Sweeps SOI on Redirect traffic (17)
Offer: Win Device USA, SOI, ClickDealer
Payout: 2$
Traffic Source: ZP
Traffic types: Pops, Redirects
OS: iOS, Android
Result: -130$
Leads: 37
I took most popular FB winning wall lander as base. Also i made my own lander "casino" with some interaction but it made 0 conversions. I spied for landers and took 2 another landers.
Most conversions came to PS4 Bundle, as its relatively new offer.
Traffic
I decided to use ZP as - quality traffic and competitive.
After looking at zeropark.com/volume i decided to use pops (mistake 1), becase they are really cheap ~ 0.006$. I targeted Android only (mistake 2).
After initial tests i saw 40% of tablet traffic, so i turned off pubs with 50%+ tablets. Also i saw that in the evening offer converts better for me, but i asked AM to see overall % of convertion by hour and its relatively the same except a few hours in the morning.
After i spend about 100$ in pops i got about 12 leads 
I decided to try redirects. Thank you guys for follow alongs for the same offer but in another geo - i saw, that this offer can be at least non failing with bids ~0.04$ for MY. So i put bid $0.016 for initial tests. In evening i was almost profitable but i had really small amount of converting traffic:
most of my conversions came from pubs with 1-2 visits, in the same time i had about 10 pubs who brought 500+ visitors with 0 conversions. After disabling such pubs, and bidding high (0.03) for converting pubs i got no more traffic from there in next few days.
Also tested this offer in AdCash, after spending 20$ and 0 conversions, stopped traffic.
I d like to test it in PopAds, but payments from my country does not allow me to target 1-tier countries like US, Australia.
Landers
All optimization was made on redirect traffic only
After spending 3-4 conversions on each lander i decided to disable "casino" lander and a few more quizes. They made no conversions, "base" lander was doing +-0$ ROI.
So i made about 6 variations of this lander and spent 2-3 conversion payouts on each of them. I tested vibration, sound, alerts, cta, different comments.
I got really bad result - my very first lander was much better then anything else.
One more thing
After my mistake i turned on iOS traffic. And gues what? It converts at least not worse then Android, even tablets. But still not profitable for me. My latest results were -40$ ROI, so i decided to test something else.
I have about $1k for testing now, so i decided to choose more 5 offers with payout <2$ for tests.
Lessons Learned
Excel spreadshets with details for every lander and spendings really helps to see overall pictures.
P.S. Still looking for masterming group on POP traffic, please pm me if you have 1 slot for noob 
Very nice start to a follow-along!
One suggestion: Consider changing the topic of your thread from "FAILED" to "APPROACHING SUCCESS". Please tell me if you agree that this would be a good change. If you do, I could change it for you. This may sound like a little thing, but trust me, it will change your expectations, which will influence the result!
I think you have accomplished a LOT - by just testing a few landing pages you've achieved +-$0 ROI! Did I read that correctly? Are you breaking even with your best lander?
You can certainly test other offers. If you do, use only your best lander to do this. I would HIGHLY suggest that you set up a new campaign to test offers, and keep using your first campaign to test landers. This way things are much less confusing, and you will have enough data to make decisions much sooner.
I would also suggest that you do more spying, and rip several other landing pages that look different from the landers you have now. 4 landers are not enough!
Also try different themes. The fb theme is everywhere and people are getting tired of seeing them. Try themes that look like system messages for example.
Great job on testing different traffic settings! Cut zeropark placements that are draining your budget and not converting - this will make your testing cheaper. (You'll probably want to keep placements that are losing money but still making conversions, unless the loss is big - because at this point you want to receive enough traffic to test your landers. ROI at this point isn't the main consideration. Getting enough traffic to optimize your creatives IS.)
Since you're targeting mobile, try adding a backbutton script - a lot of visitors tap the back button so you want to capture them! Send them to the offer page for now. (You can also split-test that later by sending them to another offer or another lander.) This can make a big difference to your ROI!
