newbie CPI app installs (25)
hi,
previously i did mobogenie and quikr IN, so i thought instead of doing a new thread for every campaign i might as well do a consolidated one so that i can always post here whenever i change a campaign.
right now im working on an antivirus targeted at brazil, heres what i done so far
1. i researched using a vpn to find out exactly what ads are being shown and what kind of antivirus ads are being shown, their landing pages etc and basically modelled after them.
2. i launched a campaign and its still running, apparently the better performing one is a direct CTA >>> remove virus now
3. the same campaign shows that the ones that perform better are the ones that directly creates fear and looks somewhat professional like a real error (7) virus on phone? etc
obeservations
1. i made the mistake where the landing page did not match the banner. the banner says 7 virus but the landing page says 27 virus.
2. the direct linking did not convert as well as the landers
3. the lander that was less "aggressive" showing the 2 steps to remove virus did not convert, maybe its too lengthy
4. the lander that converted was very simple, >>> remove now, im guessing the user wants an instant solution???
5. the landers that converted were congruent, they were talking about the same issue in the same light. those that switched focus did not convert, for example the one that converted the banner said 7 virus, and the lander that converted said, remove this 7 virus. the other lander that said , protect yourself here are 2 steps, apparently did not convert
personal notes
ok i got demotivated these few days because of the reptitive tasks like compliance with ad networks and it makes me sick. however i recognized that this is part of the job and is about to get worse and worse. i looked back at my past experience and found that what really motivated me was working for the people around me, so i created a wallpaper that shows a 3 month goal of what i want to achieve for my parents and gf, lets see how that works out.
budget wise... not so good, but im working on it
traffic sources, im not sure where to run my traffic really, everyone is recommending a different traffic source, im confused, could someone link me up to a successful antivirus campaign?
so far the people i know did not run their traffic on decisive, so im guessing perhaps decisive might be a problem and im going to try another traffic source maybe leadbolt? or adult traffic
you'll get the best feedback for your follow-alongs if you post images of your ads,landers,and/or stats from your ad/aff network, and organize your questions very succinctly. Since your not profitable yet, you dont have anything to lose, and everything to gain.
From the sounds of it though, your on the right track, I would just stick with a single traffic source that you know works and hammer at it until your seeing green. good luck
here are the banners
banner 1
banner 2
banner 3
banner 4
banner 5
banner 6
banner 7
banner 8
banner 9
banner 10
banner 5 is rejected due to misleading content, but i ripped it off some sites so i think its possible to find a traffic source that allows this?
banner 1 is rejected for same reasons as banner 5
banner 6 got rejected in one ad but accepted in another... both on decisive... errr... maybe they missed it or something.
and here are the landing pages
lander1
lander2
i realised lander2 was associated with this banner which i forgot to upload
so obviously it did not convert. and it couldnt help the other banners too
what do u mean by sticking with one traffic source, u mean stick with decisive which is the one im using? or stick with using mobile as a traffic source?
what other stats should i be looking at?
your data is statistically insignificant, collect more data
Yep, definitely - spend $15 per campaign, then come back and analyse the data. Currently you don't have enough information to make the decision - see Part 8 of the Getting Started Guide for more on statistical significance.
Also, keep an eye on your placements and kill anything that's spending a significant fraction of your budget with no conversions.
Not to be a wet blanket, but I think the whole AV "Zomg you have 7 viruses >> remove now" approach is dying out quickly and I would suggest taking a completely different approach.
It wouldn't surprise me if a month from now almost no mobile sources would allow those types of banners.
Of course there are plenty of ways to make similar, less misleading banners.
From the very beginning I have looked at those ads and just thought "wow, what scammy marketing". I pity people who can only make money in AM by running misleading ads that don't help improve their marketing skill set.
Originally Posted by zeno
Thanks for the slap!
Don't think the method of throwing shit and see which one sticks suits me.
Time to rethink and revisit my notes again.
some things i learnt
1. its hard to work alone, so i asked for help from some of my friends and got myself a partner.
