Home > Mobile > Follow-along Campaigns

---

[Mobile] 15 campaigns / mo until I make it (23)

---

Hi there,

Volkan here from Istanbul / Turkey. I have joined STM last week and I have been reading anything I can find about mobile. Actually this won't be my first time creating a mobile campaign, I have created 5 mobile campaigns in about 10 months and 1 of them was almost profitable.

I have around 500$ / mo budget. Since it's not much I am going create at least 15 campaigns per month. Any revenue earned from the campaigns will be used for testing/scaling existing campaigns & creating other campaigns so I may create more than 15 if my ROI isn't -100% (P.S. I count "wifi/app", "wifi/web", "mobile/app", "mobile/web" as 1 campaign and not 4. I can track it with Voluum using custom variables so I don't see a need to create seperate campaigns for that purpose.)

I have studied computer engineering in university so I have coding skills, I also know how to use PhotoShop, so hopefully I won't have any hard time creating banners and landers, just making them convert

DAY 1
I am already approved @ PeerFly, NeverBlue, YeahMobi, MobPartner, Matomy and I have just joined F5 Media.
I have accounts on Decisive, mMedia, TapIt, BuzzCity, Mobfox, AdModa & JumpTap. (But really going to use only one for now, jumping around from one traffic source to another never did me any good in the past)
I will be using Voluum for tracking and a CDN to host my LPs.

Got the top offers sorted by volume / epc from YeahMobi waiting from F5 Media.
MobPartner doesn't have or won't give any info on volume. But at least I got this list of campaigns sorted by EPC : http://www.mobpartner.com/files/publ...gn_Silver.xlsx
I don't know if it means anything without volume, as far as I know someone may have send 10 clicks to a $10 offer and converted 1 which makes it's EPC $1.

DAY 2
Selected an offer from each network, brainstormed angles and created banners for each but created LP only for Offer #1[CPI]. I just want to first test Offer #2[CPS] with DL.
F5 also have Offer #1 with exact same payout as MobPartner, so I applied for it too. Sent my ads & LP to my affiliate managers, waiting approval from the advertisers.

---

All the best mate

---

Thanks mate.

While waiting for my ads to get approved, I revived the campaign I mentioned in my first post, the one that is almost profitable. The only thing I changed was to raise my bid from $0.03 to $0.05
So far it's good:

---

All the best and i'll follow your journey

---

Nice man, love the attitude!

Since you're creating so many campaigns try to focus on keeping a good flow, you don't want to become lost in it all and for frustration to creep in, but nonetheless get at it man I'm watching this thread !

---

stacking your money!

---

Thanks alot for your good wishes guys.

So I have got one question, publisher X has 60 clicks and no conversions. The average payout of the offer over 3 networks is around $0.35, and the cost of that publisher to me is 60 * $0.05 = $3, it's more than 8 times the offer payout.
I am tempted to kill it, BUT the clicks are spread like this:
Carrier 1 : 33 ($1.65)
Carrier 2 : 6 ($0.3)
Carrier 3 : 3 ($0.15)
WIFI : 23 ($1.15)
Is it safe to assume that publisher X is no good. (My other campaigns which is profitable has 1 publisher with almost 10% CR on Carrier 2, and zero conversions on Carrier 1)
Should I block the publisher only on Carrier 1? (Which will require me to create 3 seperate campaigns on the traffic source for each carrier)

---

DAY 3

Stats:
Expense : $50.53
Revenue : $0.64
Profit : - $49.89
ROI : - 98.7%
(The other campaign that I resurrected is profitable like $5/day. But I will not add it's stats or mention it from now on in my updates since it was a campaing created long before this follow along.)

I spent too much! You can read the reason for that in my previous post. In summary, I tried to give each placement enough clicks from different carriers but while waiting clicks from Carriers 2 & 3, Carrier 1 kept eating my budget. That's why I am changing the way I create campaigns on the traffic source, from now on it will be like:

Offer Name - Country - iOS - Wifi
Offer Name - Country - Android - Wifi
Offer Name - Country - iOS - Carrier 1
Offer Name - Country - Android - Carrier 1
Offer Name - Country - iOS - Carrier 2
Offer Name - Country - Android - Carrier 2

Each campaign will have a budget set depending on the payout, and will be stopped if it's not at least -50% ROI after the initial budget is spent. Also each placement in each campaign will be cut after spending 1x - 2x the payout without any conversions.

