Mobile Sweeps - Win iPhone6/Zeropark US Redirect/AWS/Voluum (15)
Mobile Sweeps - Win iPhone6/Zeropark US Redirect/AWS/Voluum
Hey guys!
Before we start I just want to say thanks to all you big contributors on the site. You guys have helped me a ton, and I'd love to buy you all a beer at some point 
.
Also, other than hopefully making a profit with this follow along, my goal is that it will help someone else.
Ok here we go:
Offer:
Split testing 1 US offer from 2 networks (convert2media/clickdealer). It's an email submit, enter to win iPhone6.
Payout is $1.40 bumped to $1.60 (i have a really cool affiliate manager).
Note: You should never run an offer without trying to get a bump. It happens more than it doesn't in my experience.
Landing Page:
To get a landing page for this I used mobileadscout (Thanks Evan Christopher for making this!). I can vouch for it with redirect/pop spying.
It blows WRW out of the water when it comes to redirect/pop.
I dug around for a few minutes and found a decent lander with a clean apple looking design. I ripped it and did the following optimizations:
- back button redirect to offer
- android or apple noise depending on device. (Note: autoplaying sound doesn't work on all devices... is there a way to get sound to work across the board?)
- ip geo detection: when on wifi I use the location of where the user (using telize's GET REST call via ajax) and use that to call out
- image optimization
- gzip
- minimized all JS
- added vibrate
Screenshot:
Here is a screenshot of the lander (or the top part at least):
Hosting: Amazon AWS Cloudfront
Using AWS s3 & cloudfront to host everything. Very fast, but not really super easy to manage deployment and get everything launched. Does anyone else feel the same way about current CDNs? They seem slightly over complicated for simple things like massive & fast deployments and quick edit. I am using a script I wrote to deploy using the aws-cli which helps out quite a bit, but I'm curious on your thoughts here.
Tracking:
Note: Zeropark automatically imports conversions from
Also for anyone unfamiliar with conversion tracking and wants a quick step by step: (these steps are only for an offer network using the cake platform)
1. get your postback url from the settings tab in Voluum.
2. Paste it into your the offer's network's cake backend (for the offer you're running) and add this to the end of the postback url: "?cid=#s2#&payout=#price#&txid=#tid#"
3. When adding your traffic source to Voluum make sure you check 'automatically append click ids' option
4. When adding an offer to Voluum make sure to append this to your offer url (only if its not already there): "&s2="
I realize there are other ways to do this (like manually appending click ids), but hopefully this will be useful to someone. There are more things to do here, but the above directions will at least get you tracking conversions in Voluum. I need to track site ID's back into Voluum from Zeropark still and will provide updates when I do that.
Traffic Source: Zeropark redirect US
I chose Zeropark because I've heard they have good quality traffic but I don't have much experience with them yet. The first (and only other) campaign I ran with them I bid $3 instead of $.003... Lol. It blew through $83 in about 3 seconds flat. I asked for a refund and they said they'd take a look but now won't respond. I blame my own damn self on that one though, lesson learned!
Campaigns:
First I am creating a test campaign to collect profitable targets. I'm making it mobile RON redirect traffic for the first campaign and using redirects.
Set initial bid to $0.018 - $0.02.
Setting max daily budget to $40.00.
Note: Targeting android only because I can't get iphone's to vibrate or make noise. Apparently, it's just not supported?
My plan is to make a profit camp with the profitable targets I get from this initial campaign (if we have any).
Final thoughts before starting:
Just a reminder... I'm totally new to all of this and would sincerely love any advice you guys have for me. Please don't hold back!
Verrrry nice! Great job with the lander optimizations, load time optimizations, CDN setup...and thanks for providing instructions on the zeropark-voluum setup!
Good tip on asking for a bump before you start! This is not always possible and will depend on how your relationship is with your AM, and also your past performance - but it never hurts to ask!
The only advice I have at this point is to rip more landers than just one and rotate them in ASAP. Start with the ones you see everywhere - chances are at least some of them will work (otherwise there wouldn't be more and more people using them). And of course mod them up with your kickass optimizations!