2. apparently the better approach to antivirus is on battery draining? its performing better, i know people who ran it and got better results without a lander. so il lwork on that
3. there are many other cheaper traffic sources other than decisive, so ill explore that too, but im not sure if now is the right time??
4. i developed further in the you have a virus angle and made the lander look even more like a error and right now it seems to have higher conversions and better CTR too
5. watched zeno's video and downsized my landers from 80kb to 10-15kb. amazing!
Looks like a good follow along.
The banners you posted look pretty good, just need to buy more data.
Keep in mind on decisive.is you won't be able to profit running RON so you need to gather data and cut out the non-performers.
I would keep the banners, maybe add 1 more lander to the mix and spend a bit more before making any decisions.
hello bbrock32, firstly thank you for your response
can i find out what do you mean by wont be able to profit running RON? im guesssing RON means the entire traffic. how much data do i need to cut out a non performer? and can i seek a better understanding of why the site is not performing? or maybe why other sites are performing? maybe the ads are shown in a way that is not well received? like another tab which the user closes? or maybe 20-30 ads grouped together where the user avoids? below are some statistics, please guide me along on what to cut, maybe an explanation why, and what i should focus on?
i have also added a few banners to develop on this angle and landers, which are below
here are the creatives i kept from the first run, which is used in the 2nd run. the angle being used here is >>> you have a virus! remove now. ok like zeno mentioned this is kind of a bullsh** angle and in fact some traffic sources have already rejected this as misleading (obviously it is). and from my mastermind group i noticed the battery draining angle seems to be doing better, but im really very very curious here.
how much worse is the "you have a virus angle"?? i mean i know its crap but we can see from this thread by bbrock >> http://stmforum.com/forum/showthread...azy-New-Angles
that virus concerns are number 2. we can see that many people are using this angle, possibly reducing its effectiveness. so how about the 3rd angle on slower performance (which is also widely used) or say the rank 5-6 like wifi not connecting? they are not often tapped, so im thinking lesser banner blindness? so maybe it could perform better than "you have a virus???" i'm just really curious how much worse this "you have a virus" angle will perform due to this overuse by people and the BS tactics
and here are the newly added banners my partner helped to make
again they are an expansion of the "you have a virus" angle.
i also tried to develop the landing page, this is the original
the text in english is as follows >>>
Your Web Browser May have (7) Virus !
If it does not resolve in a few days,
the Virus may leak your personal contacts or other important information
Please DOWNLOAD and INSTALL Mobile Security Antivirus App!
Then Open the App and Remove ALL Viruses
REMOVE VIRUS NOW
---------------
research into this area suggested that if it looks more like an error it might perform better, so i went ahead to develop another lander which is the same text but this time with a box, perhaps looking more like an android error.
-----------
previously the lander was around 70kb, but now its been made to around 10-15kb through zeno's video on optimizing mobile landers, i downsized the images, from around 15kb to 8kb. and the CSS files were deleted, not sure how much space that saved but its probably alot seeing that the file size dropped from 80kb to 15kb according to GTmetrix.
------------
ok and here are the results from decisive
TOTAL SPEND - $73.92
total revenue - $7.36
ROI - -90%
ok i am not really sure what to conclude here. i have been watching this since it started (in brazil time) around 2am, for about 24 hours. the CTR on my lander was pretty high from 2am-7am at around 30-35% on both landers, and i was winning bids at 0.20 CPM at a around 30-40% rate. however as the day came, around 12midnight to 2am, my win rates dropped down to like 10-20% in terms of winning clicks, and the CTR on my landers have reduced to like 10-15%. what is this because of? or am i just reading too much into this? previously when i ran this same angle with different banners the CTR on my landers was also around 25-30%. and now it dropped towards 10-15%?
a1r3 on the mobile site seems to be performing well... but the conversions all came in the day from 12midnight to around 8pm. whereas the conversions for the first 3 banners (run2) came mostly in midnight.
is the mobile site better? does wifi site suck because there is so much impressions with little conversion? or is there insufficent data?
the payout for this is 0.46 per install, and in fact on some affiliate networks they are paying like 30 cents? so i guess 10 bucks per campaign is enough to test? thats already 20x the payout, maybe even 30 if we consider the other affiliate networks.
