Home > Mobile > Follow-along Campaigns

---

Mobile Pin Submit-Follow Allong (28)

---

I am gearing up to run my new mobile Pin Submit campaign.

I have 15 banners with different images/animation/color.

I have about 7 angles with each angles having 5-7 ad copy to add to the banners to test, though i'm only using one CTA for all banners. Q: How long do you recommend an ad copy be for a mobile banner?

I have 4 LP of different styles to test.

Let me point out that all of my creative and LP are all unique none were copied. While working on the LP's I decided to throw $60-100 at some of the ads DL and see what kind of data i get. I got data but i would love some help reading it and understanding it. I understand there's not much data to make any conclusion I just got anxious while the LP's were coming along and ran the campaign. What i wanted to see mainly was if my banners were any good.

I did Smart CPM in Decisive. I'm am 100% sure i did that wrong. Blew through $100 like there was no tomorrow, not that i really care just want to make sure i am setting this thing up correctly.

STATS: The two Different campaigns you see in the first screen shot where two different angles. I wasn't sure if i should test all the angles in one campaign or not. Any advise on this?

http://i.imgur.com/Q4ywUBa.png

http://i.imgur.com/a2mP4zE.png

http://i.imgur.com/wkKJzXC.png


Once my LP's are done i will be running this campaign fully and updating this follow along so ya'll can help me if i am doing anything wrong.

---

Do you have conversion tracking set up in Decisive? It might be worth firing a test conversion, if you haven't done so already, to see if things are converting OK.

---

I set it up in Voluum as recommended here . Not sure how to go about testing the offer since it's international mobile. The testing link on the network is browser base.

---

Damn forgot to add a Follow Along tag. Is there a way to modify the post.

---

Edited title for you.

Have any conversions occurred according to the network you are running the offer at? The Voluum set up should work fine you just have to make sure you are passing unique click IDs from Decisive with the {{{bidhash}}} token.

Looking at your stats, all the creative performances CTR-wise look similar. Are you just changing ad text or are they very different banners? Ideally you would split test completely different banner designs to start.

Look at the carrier/app/site tabs instead and look for poor performers. These can be culled off before you even put a lander in since a lander isn't going to boost a dismal placement CTR.

$100 direct linked with no conversions isn't a great sign but it is a pin submit, so a good pre-sell lander is critical.

---

I should mention I was testing the same offer from three different networks, and no there was no conversion with either network, though i was hoping even if i wasn't using a LP i would have had at least one conversion. Here is a screenshot of my Voluum setup. Please advise if i have done something wrong.

Voluum Traffic Source setup:



Voluum Offers Setup:



Voluum Campaign Setup:



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

The Banners in each campaign were the same except for the text. What i was thinking is creating a campaign for each different banners that design i have, since i was matching the banner design with the angle i was taking and just changing the text and find the right banner/text combo (10 unique banners x 7 angles x 5-7 ad copy/angle=$$$$$$$$ TEST BUDGET ). But maybe i should put all 15 into a campaign and find the best performing ones, then experiment with changing the text.

These are the data for the Apps/Carrier for the campaign i spent the most on in the pic i first posted.

Apps:



Carrier:



Note: That $100 was spent in a day . But i was thinking that i didn't care i just wanted data. Like you mentioned this is a pin submit so i am not shy with spending money to test.

---

I would like some feedback on my campaign setup.

Campaign #1:
Target: Mobile-App
Banners: 5 to start(different animation, one angle)-My idea was to start with 5 banners, then delete the non performers and rotate with new ones. When i find the best performers then i would test different angles with them. Should i continue down that path or try multiple banners/angles from the get go.

Campaign #2:
Target: Mobile-Site
Banners: 5 to start(different animation, one angle)-My idea was to start with 5 banners, then delete the non performers and rotate with new ones. When i find the best performers then i would test different angles with them. Should i continue down that path or try multiple banners/angles from the get go.

Also should I cull in this order-Placements-Device-banners or Banners-Placement-Devices?

