Mobile: the beginning of the learning curve (15)
I selected a few initial offers I wish to try. My creatives are still in the approval process after 4 days so in the mean time I decided to familiarize myself with the process with another offer because I saw a fun angle to try, and it did not require creative approval. Spend my wait time with a bit of learning-action instead of learning-reading.
Traffic: Decisive
Offer: Antivirus (Accepted traffic - search/social/display/contextual this is all good on decisive, correct?)
Geo: Ukraine
Payout: 0.75
Bid: 1.125
This is mostly a test campaign to ensure tracking is set up correctly, and that I am looking at data accurately and not making silly mistakes in understanding.
As a result I spent $0.83 running traffic lol. 0 conversions.
I made 10 banners, a mix of languages, styles, and a couple I simply jacked from google images as a control. Direct linked for the time being.
I know I need to buy more data for statistically valid conclusions, but I have a few concerns everything is set up correctly.
1) Proper testing of mobile offers. In my affiliate network I first went to my mobile link in test mode. It just kept loading and loading. Clicks were registered but i never reached the offer. I tried using http://www.mobilephoneemulator.com/ however I encountered the same issue. I then connected to a proxy in Ukraine with my browser, accessed the emulator and then offer with similar results. Not sure if my method of offer testing was correct. I know I can get redirected to odd places with mobile offers from outside the geo, however I simply could not access the offer page at all. According to my AM, nobody is on this offer currently. (offers waiting on approval for are top offers, just curious about this one)
2) Similar set up as previous, but with an email submit offer I am waiting for creative approval on. No issues accessing the offer, my clicks were being registered however my submission was not registered as a conversion despite the proxy in the desired country I was using. Is my method of testing mobile offers from a computer in a different country correct?
3) Banner CTR: Higher is obviously better. My banners ranged from 1.3% to 4%. What is the cut off for banner CTR? Of course I should gather more data first, but under what performance should I kill it and replace with a new modification? Splittester.com says I have enough data to be 95% confident these banners will perform differently over time.
4) tracking discrepancies between decisive,
5) Voluum data collection: In voluum I see where all the views are coming from as I switch between tabs, however my costs do not seem to be being imported from decisive, other than the clicks all columns are empty. No costs, CTR, or negative ROIs. Where have I gone wrong in setting up voluum so that this data is missing?
6) I want to eventually use my angles for an offer on another network, however the payout of that one is only $0.16...Is this too low to be workable or still worthwhile?
7) Right now I can only see individual banner ctr on decisive... is there a way to see it in voluum?
cheers
Where are your creatives in an approval process for 4 days? This should not take so long.
Good choice to take action and get used to everything!
Decisive is display traffic, so yeah it's allowed for the offer.
1) Ask your AM to check if the offer page works. Usually the preview should work, when using a proxy the loading time is sometimes a little bit long but it should load as well. For the payout it guess it's an app install - so there is no reason why you should be redirected to something else using a proxy. In general I'd pick top offers for the beginning, there are enough variables already.
2) Maybe they somehow saw that it's you. Even if you fucked up the tracking a conversion should show in the affiliate network interface. I tried an email submit as well when I started, check my Follow Along to see how it went haha. I'd try lower payout easier converting proven stuff for now.
3) On Decisive I usually do at least 10.000 wins for every banner to make CTR decisions. You can do it earlier if the difference is very big. Keep in mind that the CTR ultimately does not matter, what matters is how many conversions each banner is making. I've got some banners with higher CTR which are making less money than those with a lower CTR but better conversion rate. For now it's hard without having a bunch of conversions so I'd just cut the really bad banners CTR wise - they're unlikely to outperform the best CTR banners.
4) Little click loss from decisive to
5) Sadly voluum can not automatically important clicks from Decisive. You can set the CPC manually after looking at your decisive stats in the voluum campaign setup window. For exact placement costs you'll always have to look in Decisive.
6) The payout does not matter so much, if the offers are similar split-test different networks - often the highest payout will not make the most money. $0.16 is very low - I would stay away from it until you've got more experience. You could also give it a shot with $10 and see how it performs...
