Home > Mobile >

---

Mr Green's Mini Mobile Case Study. (48)

---

If you aren't already smashing it on mobile, then this should help sort out your process.

Below is the exact process I used last week to make a $25 Whatsapp Netherlands work on Decisive (as well as on other countries).

Step 1:

Create 3 campaigns:

• Android
• Windows
• iOS

Creatives: Use WhatsApp branded banners (Mr Green tip: this will allow you to get access to premium inventory that “download/play/antivirus” style banners won't get you).
- Bid: I set to 0.31 CPM (the extra 1 cent is to bid above those lazy people who bid in 0.05's and 0.10's). My goal was to see a decent chunk of traffic. I was getting around 70% of wins.

Step 2: Let it run. Cut out the crap that is spending a lot but not converting. Usually placements like “6snap”, “6tag” etc.

Note down the gold nuggets.

Exhibit A (this was data from the Windows campaign):


Step 3:This is where things get interesting! Separate the gold nuggets into individual campaigns.

I want to win as much traffic as possible here, so I set the bids so I would be winning 90%+ of the traffic.

Bidding around 0.51 manual CPM got me sitting pretty.

Step 4: Optimise handsets, carriers and bids. Test more banners. Use the same targeting to scale on other sources. Ride the profit wave!

-----------------------------------

FAQs:

Did I use a lander? Yes, it's a simple Whatsapp download lander. No, I won't share it. No, it's not magical. Yes, you can easily create one yourself.

How do I know if I get access to Decisive premium inventory? Email support. Can't be bothered? Create a Hotel.com app campaign with a branded banner you will see the placements you wouldn't usually see before.

What is defined as a "crap" placement? Since the payout here is high at $25, I will cut any placement that spends $25 and doesn't convert.

Why aren't I bidding Smart CPM? I found that you need to spend a boat load of money to find winners if you are running high payout offers. I typically use SCPM for low payout offers.

Chur Chur Chur!

Mr G

---

Fantastic!

What sort of volume did you end up seeing on the individual placements?

---

Each placement seemed to spend between $20-$40/day, but when you add those up, combined with the ridiculous ROI it was a pretty sweet autopilot set up.

---

Awesome case study.

What's the lifetime on smaller-volume mobile display campaigns like these? How often do you need to refresh creatives?

Thanks for the data screenshot!

---

It really depends, sometimes they can last for months. The beauty of this strategy is if you bid high enough then other affiliates won't even see these placements when they create new campaigns. Once I have a winning creative I'll test a new batch every week or two.

---

Great case study short n sweet.

---

Hi Mr Green,

this is a super newb question so I am apologising in advance.

Where is the $25 coming from? I was under the impression CPA is cost per acquisiton, or the amount your are being paid per action. any chance you could help some of the newbs out with the math here. I am sure it is something very very simple I am missing, just can't for the life of me see it

---

@aady83 $25 was the payout I was getting every time someone converted on the Netherlands Whatsapp pin submit offer I was promoting.

---

yikes, thats pretty sweet and completely mental ROIs!

what i dont quite get it , the payout was coming from the aff network, for example I have only seen whatsapp at .5 cents per install. if its the case that that was what you were getting then got it all and hats off to your good sir! if its not, I am a little lost as to where the payout comes from.

take line 2 for example.

spend = click (622) x CPC (0.02) = 12.44 spend
conversions (12) x payout (25) = $300.

although i have a feeling i am missing something as line 1. cost (same calculation is approx 350) conversion/payout is 550 which is still a + so unclear to me why it was cut.

---

Yep the payouts differ a lot based on what action is expected for the conversion to occur. Those .50 cent offers are for app downloads, the $25 pin submits actually charge users. Also payouts by country vary immensely.

Your calculations are correct. However, I didn't cut line 1, I just didn't separate it into it's own campaign. I cut lines 4 and 5 though.

---

Thanks for confirming that calculations - that is insane! I am almost speechless (apart from silly questions :P)

---

May i ask what you mean by "Whatsapp branded banners"? Do you get them from whatsapp or do you create them yourself?