Does anyone know how to work the Smart-CPM in Decisive? How do you calculate the CPA?

Thank you for the feedback

---

UPDATE

Campaign #1
Target: Mobile App
Spend (so far): $20.48
Bids: 19,076
Wins: 13,122
Clicks: 90 (67%)
Banner 1: Wins 2,751 (70%)
Banner 2: wins 2,647 (68%)
Banner 3: wins 2,611 (68%)
Banner 4: wins 2,655 (94%)
Banner 5: wins 2,631 (30%)
CPM: $1.55 CPC: $0.23

Campaign #2
Target: Mobile Site
Spend (so far): $20.48
Bids: 37,882
Wins: 32,037
Clicks: 147 (46%)
Banner 1: Wins 3,471 Clicks 21 (45%)
Banner 2: wins 4,742 Clicks 21 (44%)
Banner 3: wins 4,542 Clicks 20 (44%)
Banner 4: wins 4,493 Clicks 21 (46%)
Banner 5: wins 4,609 Clicks 14 (31%)
CPM: $0.47 CPC:$0.11

LP 1: 90 visits 5 clicks CTR 5.565
LP 2: 90 visits 2 clicks CTR 2.22%

NO Conversions yet.

Budget $100/d

To Do
Add 5 more banner designs to each campaign
Start Cutting placements
Then Cut Devices

---

Looks to me like your banner CTR is too low (Camp 1: 90 / 13122 = .6% CTR;Camp 2: 147 / 32037 = .45% ). Try to get that to at least 1%.

Also your win rate is pretty high (85%) so you could probably lower your bid and still get some decent traffic. With Smart CPM, I read somewhere on here to bid 30% above your offer payout as the CPA (If your payout is 10$, set Smart CPM bid to $13 CPA) and that's worked OK for me. You just want to be sure to watch it closely.

Also your LP CTR (2-5%) looks very low as well, I shoot for 20-50%+ depending on the vertical.

It seems like you don't have solid creative/LP yet, which is going to skew the rest of your data. (Placements/Devices might not be bad, but rather your creatives).

My recommendation would be to nail down a solid angle/banner/LP using a lower payout (perhaps CPI) offer in the same vertical, and then take that knowledge and apply it to the higher payout PIN submit campaign. Or maybe switch to a country with a lower payout PIN submit offers that will give you data much quicker.

You have lots of "data" but no conversions, which is the only data that matters.

---

Thank you so much goodrich for taking the time to right that feedback, i really appreciate it.

My question would be how do i test angles/banners. I have 10 banner designs should i upload all 10 with one angle let them get a few impressions see how they do. IF the CTR is not over 1% then upload them again with a new angle and so forth until i find the right angle/banner that gets over 1% CTR. Then i can work on the banner/LP.

Also what what point do you know the banner/angle is not working. Is it by impressions or how much you've spent on them compared to your offer payout?

Thank you for taking the time to look over my follow along guys.

---

Read this gold for more info about angles: http://stmforum.com/forum/showthread...gnoring-angles!

Basically you want to test angles and not designs first. So instead of 10 designs for 1 angle, 10 angles with one design/layout.

Pick out the best few from the initial batch and launch variations on those, rinse, repeat.

What you've spent compared to offer payout, aka ROI. CTR is important, but only if it converts as well. This is why it helps to start with low payout offers, because if I want to spend 10x offer payout on 10 banners/angles that would be $30 for a $0.30 offer, but $1k for a $10 payout. And this is just the first test...

---

UPDATE 2

I have 4 campaigns setup. 2 on WiFi, and 2 on mobile Direct Linking. Below is my overall stats from the day so far on my 4 campaigns. I have 32 conversions between two networks, one more then the other. All conversions came from one connection type. my tracking stats doesn't show all of it since i had problems tracking conversions.