And have you tested your page load time yet? If you haven't, be sure to do so! Read some posts on how to decrease page load times. Most importantly, USE A CDN - especially if your webhost is far from the geos you're targetting. Load speeds can make a huge difference to ROI as well!
Test more landers and more offers - your ROI WILL improve!
Will be waiting for your next update...
Amy
Thank you for your feedback, vortex. Some time ago i saw your posts and they also were very usefull.
I decided to continue this campaign.
Here some stats for yesterday (its 7 hour difference between my location & USA).
Date: July 4
Spent: 27$
Earned: 14$
ROI: -47%
So i was wrong about +-0% with best lander, its -30%. Tests were made for evening only, as i see better convertion in evening.
I disabled one of landers after 70 clicks because of 0% ctr. Today i checked it and saw there was broken final link Will fix it and try more.
Also yesterday i tested similar offer for Poland ~0.5$ payout. WTF poland mobile iOS traffic costs more than $0.01? Really expensive for such payout.
Spying
Used vpn + smartphone/desktop and my automation spying script.
Found only 2 new landers and added them to split testing yesterday.
Found one good lander in another geo but could not get it
After opening the same link in phone in second time, i was redirected to its_domain.com and got image like: Are you angry bro? 
Also i saw a few popular landers but they had no relative address. I expected to see domain.com/us/sw/land/1234.html?blablabla but iubt i saw domain.com. As you understand, this lander is possible to save only from desktop, because you can't see it after entering such domain in browser. Its something new. I think its done from server side, will try the same, but i need own landers first.
Saw a lot of old landers with a lot of unused code, dublicates, errors and so on. And they are really missleading - You Won! You Will be Invited and so on.
Landers Optimization
I worked as software developer for some time, so it is the easiest part for me.
1. To detect bottlenecks i use gtmetrix.com
2. I use my own version of jquery ~50kb insteed of ~90kb in almost all landers. And try to avoid it when possible.
3. I use tinypng.com to minimize pictures
4. All JS/CSS is in html file, minimized. All js event handlers are firing after page load.
6. VPS in Central Europe (only 1 location for now): optimized nginx configuration + images are located on another domain with enabled cloudflare.
Just optimized 1 lander 700kb -> 200kb is it enough for USA internet & Europe VPS?
Currently this lander does 18 requests also. Is it too much even with CDN?
Looks like my USA campaign was not such bad, as i thought.
Today i tested different SOI offers in DK - Apple products, vouchers.
Got translated FB win wall lander, quiz, + 1 more custom design.
All landers with vibro, alerts, sound, back&close scripts, no conversions
Spent: 20$
Got: 0$;
Traffic sources: PopAds (with STM blacklist), ZeroPark redirects (less traffic than PopAds, need at least 500 visits to make decision regarding to it).
My bid in ZP for DK >0.02, got about 170 clicks for a few hours.
Will continue USA campaign and try another GEO. Also have one more idea for lander, need to implement it.
Impressive lander load time optimizations!
What's the load time down to now?
http://tools.pingdom.com/fpt/ (don't forget to click on "settings" and set it to the closest geo location)
http://www.webpagetest.org/ (don't forget to set "Test Location")
Amy
P.S. Title changed as advised in your PM. You're well on your way to success - don't need that big "FAIL" in the title!
Here are speed tests for my best lander:
First of all, latest post about Sweeps by mr green gave me some motivation and mind-set (Thank You!). Its not very clever to stole some landers and hope to get good money with them without any changes and creativity. So i decided to create own lander. Using google thrends, i saw Game of Thrones is much more popular than any other tv show in USA, so i found theme. I took questions from fan sites.
TESTING
Theory 1
Everyone should know this tv show so it can work from redirect/pop traffic.