As kepe95 said, the approval time doesn't take so long. Just shoot them an E-Mail. The same thing happened with my campaigns and they approved them 20 minutes after I've sent them the E-Mail.
Thanks!
re: 1/2/4 my AM has asked the advertiser to check but she suggested I try something else in the mean time.
This is with clickdealer. She got back to regarding creative approval and said that they denied them, but could offer no guidance on what the regulations are for creating permissible banners when I asked.
Then I asked for top performing offers between 0.5 and 1.0 payout in the app download space. And she gave me a list of sweepstakes with payouts between 2.50 and 6.50.....
Rather frustrated by this AM. Do I have any recourse here to switch AMs?
Right now I cannot seem to find individual banner ctr on
Also, can a mod edit the title to put the red "Follow Along" in the title of this thread?
Thanks!
Be firm with your AM - if they send you a junk report then tell it's not what you asked for... if they are continually unhelpful then ask for another AM.
When you say banner CTR, you mean clickthrough rate, not conversion rate, right?
Clickthrough rate comes from clicks/impressions. The traffic source is generally the only platform that tracks your ad impressions and is therefore the only place you find CTRs.
The exceptions are platforms that can use impressions tracking tags (view tags) to also count impressions, though the traffic source must explicitly allow these as they have to be embedded in your ads. Ad servers also track impressions but you'd know if you were using an ad server.
Title edited!
Thanks Zeno, I re-asked my AM and will see what happens.
Yes, I meant clickthrough rate.
Bit annoyed right now. Spent 4h making 25 banners for a new campaign.
OLX app
Venezuela
$0.92 payout
Decisive rejected because it was redirecting to alibaba instead. Asked AM about it. Told me they cannot remove the geo redirect and I should avoid targeting Opera, Dolphin, UC, DU or devices like Blackberry & kindle fire. Correct action is to simply block these sources in decisive (correct?). However how do I block these in decisive? I can see BBY and Kindle fire in devices, however is there a simple method to block them while setting up? seems as if I would have to select every device except these. I could of course blacklist after the campaign is running, but not sure I can get through the approval process without them blocked if there is a redirect.
Can't even find Opera/Dolphin/UC/DU in the app list while setting up a campaign, so not sure how to fulfill my AM's wishes?
Then my AM told me only stock creatives were allowed...odd because it did not say that in the offer. aarg, and after I made 25 banners.....
Just want to actually get some data and start refining and split testing!
Don't worry about them... they may convert, so who cares. It's no problem for the advertiser since they don't pay for what doesn't convert. If they can't convert or the advertiser explicitly doesn't want them, or they make up a hefty % of your traffic and you've been advised not to send that traffic, just blacklist them as you mentioned.
Your issue with geo-redirection is trivial to fix - just add rules to your tracking system in the campaign section to do geo-redirection first, before the affiliate link.
See here: http://stmforum.com/forum/showthread...king-Campaigns
In a nutshell, add rules to your campaign so that if country =/= Venezuela ---> send to "Passthrough offer" which is a second offer you have added that is the app store/destination page, no affiliate link.
This way users only go through an affiliate link if they are eligible to convert and reviewers won't get redirected by the network.
Finally, if only stock creatives are allowed, I would really move to another offer. OLX has been around a while and with no creative freedom you may find it tough to get any traction - since your angles and ads are completely oppressed from the start.
Alright, so after the issue with OLX, I switched to a dating app install for android, and an android antivirus pin submit
Turns out the pin submit offer had a LP that did not comply with decisive's rules, so I'll save the banners I made for when I get all set up for VPS this week. Hopefully my lander between the two will avoid this issue. If not I'll run on another traffic source. I'd hate to waste another 25 banners....
Dating app I split tested against the stock creatives. a couple of those were denied for being too suggestive lol. Not even any tits. Just "I want to fuck" lips.