I tested some angles with the same banner design, got some one them at or above 1%. I'm going to let it run some more them pick the winning angle(s) make variations, upload with new angle and rinse & repeat. For the angles that did the best i will split test with a landing page. I am looking to expand to a new traffic source Decisive is not spending my budget. Any suggestions please PM me ASAP. need to get this ball rolling.

http://imgur.com/gS36WQj - Overall Traffic Stats

http://imgur.com/GBi3Hhy - Banner Stats

To Do:
Create new ad copy for working angle(s)
Create new ad sizes for working angle(s)
Test LP vs DL for working angle(s)
Create/Test New Angles/ Ad Copy
Test new traffic sources (recommendations welcomed)

---

UPDATE 2 Cont...

I was so tired when i posted the first one i didn't know what i was writing. (Still tired)

As of now i spent $223 between the 4 campaigns (1 day running) All the campaigns have 16 banners SOOO YEAH!. I've made $xx back. I'm not sure if i should let the campaign to continue running before make tweaks.

What sucks is since i started recording conversion late i can't tell if app, site or banners is bringing me most of my conversions. With the data i have it shows app, even though it has the lowest banner CTR. Maybe my AFF network have the information.

Should i let this campaign continue to run or should i pause and start optimizing?

The thing is some angle/banner have a higher CTR on site then on app no matter wifi or mobile. maybe the color of my banners are not working well with the apps.

From now on i will test 5 angles with maximum 2 banners/ad copy per angle. 16 banners is crazy. Then after 10x offer i will delete/replace non performers, and make variations of good ones.

Some ideas i had was like bigeasy123 mentioned creating pools. Now i see what he meant. Testing 16 banners in one campaign and having those mix with good performers and bad is a bad idea.

So I will have a test pool, where i throw in new angles/banners to test. I can bid high on this pool to get data. Then when i see a performing angles/banners, take it out of the test pool make variations of it and add it to the profit pool. Also doing that for placements/devices. The "Profit pool should only contain the placements/device that profited on the test pool, blacklist all the rest.

WHile you test new angles/banner in the test pool once you get profitibale angles/banners & placements/devices (If different from what is in profit pool) add to "profit" pool.

MATH TIME:

4 (wifi/app, mobile/app...etc) campaigns x 2 pools = minimum 8 campaigns




TIME TO WORK

---

Did you pass the subids through to your affiliate network? If so, you should be able to get a list of converting subids, and update them in your tracking software to get the full information.

---

I will hit up my AM an check. How do i upload it to Voluum. I guess i will hit up support for that one to.

Edit: I am not running a Pin Submit any more. I took the advise of colin and I am running a lower payout CPI offer.

---

I need some guidance in reading the data i have. After spending $350 on the campaign I have 65 conversions from across 2 days realistically. I fee like there isn't enough conversions to determine winning angles/banners.

Should I start cutting complete nonperforming angles/banners and creating new ones?

with the winning angles should i start to test different banner variations now?

---

This is KEY when testing ideas.

Test the core angles first, the deisgns itself don't matter all that much. Once you have the angle infiltrated, THEN make variations of the design.

---

Are there any patterns emerging? 65 conversions is plenty to start telling which angles and which placements are showing some promise. Are the conversions clustered anywhere, particularly on an individual angle?

Do you have any placements or placement / angle combinations that are in profit, or close to it?

---

Some of the conversion report was not coming back to my tracker, there was an error with the global postback my AM setup in my account. I ask him to send the converting subids but i am not sure i passed the subids for the creatives. I have about 23 unaccounted for conversions, which could sway the winning angle.

---

See if you can get a list of converted IPs. You could then contact Voluum support and ask if they can give a list of click IDs for those IPs or if they can add conversions to them manually. At the moment in Voluum you can't ever see the {clickId} values which can make situations like this a bit more difficult. Perhaps they will make a Click ID export function at some point!

---

Thank you caurmen and zeno. My AM was able to get me the subids and IP and i was able to add them into Voluum which really helped my in figuring out my winning angles.

I have a question. Should I create a separate campaign for each angle, then create a separate campaign for app/site, network/wifi for each angle?

---

Generally this is a good idea - try not to mix separate angles into the same campaign. It makes the stats easier to digest.