Its 3rd variation with biggest CTR ~ 3% on redirect traffic ~ 1000 visits
Failed
Theory 2
OK, lets try some banners
BuzzCity
Spent 18$
LP CTR 3-4%
Go2Mobi
Spent 12$
LP CTR 2-3%
Stopped because with cpc 0.1$ and ctr LP 2-3% i need at least 2 mirracles to have at least 0% ROI
AirPush
Spent 40$
-20$: At the beginning i did not choose anything in geo targeting, so i got a lot of clicks from Marocco, Nigeria, Iraq with CPC 0.03$ 
More 20$ were spent at least ot USA, but i got 1% LP CTR, stopped campaign.
After all this, i got a bit stressed, because i thought angle is really good, and at least banners will help to get break even.
0 conversions from angle - really bad sign, i think
I was thinking to stop this campaign, but remembered about ZP campaign with old lander (first pic in post), so i started it in best converting hours and got at least some conversions:
I disabled some more pubs and hope to get more ROI today
Really, i like results from ZP, but when i look on converting pubs, i can't see what should i optimize besides disabling all big pubs:
Almost all conversions come from pubs with 1-2 avaliable visitors.
Plans:
1. Test another SOI offer directly on redirect traffic
2. Make & test new angles with banners
3. Spend 50$/day
My familiar superaffiliate in adult told, at the beginning he tested 10 landers / 200 banners per day.
So i know i am VERY SLOW and LAZY and i should do something with it.
P.S. At start of this thread i did not know how to paste full sized images, now i do not know how to resize them
Tried [RESIZE=100][/RESIZE] and [IMG WIDTH=x HEIGHT=x] - does not work. If you know how to do it correctly, plz send PM.
Hey, nice start to a follow along,
I'm a newbie as well, but i might as well start helping out as others are helping me
Good job on creating your own landers!
I might start the questions off a little easier - try not to make the person think too hard on a specific episode on a specific scene.
For your banners, your grammar is a little off. This would probably be better - "Do you know who he is?" - "are you really a fan of" - " are you a true fan of"
keep it up! can't wait to see your progress.
Major action and progress! Major respect!
A load time of 0.5s - for a sweeps lander like that with multiple questions and all the graphics that's pretty darn good!
Several suggestions on your lander and banners:
-Correct the grammar a bit: "Are you a real fan of the Game of Thrones?" "Do you know who he is?" (Try the shorter version "Know who he is?" or "Recognize this guy?" or "Do you recognize him?")
-In fact, I would suggest replacing that lander headline with "Love the Game of Thrones?" (I've tested similar headlines and found that to work better than "Are you a real xyz fan?")
-Instead of saying "Prove it and get a prize", try something like "Answer 3 questions correctly for a chance to win a playstation 4!" and make the font bigger and put it right beneath the headline where it will get seen first - because it's important (may also be good to put a similar message on the entry alert).
-Make the questions shorter - people don't have the patience to read long questions.
-4 questions may be too many. In my own testing I've found that 3 questions work the best - but you can test this.
-4 answers may be too many. Test 2-3 answers. Having fewer answers will also allow you to keep all the content above the fold after you add that "Answer 3 questions correctly for a chance to win a playstation 4!" line.
-In case you don't already have this: when the visitor clicks on "CONTINUE" have an exit alert pop up to tell them what they need to do on the next page (e.g. "Enter your information on the next page to reserve your prize.")
See if that doesn't increase your CR.
Amy
Originally Posted by oleg2014
Thank you all for advices, especially vortex.
You really help to keep on!
I have already made changes to LP and banners, removed grammar mistakes, simplified LP.
Also i enabled back button & close page scripts for traffic sources that allow it.
Here what i got:
Results from Buzzcity:
Spent: 3$
Clicks: 160
LP CTR: 1.8%
Conversions: 0
Go2Mobi:
CPM: AutoBidder
Clicks: 120
CTR: 1%
Spent: $12.81
LP CTR: 0%
I stopped campaigns when saw the LP CTR. Its something weird, because in redirect traffic after 100 clicks i got 4% ctr!
Should i spend more to find placement with at least some LP CTR?
