Home > Mobile > Follow-along Campaigns

---

How many Pins can a Newbie Pop with Sweeps? Hopefully a lot... (15)

---

Okay so, I have four weeks to make this AM thing really start to happen. Sweepstakes certainly isn't the most original vertical to start in or the least competitive but it seems a relatively good entry point and at least I'm getting conversions with some of the pin submit campaigns I've tried and getting some tastaay data to digest.

I started this thread about my opening strategy a month ago, really just wanted to get some feedback on my initial testing strategy and it certainly proved worthwhile with some useful discussions with @Caurmen & @Vortex.

That campaign was a little all over the show as I made changes to my campaign earlier than I should've, but hey it was bloody useful to learn from and I haven't given up hope on that one yet. But I wanted to start fresh with a follow along on a new campaign and document what I do as I go so I can get some feedback to ensure I'm not going about things the wrong way!

So... Here we go:

Geo: Iraq
Offer: Win iPhone 6s pinsubmit
Traffic Type: Pops
Landing pages: 7x ripped landers.

Note: I will not touch anything over the first few days of launching!

Day 1:

Spend: $3.13
Revenue: $0
Profit: -$3.13
ROI: -100%

Day 2:

Spend: $5.24
Revenue: $0
Profit: -$5.24
ROI: -100%

Day 3:

Spend: $5.85
Revenue: $5.10
Profit: -$0.75
ROI: -12%

Day 4:

Spend: $6.55
Revenue: $8.50
Profit: $1.95
ROI: 30%

Actions:

1. Finally some good data and some green numbers! Only $1.95 worth of earnings on Day 4 but a huge motivation just to see some profit. Most importantly I can actually start to narrow down some of the landers thanks to some conversions.

I slimmed down from 7 to 2 landers with 90% certainty based on 3,000 visits each and a total of 9 conversions. They were also the only two with positive ROI. So happy days...

2. I started a high bid test campaign to see how that would fare. Whether it can speed up the process and I can gain a bit more of an understanding of the bidding process.

Day 5 (01/12/15} - Original Bid Test Campaign:

Spend: $4.76
Revenue: $1.70
Profit: -$3.06
ROI: -64%

Day 5 - Higher Bid Test Campaign:

Spend: $12.90
Revenue: $5.07
Profit: -$7.83
ROI: -60%


Quick summary: Glad the higher bid test campaign is still converting and obviously with more traffic coming through so a bit more of a chance to move forward quicker. Hope to post some more stats going forward which will help with more constructive feedback from everyone.


Actions going forward:

+ Keep the traffic flowing..
+ Start to split test different angles using the winning landers.
+ Add another lander (proven from an AM friend) into the mix.
+ Once I'm back down to 1/2 landers - split test a similar offer from another network (just found out they have something similar...)

Once consistently profitable with some winning landers:

+ Expand to other sources - perhaps Popcash (haven't tried there yet but from spying seems it could work- any suggestions?) and/or Zeropark.
+ Look for similar offers in other Arabic geo's and test my angles/landers there.
+ Tighten up my 'funnel' - perhaps try and convert some of this backbutton/exit traffic with other offers or YTZ see how it affects my ROI as it will affect the lander CTR's with their backbutton scripts.

QUESTIONS:

I know there's no easy answer to this apart from testing..But any suggestions for other traffic sources I should be testing with this offer in the coming days? And should I be doing this now or wait to get this consistently profitable even if it is only going to be $XX a day?

---

Mad props for realizing that so early on in your career! You're doing so many things right - I'm seriously feeling like applauding you if it wasn't 3:33am in Bangkok and there are people already asleep in the unit!

Iraq is less competitive than a lot of other geos. Although there's not much traffic there, it's a great one to sharpen your skills with. 7 ripped landers is a great number for a first round of testing assuming they're all very different. Sitting on your hands until reaching stat sig is commendable.

Testing different bids is also a great thing to do, as you've proven to yourself with your finding that you can do similar ROI but collect data that much faster at the higher bid. This kind of testing is what will allow you to understand a traffic source better than your competition.


All sounds great!


Since you already have a green campaign - it wouldn't be a bad idea to start scaling even now. Keep in mind though that your best lander for one traffic source may not necessarily be the best-performing lander on another traffic source, so I'd suggest testing your best few landers on new traffic sources.

Thanks for starting one of the most refreshing follow-alongs I've come across! And congratulations on the green! Here's to much, much more!


Amy

---

Thanks Vortex for the kind words, some useful info as always!