---

Do you let the RON campaigns run in hope of finding new golden placements or do you just create campaigns targeting the profitable and scale that up?

---

How do you handle Decisive being so bloody slow? I tried running some traffic there and had to abandon. It was unusable for me.

---

Great share bro! Shows how simple it can all be, just need a good working process!

Chur Bo!

---

Great share. Thanks for your thread!

---

Nice case study! Were you only bidding on carrier traffic?

Cheers

---

Loving the simplicity of this, a lot of the time this is all you need.

Scaling this takes work, but this is exactly how i test new campaigns.

---

wow it looks so easy when you do that haha , so inspiring thanks bro

---

Basically an ad that looks like an advertising agency designed it, not an affiliate.

Yes I continue to let it run, while constantly cutting the fat. Like I mentioned before, sometimes you won't see placements because other advertisers are buying out the inventory...sometimes they let their campaign slip, and that's when the gems show up.

The only problem I have is with pulling big reports...I will go and make myself a coffee while they are loading haha.

---

Mr Green what do you do when you are getting traffic from 200 pages of apps. Eg. few thousand apps and only some of them give you much imperssions. Most are slowly eating your budget by having 2,5,10k impressions per day.
This is a problem that i see lately i and i am not sure how to solve. Do you target just the top 10-20 placments by v voulme or you run ron? If you have to wait for 2000 apps to acumulate $20 spend when most of them dont work its pretty bad situation.What is your mindset for targeting/cutting placments ? Do you cut based on ctr ?
tnx

---

That's awesome indeed!


I have a question for you Lorenzo:
Looks like the WindowsPhone_SocialNetworking is not blocked yet? Were you still trying to get it profitable with different Banners/LPs or 8 conversions were enough for you to keep testing that source with the existing creatives despite the $46 CPA?

---

I took this screenshot before I did the blocking. Those bad placements ran wild on me, I didn't check them fast enough.

---

So I did a little math.

You had 4,138,327 Impressions

You spent $1282.88 on your testing just for windows?

---

That sounds about right. However, it doesn't mean you can't test effectively with a smaller budget. As you can see I let some placement overspend by a decent amount.

---

#mobilecpamarketingissoeasy #nailmobilecpamarketinglikeaboss

Great case study Lozo! Short and straight to the point - easy first blood

+ I always heard people say that it is impossible to make Decisive work - where this statement is again, proven to be invalid.

---

Great case study! Did you split wifi and mobile for each campaign or is that something you optimised after getting some data?

---

question re how you chose the offer.
reading the mobile appetiser it recommends going into a campaign taht is sub $1. Would you recommend looking at campaigns like this for newbs, or steer clear until there is more experience in mob marketing?

---

We recommend that because low payout offers tend to pop up when a) in geos with low traffic costs and b) when the conversion point for the offer is low-barrier, i.e. easy to achieve.

This makes it easy for newbies to get actionable data based on ad cost and revenue.

For high payout offers like this one you need to be ready to commit more to testing - the kind of ROIs seen here happen infrequently! It could well have gone $1200 cost and $1300 revenue, with the campaign running at a loss until day 3.

For a newbie, spending say $500/day without yet having much experience is a hard ask.

---

I optimised it after. Mobile killed it though, having half the CPA as WIFI.

What Zeno said!

---

Not horrible, just not as high as you would get for those download/play or facebook style banners.

---

Thank you very much

LOL… so embarassing

---

So what kind of CTA will you use so the banner will "feel" branded?

---

You can still use "download", "install" or "click here".

Just step into Whatsapp shoes; use their colors, use their logo, keep it professional looking. Their homepage is your cheat sheet http://www.whatsapp.com/

---

WOW! Mr Green another amazing post.

Some Questions!

You say "And let it run".

1. For how long did you let it run before starting deciding what to cut on placements?
2. How much did you spent till that point ?
3. Do you first cut placements, then banners and then campaigns ?