Other than that I am having difficulties choosing offers, most of the offers in the top offers list I got from F5 Media are sweepstakes and I am completely useless when it comes angle creation for sweepstakes.

---

In this situation I would advise killing placements before killing carriers.

However, in some countries certain carriers may be a PITA, so your approach of splitting by carriers is sensible - provided the volume is there.

It will at least make the testing compartments a bit easier to digest.

---

DAY 4 & 5

Stats:
Expense : $56.1
Revenue : $0.96
Profit : -$55.14
ROI : -98.2%

It's taking too much time and money for me to test a campaign, I have to simplify how I do it. I also think I am over analyzing, even if it has only converted 9 times after 1400 clicks (cpc $0.05 - payout $0.3) I still try to block the publisher that has spent too much and give other publisher a chance to convert. I don't know why I do that it's stupid...

While I was promoting PPV offers I would DL and rotate 3-5 different offers from the same vertical and throw $15 worth of traffic to each offer and then only work on the offers that has generated conversions. I wonder if I can try it with mobile offers. At the moment I come up with angles, create banners and LPs, get them translated, create campaigns on Voluum and ad network which costs me around $50 per offer and a day. And another day to collect data. What if I choose 3 to 5 offers from the same vertical create some angles which aren't too specific so I won't need a LP and test them on 1 or 2 traffic sources. Even a -70% ROI for an indivdual offer would be a good sign for me, and I can concentrate on that offer.

---

1400 clicks to the offer?

If that's the case then it's your angle, the traffic, or the offer that aren't performing.

There's no point in sending more and more money through the funnel when it's simply not converting.

Going broad and rotating multiple offers is a good approach provided you are comfortable with drilling down into the data to figure out where your highest earnings are coming from.

---

Thanks for the reply zeno, I have already created a campaign in France rotating 3 similar offers waiting it to be approved. My angles are sort of ordinary as I am not using a LP. My thinking is if I can get even a few conversion with simple angles & DL then the offer-traffic source match is good and I can make it work. Since I am a noob, I may be completely wrong =)

---

What kind of offer is it?

Some offers and approaches need a lander more than others - so much so that starting direct-linked can be a waste of time.

Additionally, landing page CTR can give you a preliminary metric on how well your angle is engaging your audience.

---

They are App & Game Installs.

Hmm never thought of it that way, excellent tip thank you very much.

---

Game installs generally don't need landers but some app installs will do better with them.

For simple game/app installs I would focus on direct linking and playing with angles to get traction. Traffic source matters a lot here, so either test a few or - if Decisive, try raising the CPM drastically to access higher quality placements.

---

I would advise Never try Web adnetworks for Appinstalls.

Try Mobile SSP networks (The ones that provide developers with SDK's).

---

I was just gonna ask how much of an impact does the traffic source makes with app/game installs, thanks for answering my question before I even asked it


Fortunatelly I am using 3 networks and they all have SDKs, but what's the reason behind it?

---

Hmm, just looking at these stats - have you done any testing to see if you're getting a lot of bot traffic?

http://stmforum.com/forum/showthread...367#post180367 is the method I developed a while back - seems to work pretty well to detect dumb bots at least.

---

Day 6 - 10
Stats:
Expense : $116.2
Revenue : $33.45
Profit : -$82.75
ROI : -71.2%

No, I haven't. Would you suggest making a quick bot test for each country on every traffic source before I create my campaign, and exclude the possible bot placement right of the bat?

---

Generally, yes. It depends on just how wide a net you're casting, but IMO it's usually worth doing.

---

Update:

So I have been creating campaigns and failing miserably. Below are some of the campaigns I have tried over the last 2 weeks. I have uploaded some banners & LPs to imgur so maybe you can tell me how they are.
I don't know what I am doing wrong but all of my campaigns end up doing -100% ROI or -90% at best. So I have got some questions for the more experienced marketers, I would be grateful if you can shed some light on them.

1. Choosing Offers : I ask my AMs and get the list of offers, but most of them are sweepstakes and I think I lack the imagination to create an angle for these kind of offers. The others are utilities which have been around for some time such as Du Speed Booster (BR) it has been downloaded 50 - 100 million times and probably every angle I can think of with my limited imagination has been done to death.
   