Ok here's some latest stats

Day 6 (2/12/15) - Original Bid Test Campaign:

Spend: $4.27
Revenue: $6.76
Profit: $2.49
ROI: 58%

High Bid Campaign:

Spend: $18.05
Revenue: $11.84
Profit: -$6.21
ROI: -34%

Summary:

The morning performed well, with some more conversion coming through. I noticed on the High bid campaign that Lander A was outperforming Lander B by quite a bit - not enough to cut Lander B and I considered adjusting the 'weight' so Lander A received more traffic - I decided against it so as not to skew results halfway through the day and wait for complete statistical significance. Would you have done this or have I gone about this the right way?

I also launched a 'super high bid' campaign for a laugh to see what would happen and try and get as much traffic as possible. Stupidly once it was approved it ran automatically and ran out very fast - but it was in the very early hours of the morning in the target geo so no conversions and lost $10 just like that - but will give this another $10 spend at the best converting time to date.

Day 7 (3/12/15):

Super high bid:
Spend: $19.52
Revenue: $10.14
Profit: -$9.37
ROI: -48%

High bid:
Spend: $18.60
Revenue: $5.07
Profit: -$13.53
ROI: -72%


Avg bid:
Spend: $4.39
Revenue: $10.14
Profit: $5.76
ROI: 132%


Turns out that WAS the right decision to not touch the landers just yet. Lander B managed a few more conversions and still cannot cut one or the other. However as Lander A is performing slightly better than the other, I decided to make out 10% of Lander B's weight in the campaign and offered that to new Lander C, which is Lander B with the Lander A 'angle' and made some minor adjustments.

I'm considering stopping the 'High Bid' campaign and just keeping the profitable 'Average Bid' and 'Super High Bid'.

Like Amy (Vortex) said, it may be worth cracking on with taking this to another source. So I gave that a go and basically replicated the tests from scratch again (e.g. with the 7 landers + the new Lander C - so 8 in total). But this time I wanted to separate each test into handset types as certain handsets will most likely convert better with a certain type of lander. So have 1 Android, 1 iPhone, 1 iPad, 1 Android tablet and 1 for all over devices. This will help me be able to control the bidding a little more for each device and see from the get go what type of devices aren't converting.

Here were the results from Traffic Source B:

Note: I pissed around with the bidding on the Android campaign and was testing Caurmen's 'mapping the bidding landscape' theory - which proved relatively helpful in understanding the bidding process a bit more and was able to find a level where I was still receiving a decent amount of traffic for less the cost. But I was looking at testing 500 impressions instead of the recommended 40 pops (I went for 100 in the end), so wasted about $20-30 with little to no conversions so results for that campaign are a little skewed.

Android:
Spend: $55.18
Revenue: $13.60
Profit: -$41.58
ROI: -75%
(on the face of it, doesn't seem promising. But hoping with pausing a few placements that got away on me during the bidding process and narrowing down to the winning landers I can get this into the green).

iPhone:
Spend: $35.00
Revenue: $17.00
Profit: -$41.58
ROI: -75%

Other Three Camps:
Spend: $3.36
Revenue: $5.10
Profit: $1.74
ROI: 52%

Tasks:
With a healthy bit of data now after a day and a half with ZP I feel like I can get somewhere and start cutting landing pages straight away as well as. Keeping an eye on placements too to ensure there is an even spread across the board. So time to dig back into the stats this morning.. As well as adjust some bids for the iPhone and Android.

I've got a whole list of angles for sweeps that I really want to start testing. So I want to start trying a couple of these using the winning lander's design.


Question: What is the best method of split-testing more landers? For now I've just given the new lander a low % of traffic from my average spending campaign and high bid campaign, but have given the best overall lander to the super high bid camp.

If I want to cull a certain device type and I can't do that on my traffic source I guess the only way is to redirect it to another offer right?

I'm tracking with ads bridge at the moment, had to upgrade to the Pro account (the free version is great) as I was receiving over the 100,000 visit threshold. It's now really starting to piss me off as they are tinkering with the layout, tweaking things and stuff just doesn't sometimes work. Updating costs is a complete pain the ass, I've waited an hour and it's still 'waiting'. Is Voluum efficient with updating costs manually?

---

Nice progress!

Wait for a round of lander testing to finish before starting another one with the winner + new lander(s).


Yes assuming you can find an offer that converts better than the current one, for that particular device. This may be more trouble than it's worth unless you're doing volume. Alternatively you could just absorb the loss (assuming you're still in high enough ROI overall).


AFAIK there's no way to accurately track costs. Certain traffic sources, such as popads, allow you to post bids/costs from the traffic source back to Voluum. But the results are rarely accurate. I remember seeing other threads that discussed this topic - do a search and you may find them and hopefully see comments by more experienced people. I personally almost never track costs in voluum.