Thank you again for another so informative and eye-opening post!

---

Hi Lorenzo, you created 6 campaigns as below

Iphone app
Iphone site
Android app
Android site
Windows app
Windows site


Or you just targeted apps?


Thanks!

---

I have printed this post ^^^^^^ and have it near me in my desk. Short, simple and effective -Thanks

---

Is there any software or method to see what other lazy affiliates are biding? So we can have the advantage to bid higher to gain a concrete position.

---

would love to know that^^^^^, but i have never read of such a thing for mobile..

---

Haha I don't think there is any public software that offers what you are asking. You have to get the edge on your own.

If you look at the planner on decisive you can get pretty good insight. They also have a n API that can make bidding easier.

---

@mrgreen quick question.
i am running some campaigns in AU where i am getting conversions but struggling to get any traffic at all and i am winning 64 perfect of the bids on that particular campaign, but i have it segmented out in Appetise style which i am beginning to doubt specially in Decisive. I reached out to my Decisive AM and he advised to group everything together kinda like what you did here but he said to group everything android, iPhone, windows, wifi, mobile and see what converts and then separate it.
The idea sounds OK and i am gonna test it, which should be easy cuz i already know what angle works and banners, but lets say that i do that and then i get the conversions similar to what you are doing here, after that i have to separate the winning placements, banners and all that in their own campaigns and i am thinking that once i do that the traffic will stop again and will be in the same place again.

Whats your take on this? how do you deal with Decisive and this slow traffic deal.

Thanks

---

This is how much traffic they have available for you to bid on in AU:



If you break it by content type you can see it's heavily in favor of apps:



If you aren't getting anything close to that volume when you are winning 64% bids, I would put it down to either you campaign being too aggressive, or your bids to low and skipping bidding altogether on some of the placements.

Also make sure you have more than $100 budget in your account funds and campaign, or you will be throttled.

---

@mrgreen,
1-attached is what i am bidding, and the other attached shows the traffic i am getting over all in this particular one. Which by the way this particular campaign is already green but very little gain as i get almost no clicks and i haven't separated the winning placements into their own campaigns, also like i said if i am getting this very little traffic now i fear that when i separate them out i will get no traffic.

2-Another thing that is killing me in AU is the high CPC across the board, which is burning thru my test budget, this has been the only campaign i have been able to get CPC at around .04 to .01 the rest is too high, i understand how CPC is impacted by CTR is different things, i guess my question on this would be based on your experience what would be an achievable low CPC in AU?

by the way the payout is 2.5 for this. Thanks again for your help mate!

---

There is definitely some potential in your campaign based on those initial stats!

Your traffic will definitely be toggled because of your budget. AU is usually pretty competitive. Another thing to note is this month is typically the most expensive month traffic wise for all mainstream sources.

If I were you, this is what I would do:

• Pause all creatives other than your top 4. (you have too many running right now, you can add more in later).
• Separate your campaign by android, iOS, windows, apps, and ,sites (so you should have six total, $30/per campaign).
• Bid SCPM at around $1. (This will let the Decisive platform find your CPM sweet spot. You will be able to see what CPMs you need in order to get traffic from certain placements. If you still don't get any traffic then bump it up).

---

what you described here is what i did at the beginning, i separated them out in windows, iOS, android, app and site and i bid SCPM at around 1 dollar!
then i started seeing the super high CPC for a lot of the separate campaigns and this particular one (windows-app-mobile) is the best one. that with out optimizing is already green, i guess i can clone this one and try to bid at 1 dollar SCPM again, kill the creatives that haven't converted get some more data and then separate out the winning placements into their own campaigns..

---

Haha ok nice getting green of the bat is great!

Definitely kill some creatives, then experiment with bids.

---

Thanks man, will follow with outcome in a few days..

---

nice post!!!!! but i have a few questions. Do you focus on apps only and block sites or do you do both? and also do you recommend selecting placement categories on the traffic source?

thanks

---

Home > Mobile >