Originally Posted by oleg2014
I thought you were making some conversions on zeropark? If so it may be better to focus on using zeropark to optimize your lander first, and only after you have a great-converting lander, move to display and mass-test banners. Right now you're working with unoptimized banners and landers, which can really drain your budget. If you have the money to spend and want to go straight for the killing please continue - but in that case you'd have to test a LOT of landers and banners.
Also, US is an extremely competitive geo with expensive traffic. I don't know which language you originally speak, but perhaps you can try some offers from a target geo that speaks that language? Or use Fiverr to translate your lander. If you don't have much text, onehourtranslation.com won't be too expensive either, and they're fast.
Whichever geo you may pick though, make sure you run your camp at a traffic source that has plenty of traffic for that geo! That should be obvious but it took me a long time to realize the importance of that. You won't want a little trickle of traffic - that would take you months to test landers, and then even if you can get profitable, the camp may only make you $x-xx profits a day in exchange for your valuable time having to babysit them (check stats, tweak bids, cut placements etc. etc.)
If I were you I would split-test the new lander version with the old one on zeropark - just rotate it into the current camp - and see if it's even close to breaking even. If not, then you'll need to test completely different lander styles/angles (as was mentioned before). It looks like you're really good at coding - so use that to your advantage!
Hang in there! There's a steep learning curve for sure, but you're doing great and taking massive action - if you keep testing enough things something will work, and your faith in this statement will keep you going until it does!
Amy
Everyone should know this tv show so it can work from redirect/pop traffic.
Sorry for silence these days.
SWEEPS
First of all, after conversation with AM from PopAds, after sending some recommendations & showing these thread, i got approved to US targetting (they have some rules, that advertisers from not all geos can target tier 1 countries). It took about a week, one more signal for me, how important is persistance.
Also now, as you now we can blacklist targets in AdCash selve-serve, so it makes sense for me to try it.
Also yesterday i found 2 more new LP for sweeps, which i will test.
Regarding to new GEO, i already tried PL (payot 0.5$), now i see such offer in EU only. I can't find any SOI sweeps in tier 3 countries, except IN with payou 0.03$
CPIs
For this time i tested 2 CPI offers in 3 tier country, in go2mobi, mobfox. I saw, that not VERY ugly banners converts, made about 6 banners for each offer, only 2-3 of them gave some convertions. Had some troubles with new policy from MoPub:
- must be branded
- must have a clear call to action
- cannot use the "download arrow" button/imagery
- cannot claim to "fix" or "upgrade" your phone
anyway, go2mobi support is available to chat, it helps to understand what kind of banners they need.
both offers resulted in -80% ROI with direct linking after spending x30-40 payouts, stopped.
Also now i am in mastermind - EMMA, starter of it made really good job, regarding to branding it
looking forward to hear more from you. Keep going buddy! I've learn a lot here. Thanks!
Network: PopAds
Landers: 2 newest, i found, the same i saw the same in another FL
I changed these landers with own ideas.
19 July
Server: DE + Cloudflare for tracker + Cloudflare for landers
Spent: 58$
Earned: 36$
ROI: -38%
Clickloss: 10%
20 July
Server: USA + Cloudlfare for landers
Spent: 29$
Earned: 14.3$
ROI: -54%
Clickloss: 10%
Also i asked my hoster to give me USA server for a few days, for test. So results from 20 July were from another server location, without Cloudflare before tracker (as PopAds asks when you entering url behind Cloudflare). So, comparing 2 servers, i do not see changes in clickloss, landers became more fast, but it's not critical, as they already were quite fast.
Thinking more and more on testing a bunch of CPIs for the same 100$..
Lessons learned
day/hours parting has great influence on ROI, because i saw profitable hours while testing, comparing with another with -70%
i would not suggest it to everyone, as i collected small amount of data, but for me EU server location + cloudflare is OK for USA.
Looking good - those ROIs are definitely promising.
Thanks for reporting your observations and tips oleg! You're making really nice progress - well on your way to green pastures!
Amy
Home > Mobile > Follow-along Campaigns