what is considered promising? what is considered NOT promising? i mean i have some conversions, but is it good enough to continue to develop this angle? or change an angle? or change the offer? i mean this offer on psafe brazil is on a few affilaite network so im guessing this is a proven offer? i mean i dont need to be profitable on this , i want to learn how to make it better so that the next time i run another offer i can apply what i have learnt, so i dont mind this offer being around for some time.
another person in a mastermind group showed his results and the angle on you have a virus seems to be doing almost the same as me, but the angle on battery draining fast somehow seemed to produce better results. i mean we are still looking at a -90% ROI here, is it because my angle is not good enough? or is it that decisive traffic for RON is not good enough and if i cut out crappy sites it can improve maybe to profit? but -90% to +ve seems too optimistic. maybe -50% is possible.
run2 and run3 seems to be the same. the angles are the same, the banenrs are slightly different, i thought that the banners in run3 were poorly looking and may perform poorly... but apparently thats not the case, it performed the same and maybe even slightly better than the ones in run2... so maybe i can try some funny non-professional banners? someone in my mastermind group had some handwriting banners and apparently they seems to work fine. but again this "you have a virus" angle is supposed to look like a real error right? and when we throw in this non-professional looking banners are they not in line with the general "professional" look?
----------------
another person in my mastermind group is running a DIRECT LINK AV offer in argentina on leadbolt, and he WAS getting -30% ROI. angle wise im not sure, i think hes using the battery draining angle, you have a virus , boost memory. a quick look at the EPC on antivirus apps are like 0.04, and leadbolt is having a minimum CPC of 0.05... so i think maybe we need cheaper traffic sources.
----------------------
TRAFFIC COMPARISON
ok is this the best performing one with more conversions, the rest have little to no conversions (ill show them if they are useful, but what can we do? remove all sites because all are not converting? does not make sense)
what sites do i remove? how many sites in terms of % do i remove? if i follow the newbie mobile appetiser guide and do what it says >>> "If there is an App or a Site which has had more than a third of the total spend on your campaign since that campaign began, has not had a conversion, and has spent more than three times your offer payout, blacklist it by clicking the "blacklist" checkbox in Decisive the the right of that App or Site. "
firstly none of them hit 1/3 of the campaign spent, pretty equal, which means nothing should be removed? what is the purpose of this statement? to remove sites that are spamming our offer?
next if i remove the ones that spend 3x of my offer payout, then i end up moving almost every single site. and some of the sites actually spent like 4-5x the payout (because the payout is so low) before getting a conversion. so if i removed them earlier maybe i dont even have the conversion at all?
how many sites am i supposed to blacklist? is the performance like this because my angle is not good enough? maybe the sites are fine but my angle does not work?
or is this part of optimization where i end up removing like 90% of all traffic sources and keeping the best 10%??? but i probably need alot more data on that isnt it?
tubidy.mobi seems to be performing well. will research more on why later. it has higher CTR and also higher conversion.
popup advertising
i was experimenting around with zeropark and found that by popping the lander directly apparently the ROI is better than showing a banner>>> lander >> offer. on decisive i spend around 5 bucks for a conversion, on zeropark thats 2 bucks for the same conversion, and i dont need a banner!!! anyone can suggest why this is so? less work for more results???
here are my conclusions
1. this angle MIGHT work, i mean if it sucks its going to be -99%, so -90% suggests SOME potential, but maybe not enough
2. we should definitely try the battery draining angle, my friend is getting -60% though. ok this is confusing because i listened to the interview by glen on mr. green just 2 months ago, somewhere in june 2014? and mr green was able to turn xxx,xxx on a campaign DIRECT LINK on decisive??? means that the traffic is fine, i probably dont know how to use it, maybe my offer does not cater to their needs? maybe my angles do not answer their concerns? i dont think the fact that everyone is using decisive makes it a sucky network, though i did get better responses on other networks.
3.
You seem to have a million and one questions, many of them highly speculative.