---

That is insane. So for 3 angles basically that is 12 campaigns. I guess that is the hard work that seperates the profitable campaigns from the rest.

Is there any tips on maintaining order with all those campaigns/banners?

Side note: the feedbacks and help on this forum has been amazing. I'm learning by doing but at the same time it is appreciated to get a helping hand understanding what I don't.

---

Are you talking about campaigns at your traffic source or in Voluum? I wouldn't do that in Voluum.

When I say 'campaign' it really depends on the traffic source and how it treats campaigns vs banners, where the budgets get set, etc.

On FB it might mean splitting things into separate ad sets. On Decisive it may make sense to split angles/targeting into separate banners but having these all in the same 'campaign', provided the banners all get their own budgets.

---

I was referring to having a campaign for WiFi/App, WiFi/Site, Carrier/App, Carrier/Site for each angle at the traffic source. So for angle #1 i would setup a campaign for WiFi/App, WiFi/Site, Carrier/App, Carrier/Site (4 campaigns). All with the same banners in them for angle #1. Is that to granular?

I ask because i remember seeing that you should for mobile split your campaigns into WiFi/App, WiFi/Site, Carrier/App, Carrier/Site. So i figured if you have lets say 3 winning angles then you might want to create 4 campaigns for each at the traffic source.

---

Yup, correct. You split by WiFi/carrier this mainly because you can't separate the stats into Carrier > placement, i.e. you can see X app is doing well but can't tell what clicks/impressions are from WiFi and from carriers... so optimising is painful.

On the other hand, you could run Carrier + Apps and sites since Decisive provides separate app/site tabs. However, apps and sites can be quite different in terms of what the user is doing and how ads are served so it is useful to have them separate, simply so you can see separate global performance stats at a glance (i.e. camp cost + rev).

Things like this are often tedious. But dodging difficult things in IM (without taking a beneficial alternative) is a sure way to build bad habits and an equally bad mindset. Build a process and things become less tedious and more automated. Decisive lets you clone ads/campaigns anyway so it's not hard to make these combos in minutes.

---

I had a quick questions to post.

When moving to a new traffic source. Do you first run your winning angles from the previous or do you start from scratch testing out all the angles you tested before?

Should i wait until i have a solid angle/creative combo to start testing LP's or should i already start testing that with my winning angles?

How often should i be rotating in new creatives and at what point do you stop introducing new creatives and just let what ever is working work?

So far this is the method to my madness:

Find Offer -> Research Offer -> Pick a Traffic Source -> Create 5 angles for offer -> Create 2 ad copy for each angle -> Test out angle/ad copy using one banner design -> DL and gain impressions/clicks/CV (I usually wait 2-3 days)-> pause and look over stats -> pick winning angles(s) -> create 10 diverse ad copy for each winning angle(s) -> create 10 new banners with different design to test ad copy/angle -> If more then one angle split each angle into its own campaign using the same new creatives -> Split test LP's ??? ->Create new angles and (go back step 5)


What do you guys think of my process? Any advise will be welcomed.

EDIT:
Man when it comes to tracking i am STUPID!
I really don't understand why it confuses me so much. I am going to write my thoughts and ya'll tell me where i am wrong. Ok so you go to a traffic source, you get their postback or pixel and you place that in your Tracker??? with their available macros ??? then you setup your campaign in your tacker get the URL (The URL should have the traffic sources available macros appended to it???) and place that in your traffic source campaign ??? then you get your trackers postback or pixel and place that in your networks offer ???

---

If i have a winning angle, i'll usually run that. If there was something that came close on another network sometimes i'll test that too. Usually something that did 50% ROI on 1 network won't shoot up to 150% roi on another.

I always testing my angle with landing pages and banners that MATCH the angle. The landing page design isn't important for this phase, just the fact that the landing page represents/matches your angle.

Always introduce new creatives. It's easier said then done and very often i get lazy and let it ride out, but sometimes i have a spurt of madness and make 100 new creatives and add them to the rotation. Ideally always be testing.. ABT!

---

Home > Mobile > Follow-along Campaigns