2. Maybe my angles suck? For example for the Grocery Voucher offer I tought of these:
   
   • Direct : "Win a $1000 grocery voucher." sort of angle.
   • "XX vouchers left claim your spot." type of angle (Banners with animation like a count down of available vouchers).
   • "Free dinner every day for a month"
   • "Throw a party / feast party, grocery bills are on us." type of angles
   • "Celebrating anything? Here is your $1000 grocery voucher. / Let us help with the bills"
   • Refrigerator empty? Fill it for free bla bla
   • Hungry? Prepare a feast for yourself. For free / bill are on us / etc.
   • Some LPs had geo script telling the visitor that he must live in "city" in order to qulify.
   • Some of them has Norway supermarket names to choose from.
   • Some were just direct "win a $1000 voucher" type of LP
   
   
3. My banners or LPs suck? Below are some examples you can comment on.
   
4. Can it be the traffic source? I mainly used TapIt & Decisive. Only 1 offer I tried on mMedia & BuzzCity
   
5. It can very well be all of the above that leaves me thinking I have to test : 1 Offer x ~3 angles x 15 Banners (5 for each angle) x 6 LPs (2 for each angle) x ~3 Traffic sources, which would take $ (Payout x 810). Even for Dolphin Browser ($0.32) I would need over $250 test budget.
   
6. I said 15 campigns / mo. That's 1 campaign every 2 days. I think I rush some of the things and give less attention to some details than they deserve?
   
7. I saw some case studies which shows 1 traffic source can suck for an offer while the other may yield 100% ROI, so should I try my failed campaigns on other traffic sources before deciding it's a fail?
   
8. I first cull placements, is that a correct approach? I got $15 or so budget before I decide a campaign is worth continuing or not. So if a placement eats half of that I may not make a healthy decision. BUT what if that placement generated 1 or 2 conversions, still below -70%ROI maybe but it could have been profitable if I culled a carrier or a brand or a device or an OS. I may never know if I cull it.
   
9. Or maybe I am complicating things if the campaign isn't near -50% ROI after $15- $20 spend call it a fail. And then try one of the options:
   
   • Give another angle a try?
   • Test a different traffic source?
   • Test different banners or LPs?
   • Test another offer?
   
   

Banners & Landers

Sam Media Angry Birds Ringtones (MY) - Banners : http://imgur.com/a/6E4Pz
Sam Media Angry Birds Ringtones (MY) - LP : http://imgur.com/fY3qViG

Black Birds (TR) - Banners : http://imgur.com/a/mNamh

Dolphin Browser (TR) - Banners : http://imgur.com/a/XYnaq
Dolphin Browser (TR) - One of the LPs : http://imgur.com/yoqihds

iFortune (RU) - Banners : http://imgur.com/a/qS4VA

Grocery Voucher (NO) - Banners : http://imgur.com/a/YSrtT
Grocery Voucher (NO) - One of the LPs : http://imgur.com/0fSbud8

LP for Utilitiy Type Offers (Text can be changed to suit the angle) : http://imgur.com/bnmUvc0

I am sorry this is a long one. Thanks to everyone who has helped me till now and thanks in advance to anyone who anwsers my new batch of noob questions

---

It can be really frustrating when you're trying everything you can think of and just can't get your campaigns off the ground - much sympathy.

First question I would ask here, before I get to your questions: is there anything you always do the same? Same approach to banner design, same image source, same couple of networks? Same amount of testing before cutting placements? Anything that's a hidden part of your process that never changes?

And first tip: you're not doing badly at all here. If you've been launching 15 campaigns a month and you're two weeks into that, that's only about 8-10 campaigns. It could well be that you just haven't found the right offer / traffic source combination yet - there's a LOT of fishing to do before you find a winner. Have a look at Stackman's post from last year about how he hit mobile marketing after BKK - he tested more than 30 offers to find one huge winner.

1) Imagination can be trained. It doesn't look like you're doing terribly badly with your angles. I would, however, suggest getting off the beaten track a bit with your offers - don't just test AM-recommended ones, test ones that look good for any reason, or that you hear about on the forums. Test widely - it's OK to test a few lemons to find a gem. Also, I notice you're not testing PIN submits - you could definitely give those a go too.

How many networks are you with?

2) None of those angles look appalling on first glance - I think you're just going through the testing process. Maybe try and do a bit more research on your target geos to find topical or off-the-beaten-track ideas for angles.