This is a clear cut case of analysis paralysis so I'm just going to tell you some key points:
1. You would kill the datawrkz and mediashakers placements shown in the image. They have spent several times the offer payout with no conversions. The "1/3 of spend" comment is not relevant to individual placement optimisation - it merely means to watch for a single placement taking up the lions share of impressions (i.e. cheap crap quality traffic).
2. Look at your campaign overall.
You have spent ~$80 over many different banners and across multiple placements, all split by WiFi/Carrier and Site/App.
I am unsure of how many angles you have tested and if you have tested DL vs landers and how many landers.
In any case, the math hear doesn't really add up to anything hopeful.
None of your campaigns have shown CPAs even remotely close to the payout.
Ad CTR is not the issue.
My conclusion: this angle/traffic source/offer combo is simply not working right now.
There could be dozens of reasons why, a major one being you're trying to run PSafe/AV in Brazil on Decisive... which really gives you absolutely no uniqueness whatsoever.
3. My Advice
Pick a country which is not so absolutely inundated with AV offers and try advertising there.
Try a completely different offer, better yet rotate a few instead of putting all your eggs in one proverbial basket.
Talk to your AM and find offers that are converting but haven't yet been smashed (at least in country X).
Come up with multiple angles from the beginning and multiple landers to sell those angles.
Make a checklist of optimisation decisions, e.g.
- If spend on placement = payout with no clicks > kill
- If spend on placement = 3x payout and no convesions > kill
- etc
TEST TEST TEST
Probably not what you really wanted to hear but honestly even I wouldn't touch AV (let alone PSafe) in Brazil on Decisive right now.
well actually thank you, i wanted some direct advice and an opinion on what i was doing. why not touch AV? whats the problem? i understand that psafe in brazil is over marketed but is there a problem with AV? or a problem with decisive?
how much should i have spent before cutting the campaign? this is just one angle and the payout is 0.46. i developed the same one angle differently.
i think you got me right on the analysis paralysis, i simply didnt know what to do so i just sat there watching, and confused lol
Yup - really quick conclusions are possible here.
Key takeaway: don't watch your campaigns continuously. It's not an effective use of your time and tends to just stress you out. Check stats periodically. Real-time stats are nice but dangerous - don't let them eat your productive hours.
I'd generally cut any campaign that was performing less than -50% ROI after the first round of testing. You've tested, the angle/offer/country combo doesn't work. Have a look through your stats, think about what you've learned, and try something else.
Originally Posted by xavierfok
If you read around the follow-alongs you will see any number of people running this combo.
Red flag #1!
Copying pretty much exactly what others are doing, especially when you are a newbie, is a great way to get no traction.
The market is already full of competitors and you bring nothing new to the table.
On the other hand, the same offer in Peru might work because there has been very little corresponding advertisement there.
On the other other hand, PSafe may have already been advertised to all of LATAM and have decent market penetration in every country.
In which case a different offer would be better.
In my opinion, so many people have jumped on the AV bandwagon that I would focus on a completely different genre of apps... not because I don't think the offers work but rather because I don't want to bother trying to convert users who have been polarised by a mudslide of annoying "you have a virus" ads.
That being said, the right angle can often make things work in a saturated market.
TIP: Stop using decisive, everyone who read caurmens guide is now using decisive. What does that matter? What does that mean? That means that traffic source is ultra saturated and everyone is running the exact same damn thing.
Get creative (and I don't mean make new landers or banners) - I've posted over 400 traffic sources here on STM - check the list here, and get a VA to contact them all, and ask who has inventory in brazil. Then do a $100 dollar test on each that looks promising, and I guarantee you'll be seeing way more promising numbers than -90% ROI!
I never understood why everyone breaks up their campaign on App / Site & Wifi / 3G combos in Decisive? You can just use the reporting tool to analyses this. Now you can even export all the click data to .CSV and analyse every possible metric combo
Originally Posted by zeno
@caurmen, cut if its performing at under -50%? but on decisive with so much competition could this requirement be relaxed? or is it still possible for newbies?
@zeno - i thought that we were supposed to pick a popular offer? so now you are saying that to get better results we pick an offer thats not too popular? and the country as well? but yet converting?