Grabbing a seat and waiting for the next installment....

Amy

---

Definitely interesting reading Looking forward to seeing how you progress and what works for you

---

Thanks Amy.

The last two remaining landers are pretty even (on traffic source 1) on all three bid campaign varieties. Like almost bang on even for conversions so I'm struggling to find a winner which isn't necessarily a bad thing I guess! So at the moment I've got Lander A and Lander B with 45% share and the new Lander C which is a variation of the two with 10%. Should I change this quickly now?

And would you literally go 50/50 split test with a new lander when down to 1 remaining even though it could flop and punish your ROI?

---

All landers should get an equal number of impressions in initial rounds of testing. However, if you're already in the green, then feel free to assign more weight to the currently-best lander and then distribute the remaining weight percentage amongst new test candidates. Just be aware that by doing so, your testing (and therefore further optimization) will probably be slowed down. So choose your battle wisely.

Regarding the 2 landers that are performing similarly - please refer to caurmen's post on the subject. Specifically this part:


Amy

---

Hi guys

Quick update with how everything is going. Have been a bit slack recently but need to keep this thread alive for some killer advice from the pros and to help out any other newbies like myself.

Here are the stats from 4–13th /12/15:

Traffic Source A

Super high bid:
Spend: $58
Revenue: $78
Profit: $20
ROI: 34%

High bid:
Spend: $137
Revenue: $110
Profit: -$27
ROI: -20%


Avg bid:
Spend: $27
Revenue: $30
Profit: $3
ROI: 11%


The key adjustments I made:

Made the Super high bid campaign my ‘whitelist’ campaign as only a few of the placements were getting the conversions so made sense to narrow down and maximize ROI on this one. I’ve now paused the avg bid campaign, although it’s not doing terribly it’s not worth the effort for so little gain. Using my original high bid campaign to find new placements and test any future lander tests.


Traffic Source B:

Android:
Spend: $262
Revenue: $164
Profit: -$97
ROI: -37%

Comments:

I cut the iPhone and other three campaigns as they weren’t converting well or had miniscule amount of traffic. This one has been so up and down.

I was 50% ROI on the 9th after $20 spend with two landers running. On the 10th I cut to the winning lander after reaching 90% statistical significance and then the camp ROI dropped to -50% by the end of the next day.

I started another split test on the 11th with the same angle but with a different look (whatsapp instead of FB). The original was running in the positives that day and outperformed the test whatsapp lander. Got plenty of data that day, ran it through the A/B split test calc and so on the 12th I removed the failed whatsapp and continued with the original winning lander and has been sitting at -35% ROI over the last couple of days.

I have found a couple of promising placements so have whitelisted these and paused in the test camp. Will see how we go over the next few days. Any advice would be helpful! Currently sitting second in the bidding order which I think has helped me increase ROI, or should I be really trying to push for top spot?



I'm not too sure about what to do next. I’m thinking of focusing my effort on another campaign in France that seems promising. I need to get to XXX a day in the next two weeks otherwise I’m potentially out for the count. Don't really see how I will get there with a campaign in IQ unless I can get this working across a BIG traffic source and significantly increase the ROI – doesn’t fill me with much confidence at the moment. I need to try a few more angles really, but from my tests so far the generic ‘congratulations, you have a chance to win..’ or ‘you’re the 1000 visitor, click here to claim…’ have been performing the best but not very exciting ROI’s yet.

Wondering if it's worth to try this Iraq campaign to mDSP or take this to something like Propellor Ads (Don't have the budget for the $1,000 initial deposit!) and go through the same process as I did with Traffic Source B which I've now got profitable on Android OS. Just not sure it’s worth the effort when there isn’t a huge amount of traffic. Regardless, it’s been a really useful morale-boosting campaign and at least seeing some profit from it.

I’m left tossing up whether this is worth it financially to keep working on the IQ landers then split-testing, when I could shift my efforts toward a tier 2 campaign which was breakeven or close fairly quickly and has much to potential for larger profits.

I have tried an initial test campaign on mDSP but NO conversions when following the Main Course of the cookbook with Go2Mobi. I’ve been told to focus on download/play buttons here, any suggestions for scaling to mDSP for sweepstakes?

With a guy from my mastermind we’re trying some sweep offers there and haven’t had any luck yet. Nothing even mildly promising. Which is frustrating considering we’re using the same landers from pops and using the recommended download/play banners using the recommended exchanges. Any tips anyone??!?


