually do not need a landers, just direct linking, but we play the angle part at the banner level.
NP, my pleasure to help. Well, to be quite frank – only offers where you can right away tell what the app/game is/does don’t need landers. Otherwise you need a lander to educate and give insights so people actually go and install it. So id’say its a 20/80 rule (20% don’t need a lander and do well on DL, 80% need one)
I see. But as the landscape get more competitive, perhaps we need landers also for the remaining 20% in near future. 🙂
well yea and you need moats
the picture in voluum though you are not calculating the cpc, so it’s hard to tell how really much you made and if you were profitable at all 😉
Andrea, my man. this case studies point is not to show mega, giga profits and dreamy numbers. it was to show how its important to evaluate angles, because different angles convert at different CR % 🙂
Hi Attila. You mentioned that you were bidding at 25% of offer payout which is $0.1. From the screenshot you showed us, even the angle with highest CR you are still at roughly -47% ROI am I right? (429 x 0.1 = $42.9 and your revenue was $22.4). Can I know how you made it into positive ROI?
I always bid high, because I want data FAST. 99% of the time you will be negative, step 1 of mine is always to evaluate out of all the angles I came up with which is the strongest (highest converting CR %) and then I know what i can afford to pay.
Next step is to move down then my max bid to a little above what my CR for best angle can do, and then optimize down.
Did you run the campaigns without any targeting? didnt you seperate carrier/wifi and app/site at all?
Or did you just cut bad placements and let it run all together? I am impressed that you had everywhere a decent breakeven conversionrate at least.
No need to do that in this stage, because the point was to find the strongest angle.
You don’t have a fuckload of time, most offers out there suck. You gotta work efficient and do stuff that has big impact, can’t make a gazillion campaigns and spend a week on it only to find out offer sucks.
Chop chop 🙂
PS: Having a team helps big, I don’t do grunt work (aka campaign setup) freakin hate it!!
Hey great post!
Did you happen to run all 10 angles to just one prelander? Or did you test multiple prelanders along with the creative angles?
this post is gold for newbies like me to start walking before running.. thanks for sharing it mate
You welcome, sorry I didnt reply sooner! Stupid Disqus wasnt showing me notifications. LOL
thanks for the post!. by the way, what is CVR ?
CVR is conversion %. Total number of people that convert from a total number of visitors you send to offer.
Great post! Angles play an important part in it.But recently lots of CPI offers don’t allow pop,adult traffic,another biggest thing is how to lower the cost of clicks,do you have any suggestions on how to choose the traffic type ? Or if I drive those traffic to my LP and running it, is it against the rules?
Your CR is astonishing I’m running an offer with 0.3% CR do you think this is a bad offer as you mentioned in your post and should I let it off?
Offers get saturated, new rules and regulations happen. This is what we call the evolution in affiliate marketing. This evolution is really fast.
When I ran it, the offer wasn’t so saturated.
I could go more aggressive back then than you can today, etc.
That’s why the higher CR.
Hi Attila.. when you say 2 design, you mean 2 totally different banners or just different in background colors, cta button, etc.. How about animation?
Two completely different banner designs.
Hi Attila! Thanks for the post! DId any of your tests have landers involved or were they all just banners to start off with?
They were all banners yes
Awesome case study Attila, I always come back to read it when I plan to start a campaign. The part I don’t understand is: set my frequency cap at 3/24 per ad (Golden Nugget: anything more will just lower your CR %!). Can you please explain this in plain English. Thank you.
Hi! Why your visits and clicks are equal on the screen?
Because voluum did it that way, i don’t know. Remember I am not selling you anything, so I have no reason to fake a screenshot if that is what you are implying 😉
Sure no, even didnt think something like that 🙂 I just wondering is that direct linking or you have LP?
Dude, I really admire the way you are willing to share information for free. The way you drop bombs so raw and selflessly to help people. You rock, dude. I hope I make a killing with your info. and hopefully you keep contributing. Awesome posts. Just found your blog today.