@iamattila - someone in our mastermind group is already working on gathering information on traffic sources and each one of us will probably try different ones and share data =D
@keepthecar - looks like its time to move to a less competitive country/offer. how do we ensure that the offer is good? ask our AM to see if its converting? but thats like 500 clicks daily? is that enough confidence?
You can also just be the scammer in forefront. Constantly pushing your marketing with new and novel scammery that tickles user to CLICK. HERE. ha.
"Pick a proven offer" is slightly tricky advice, with a hidden additional part: "but don't pick one that's been done to death".
Also, testing lots of offers is just part of the game. Finding an offer that converts for you is hard, it requires testing quite a few combinations, but once you've done it you're 80% of the way there.
There is no way to be 100% certain that an offer converts except running it. Test, record results, repeat until you find something good.
As for the -50% thing - no, do not relax this rule. Campaigns that are converting at less than that are going to be harder - possibly impossible - to optimise.
tried a new offer on an Android App Install with a low payout of US0.20.
here is the process
1. researched on many countries, over 10,000 ads seen in 40+ countries to see what kind of angles offers are being promoted
2. ripped the banners and landers, modified it and tested it
3. picked the top 10% of performing angles and started to expand on
- this part is interesting, i was confused what exactly was an angle, so for 3-4 days i was mainly working on the headers images descriptions until i realised that i was working on the same angle, then i started to zoom out to work on the more value adding activity of optimizing the angle, which is essentially writing the same thing in a different way, now focusing on visualizing what the app actually does for you. as cashvertising says, let the consumer experience the use of the product.
4. work on the things slightly less important than the angle, this angle alone is 25% off the bat (dropped to 15% because i bid higher for traffic, and another angle was 25% and dropped to around -10% not sure why probably because of my overbids for more data)
- so i worked on the images, i tried things people already knew about, the image was suppose to bring up something he already knows to add crediblity, not introduce a totally new thing and try to convince him that is correct, for example everyone knows cinderella, everyone knows about google, and probably the three little pigs etc
5. once the system was smoothed out i had more time to think about stuff, and worked on analyzing angles, based on the statistical significance calculator, however there are many times the calculator will not show significance, i simply picked the best 3-4 angles/banners and worked on them. and now im running more traffic to them to get more data
6. i started working with a single country, and traffic was slow even if i increased the bids. my daily running cost (which is low compared to many other trades) come in at around $10 daily purely for internet marketing, and if u add in the food expenses and the loss of income since im doing this instead of working, im losing US100/day, and i was spending 5-10 bucks and waiting to see data? kind of does not make sense, so i wanted to increase my daily spend to gather more data since now i am better able to process it. after listening to charles ngo interview, he was talking about scaling the same campaign on the same ad network FIRST before moving to another ad network. i was not scaling, im in the process of optimizing and i wanted to collect data faster, so i went ahead to bid for more countries. i was thinking what if the angle works on one country but not another? then that angle is not scalable and not worth much anyway, we want to find angles that are scalable and can be used on many other countries and ad networks.
being my usual self, i am now trying some countries i have been told not to like united kingdom just to see what happens, and pakistan just to see how it works out.
7. everything above is currently being systemized. it took me around 2-3 weeks to do everything, but if i were to do it all over again i think it can be done in maybe 3-5 days?
questions
1. is there an existing system to generate winning angles based on some kind of evidence? not just from the air, for example my angles are from ripped banners, i change it but its not too far from the original, and the banner being there in the first place is some kind of evidence that that angle works. and i develop the angle say using cashvertising and the deadly 7 sins, and maybe i refer to common things like childhood stories, movies, dreams that everyone has.
2. what rules should i be using to cut placements? i realised on certain ad networks with many sites/apps placement cutting can be very slow, i waited like a week and only cut 2 placements, maybe i was not running enough traffic? or not bidding enough? im currently using the rule of if no conversion after spending CPA x 3, cut placement, and if there are 700 impressions with no clicks, cut and if there are 3000 impressions with less than 0.2% CTR and no conversions cut. but thats still very little cutting.
Home > Mobile > Follow-along Campaigns