So the final questions:

Any suggestions for scaling to mDSP for sweepstakes? Types of banners, exchanges, approaches etc that are working for you or for others? As pops is interruptive, do we need to be smarter and more enticing with our angles rather than the generic ‘you have a chance to win a free iphone!’ on facebook style landers which seems to be working on pops.

Should I just keep the whitelisted campaigns for this IQ campaign running and call it a day on the testing camps which are losing me money and focus on the new France campaign that has far greater potential to scale and get me to the XXX a day mark over the next couple of weeks?

Can you see any holes in my approach so far? Any mistakes I’ve made in regards to the Traffic Source B split/test and cutting?

Thanks in advance. Welcome any comments

---

I ran some IQ campaigns earlier this year, could not get it over $150 a day in profit with pops/redirects traffic, the volume was not there .. If you want to make big $$$ it's probably not the best place, to learn and split-test and experiment a good choice though.

If you got it around break-even or even profitable it's an amazing opportunity to gain an edge by testing new stuff you can apply across the board later.


I'm pretty sure the win an iphone angle and the fb style lander works well on display too haha Do some spying on display traffic to find out ore

---

Thanks Kepe95. Appreciate your honesty.

To be honest, I'd chew your hand off for $150 a day right now! I will focus majority of my time on this France campaign, however out of interest how did you scale the IQ campaign of yours? Did you just spread it among as many pop traffic sources as you could? Or focused on a few and pushed the CVR up as much as possible? Or both hah?

---

Yes, both. As many sources and optimizing everything as much as possible.

For initial testing and getting profitable I have 2-3 proven sources, and for scaling it's 10+ . I see you experimented with different bids, that's a good thing to do, for me not a way to scale at first though. Something to do later to find the profit sweet spot, for scaling mainly more traffic sources or broadening the targeting.

There are many smart guys here, I'm sure somebody is making much more money in IQ , I was just running a quick test on pops which is not very hard to do

PS. Watch out for IQ they have specific ways text is formatted, right to left, and so on - getting a professional translation can make a big difference there, and making sure everything makes sense. Lots of retarded landing pages there

---

Why not pursue both the France campaign AND continue working on sweeps on pop?

Since you already have experience with sweeps, and also have the entire testing and optimization process down, I would suggest repeating the same process for several other geos.

Simply re-test all the sweeps landers you've collected/created up to this point (translated into the local languages of course). As for how to select new geos to target, here's how I would go about it:

1)Talk to your affiliate managers on various networks to see which geos have the hottest sweeps offers.
2)Go through inventory charts of pop traffic sources that have them (or ask your rep for them) to see which geos have a lot of traffic, but low average bids which suggest low competition in those geos. Pay particular attention to Asian and LATAM geos.

You now have a go-to traffic source, know the vertical, know the process - so just rinse and repeat. IMHO, doing so would take a lot less work than to try expanding into other traffic types, such as mobile display.

************************************************** **

However, if you want to try replicating your current IQ sweeps camp on display sources, this is the best method I know of that will make display traffic work like pop traffic as much as possible:

1)Go to a spy tool where you can spy on banners (adplexity is recommended). Browse through the intelligence data for all "generic" type banners to see where they are run - mDSPs, exchanges, placements... (By "generic banners" I mean the "download now" type banners, and also the facebook type ones with the flashing envelopes or presents.)
2)Run the same/similar generic banners at the same places.

So this is another method you could use to scale that IQ offer (and any successful pop camps in the future).

************************************************** ***

Don't be afraid to test lots of similar offers for the same geo! You have a solid foundation already. The rest is just testing lots of things to uncover more promising offers and angles. Most of what you test will not work, but the ones that will, will make everything worth your while.

Success is imminent - excitedly awaiting your next update...

Amy

---

Any update here? Did you scale the IQ campaign to display? Did you get to the xxx per day yet? I assume you gave it a try for the France campaign, where there's more room to scale, but probably it's also a sweeps offer, right? It's best to stick to a vertical and master it, especially since you already diversified with multiple traffic sources and most guys recommend sticking to one source and one vertical first

---

The IQ offer has unfortunately been paused, as did the FR offer maybe a week after getting profitable with it. GRRR. Very frustrating, but I'm told that they both should be up and running at some point in the new year. Yes, strictly sticking to sweepstakes to remain focused on one vertical.

I didn't manage to get the IQ offer to XXX a day, but did get it profitable across a few traffic sources. Bottom line was there was little traffic available on the sources I tested. That's not to say it isn't possible, Propeller Ads was next on the agenda. Will update once I have more news going forward!

---

I think it's probably harder with the countries that have smaller populations and smaller payouts. Good for learning and testing, but probably harder long term, unless you have access to a ton of traffic somehow.

---

Home > Mobile > Follow-along Campaigns