3) Again, nothing looks terrible here. You could probably try varying the designs a bit more - there are a lot of banners with quite similar designs - but you're definitely on the right track. Have you tested your landers to make sure that they're actually rendering properly on the most popular phones in the geos you're targeting? Also, are you testing direct link too?

4) As MrGreen has proved recently, Decisive still has solid traffic, although it's likely to be a bit pricier in some of the top geos. Testing more traffic sources is a decent idea, but you might just want to stick to one for now - it's very unlikely that sticking with Decisive will be the sole reason you keep seeing -90% campaigns.

5) If an angle's not working overall, it's not worth exhaustively testing every combination of banners and LP with it. I'd recommend killing angles that aren't working fairly rapidly - more like after $120 spend than £250, and even lower than that for small payout offers. For the Dolphin example, I'd probably be killing sub-campaigns after one round of optimisation and placement removal if there were no obvious red flags (like one placement eating all the spend). You can probably stick with 2 angles per offer rather than 3 initially, too, to keep the spend down.

6) That sounds about right to me. From looking at your landers and banners, it doesn't look like you're rushing too much. Volume of output is important.

7) I wouldn't recommend that unless you have a large budget to play with. Find a traffic source that you know is profitable for other people, and focus on getting to profit on there. Adding in "test all of the traffic sources!" to your to-do list will make the problem space unacceptably large.

8) Yes, that's the correct approach. However, I strongly recommend doing a test for bot traffic too - a placement that's converting minimally might be converting that way because you haven't cracked the best banner combo for it yet, or it might be converting like that because 95% of its traffic is bots!

9) Taking this approach wouldn't be the worst idea in the world. It sounds like you're over- rather than under-thinking things: taking a ruthless "get through as many campaigns as possible" approach is a pretty solid way to success. Do watch out for confounding factors like single placements eating all your spend, but other than that, a "did it work? If not, dump it and move on" approach will probably work well.

As to what to do next: I'd recommend testing 2-3 angles per offer/geo combination. If the offer's not working in a geo after 3 angle tests, try a new offer.

Hope that helps! Looking forward to hearing how you get on next!

If you'd like more feedback, post details on some of your more successful campaigns and we'll all see what we can glean!

---

Thanks a lot caurmen, I really appreciate your time and help.

Actually yes, as you have noticed my banners looks like each other but I am checking other peoples banners and trying different things. I think I am improving on that.
I generally do my first tests on TapIt and sometimes on Decisive. I hate decisive for not allowing gifs that are more than 3 frames.
As for affiliate network, I mostly use MobPartner, F5Media, I have recently joined F5Media so it's mostly MobPartner. I also have accounts with YeahMobi, NB, PeerFly.
I cull the placements that has spent a reasonable amount and didn't convert. (e.g 3 times the payout for payouts < $1, but 1 - 1.5 times the payout for payouts >$3)
Same with OSes. If one didn't convert for 10 times the payout for smaller payout I block it.

Yeah I have read it wish I had more budget. Actually not finding a profitable offer is not the thing that's bothering me. What's bothering me is none of my campaigns were above -90% ROI, only 1 was and that was a campaign that I have created long before I joined STM. I recently launched it on Decisive, MobFox & mMedia (was profitable on TapIt) best campaign was around -50% ROI after optimizing. (-71.2% ROI on was my latest update was because of this campaign. )

1) Will do that. I have tried 4 campaigns listening my own gut but couldn't make them work. The one campaign that I made profit before was a PIN submit that wasn't on any AMs list. I just chosed it because it was a TR offer and I know what Turkish people would like/want.

2)I think you are right they do not look terrible but they are not great either. They look like anyone could come up with them if they gave it a thought a couple of minutes. From now on I will take my time when brainstorming angles, and try to come up with off the beaten track as you said.

3)Yep, I always test DL & LP. I check my LPs with my own phone and using the Opera Emulator.

4)OK sticking with my traffic sources for the time being then.

5)Got it. Just 1 question, how much would be the payout of the offer that would make you spend $120 for testing an angle? Not Dolphin I hope

6-7) Ok thanks

8)I will use your methods to find bot traffic on a traffic source before I create a campaign. (Should I do it for each geo, most placements are global but there are some country specisic apps/sites)

9)OK, I am motivated again thanks to you, will try new campaigns and post the results.

Again I thank you for your time and help caurmen, much appreciated.

---

Home > Mobile > Follow-along Campaigns