And US is a very competitive geo! Will take money to crack this geo for sure. If you have your heart set on US sweeps, do sign up to more affiliate networks and test more offers. C2M and Clickdealer are great, but also try Adsimilis and Mundo and Matomy.
Fantastic follow-along so far! Looking forward to seeing some stats soon!
Amy
i would try some more landers. Also there are higher paying sweeps offers available on other networks.
Following this to see results you have. Also running iPhone sweep on redirects in U.S.
Hey Vortex! Thanks! I really appreciate your advice, and all the other tips and motivation you give this forum.
I will start ripping landers! And it's awesome to hear placement stats are brought into
Also, I've just applied to Adsimilis and Mundo and Matomy, so that should be an awesome chance to keep split testing offers down the road. Bummer it takes a few days to hear back, but it'll be worth it.
Note: I probably should have said in my first post... I actually started running traffic for this follow along yesterday, so I'm nearing the end of the second day. So... I'm ready to share stats!
7/17/15 - Yesterday: First Day Recap:
My goal was to split test these 2 different offers and within the first few hours it was obvious Clickdealer wasn't converting (0 cvrs), and convert2media (3 cvrs) was. Being a n00b, I stopped the split test right there and since have been pointing everything at convert2media. I think I should have collected more data before stopping the split test, but I just had a bad feeling about the offer (didn't have a lot of data: ~3x offer payout in spend with no conversions).
After pointing everything at c2m's offer things picked up! Here's how I managed to end day 1:
Conversions: 7
Spend: $35.06
Revenue: $11.20
ROI: -68.05%
Profit: ($23.86)
CTR: 60.12%
CR: 0.43%
Here's a screenshot:
Things that affected my revenue that I can fix:
1. Split testing an offer that wasn't converting. (I will split test more offers, but I need to get approved first!)
2. I ran an iPhone targeted campaign for $6.80 spend before I realized it was a horrible idea
3. I changed my bid for awhile to be #1 and got a ton of traffic but no conversions. I think the best strategy is a slightly lower than average bid and collect more traffic throughout the day. Zeropark uses an average bid which means if you bid like $0.02, you might pay up to $0.04 for a redirect.
Possibly more things?
Question: Each one of my conversions is from a different target which has variable traffic available to buy. A couple targets only had like 1 available to buy, some had thousands of placements to buy. Do they mean the total number you can buy in 1 day?
7/18/15 - Today: Second Day Recap:
My goal today was to continue the RON campaign from yesterday and keep collecting data. Also, I started a new "targeted" campaign. In this campaign I added each placement that had a conversion from yesterday and today. It's still waiting to be approved tonight. I guess Zeropark doesn't approve campaigns on the weekends (I can't blame them .
Here's how I managed to end day 2:
Conversions: 20 (2 missed leads, numbers below include missed leads)
Spend: $63.60
Revenue: $32.00
ROI: -50%
Profit: ($31.60)
CTR: 62.72%
CR: 0.96%
Here's a screenshot (without the missed leads):
Question: How do you add missed leads back to
*Excited* I hit -50% ROI!!! Based on all of my reading this means I have a shot at making this campaign profitable?
My plan tomorrow is to split test a new lander in my test campaign and continue to add profitable targets to my targeted campaign.
Question: I'm planning on adjusting my bid for each targeted campaign's placement so I'm #1... Does this mean I'm buying all the traffic for that placement?
Thanks again for your tips tricks and motivation!
Trevor
Thanks hlyghst!
Next step: finding these higher paying offers
- Leads and
Voluum : Settings -> Conversion Upload and paste your clickids from the network in there. - Autoplay media: There's no foolproof way to get autoplay media across all mobile devices. Most mobile browsers only allow media to be played as result of/after user action (i.e a click) - so sound on clicking "next" in a quiz/running lucky tickets; an onload popup box (non alert) etc. It sucks but the reasoning behind it is sound. If a user is on mobile data in particular, they don't want shit autoloading and playing everywhere sapping their tiny data allowance... even less from a popunder that they didn't ask to see.
- Targeting: In my experience yeah targeting Apple devices for iPhone offers isn't great...
- CDN: AWS/Cloudfront is a perfectly fine CDN solution, if you want something easier then you want to stay away from origin pull and use a push solution like Rackspace.
- Campaign: As Amy said, get more landers in the split ASAP. You're leaving money on the table putting your eggs into one basket.
- GeoIP: You could consider just passing location to lander from
Voluum instead of adding unnecessary overheads for AJAX calls. If you can get away without jQuery and waiting for another server's response then your load time is already down a fair few milliseconds. Voluum already loaded that data before it redirected, it takes no extra time to just use the data from there.
Hey! Nice follow along.
It looks sweeps for US are pretty popular here, basically i run the same))
I also think, there could be easier method for uploading files + with default settings AWS DISABLES gzip .
- on my own experience, after testing on gtmetrix.com. Have you managed to solve this problem?
I tried a few guides, but still did not finish auto converting to gzip.
Hey oleg2014,
Unfortunately I don't have a great way to gzip things yet. I'll report back here with an answer when I do. I've been using a tool on my mac (gunzip and gzip). It works pretty well but isn't automatic on deployment. I'll plan to write a guide on automatically converting to gzip on deployment using AWS. It's good to know there are other people out there with my problems
Thanks,
Trevor
Good progress thus far! I'm still trying to get -50% ROI.
In terms of using the back button script, I understand that it is used to capture users to the same offer as if they were to click your Call to Action Button/Hyperlink. Do you measure the CTR between using the back button script and the CTA button? If so, how would you setup that process? I'm still a little boggled from it. Here are the links below:
Adsimilis Offer Link: http://adstrx.com/?a=xxxx&c=xxxxx&s1={clickid}&s2={campaign.id}&s3={ var1}
Postback Link for Adsimilis: lhttp://xxxx.voluumtrk2.com/postback?cid=#cid#&campaignid=#campid#&var1=#s3# <---- note: I put the "l" before the http: because the forum wasn't showing the whole link.
Hey simon_89,
One way to get the back button data separated from your actual ctr is to create a different campaign in
Maybe somebody else can chime in with a different solution utilizing the same campaign?
Trevor
Hmm, that was something I did not think of. I was just thinking you could tell the CTR difference between the back button script vs CTA within one campaign based on setting new subid variables. However, even if I made a new campaign to back button script direct link to the offer, it wouldn't work because I still need to tell the difference between the CTA button and back button(because I'm still essentially testing LP performance).
Progressing nicely! Sounds like you have a plan!
It's a good idea to use good placements and dayparting to test as cheaply as possible. However, if you find that you're not getting enough traffic for your lander testing, you may want to either hold off on dayparting or reactivate your RON camp, or both.
Also - to be able to daypart properly you'll need a lot more data, preferrably over a few weeks. You can daypart now based on your limited data for the sake of testing more cheaply, but once you're profitable or close, you may wish to retest all hours to collect more data before cutting anything permanently, so that you don't accidentally limit half your traffic (say) that could be profitable with better creatives.
And once you have a list of offers to test be sure to get on that ASAP - as that's one of the biggest influencers of ROI (duh! )
Amy
Hey Amy, thanks! I really appreciate the feedback
I plan to reactivate my RON camp and start day parting that slowly as I collect more data over the next few weeks. My average ROI is around -50% right now, best day being -30%. I feel like I'm close, just need to break through a few more walls. My targeted camp is doing ok, I have around ~40 targets now that are in it that were profitable from my RON camp. I'm using this camp to test more landers and compare CTR/CR.
Tuesday I tested out a new lander angle with all the bells and whistles and I was seeing a 45% click-loss from clicking the offer link to registering a click on my network. Is this normal?
After noticing this click loss I compared the reports for the past few days and they all seem to share this 45% click-loss. Maybe this is normal? I really need to test new offers ASAP. I'll have updates shortly on that (expect updates Friday night, Saturday, and Sunday)!!! Thank you all for your support on this follow along.
Trevor
Amy
Home > Mobile > Follow-along Campaigns