Mobile cookbook appetizer style:
0.84 payout
banners google translated, but a native speaker said it all makes sense
Day 1:
total spend: 18.83
total revenue: 1.95 (3 conversions, one of which triggered network's redirect to another geo with lower payout)
ROI=-90%
w/a
total spend: 7.17
total revenue: 0.84 (1)
ROI=-88%
m/a
total spend: 4.90
total revenue: 1.68 (2)
ROI=-66%
w/s
total spend: 4.73
total revenue: 0
ROI=-100%
m/s
total spend: 2.03
total revenue: 0
ROI=-100%
So after this day, only 3 conversions, my data almost gives me no statistical significance! But was able to kill a few site placements which I spent more than the payout and were taking up almost 1/3 of spend till that point. No apps met this criteria for blacklisting.
Banner CTR was around 2% for site placement, and around 1% for app placement.
Of the 3 conversions, one type of app generated 2 of the conversions.
Creating a campaign targeting users of these apps, which also being social apps, should prequalify leads? if they already use social apps, they are more likely to download another. Is this a good idea? Or would a hyperfocused campaign increase click costs too much?
Not sure about other apps like something like "bubble shoot mania"...seems like a kid app, hardly the target of a dating offer. But a few had decent CTRs...Should I just try blacklisting all games in the app field? I know I probably should not.
Day 2 is underway for 13h now. Unfortunately my conversions have been the exact opposite of the above. No app conversions, 4 site conversions. Not really enough to draw meaningful data from. Although CPA today for site placement is already < rev, so I am few cents ahead (for the day, not the campaign)
Read a bunch on angles last night to hopefully improve my CTR, getting them translated and will create a second campaign with them for an improved CTR, it currently seems really low.
I don't think there is much else I can do until I have more data. I doubt the total 7 conversions I have so far is statistically significant to make valid inferences regarding trends.
Why can't I just blow it up with high bids and daily spend, and then just pick out the profitable bits? Seems like that would take less time, but there has to be a reason I have not seen this strategy promoted anywhere...
Especially the m/s category...did not even reach half my daily spend...can I up the bid to get more traffic? Or let it go and take as many days as needed to reach 15 total spend.
Yeah for dating you really need to get the placements right here...
Which is going to be quite hard. Casual strategy games like Candy Crush... probably not the best spot and you have also locked yourself into a single gender - which will eliminate 50±40% of your audience depending on the demographics using various placements.
With this yes you could increase bids substantially to get more data and traffic from higher CPM placements, which may offer higher quality traffic. But you run the risk of spending $100 only to find the offer simply doesn't convert well on most of the placements at least with your angle - so in the beginning it is often better to be conservative until you get comfortable with the data handling, decisions, etc.
Day 2
w/a
total spend: 7.11
total revenue: 0
ROI=-100
m/a
total spend: 5.68
total revenue: 0
ROI=-100
w/s
total spend: 3.87
total revenue: 2.52(3)
ROI=-34%
m/s
total spend: 2.00
total revenue: 0.84(1)
ROI=-58
Very few things could be blacklisted. So have been kind of just waiting for data to roll in.
D3
w/a
total spend: 2.76
total revenue: 0
ROI=-100
m/a
total spend: 2.28
total revenue: 0.84
ROI=-63%
w/s
total spend: 2.54
total revenue: 0
ROI=-100
m/s
total spend: 0.92
total revenue: 0
ROI=-100
D4 data is almost completed for the day with one conversion for w/s.
I bumped up the bid in an attempt to spend more of my budget through the day, I raised the bid from 1.14 to 1.31 for the remaining 3 as only w/a hit 15 spend per cookbook. m/s has spent only around a third of the 15 needed! So far my daily spend has continued to decline unfortunately. Not sure what I should do...increase the bid even more? Wait 10 days at this spend rate?
In any case, I have a couple more campaigns to launch asap. One starting at midnight tonight is a rather targeted test regarding some placements that may be profitable.
The next is more my style. 45 banners. lots of angles. Some pushing the envelope. Going to increase daily spend and bid significantly and see what pops out. From looking at the data, I think I have isolated what I need to blacklist. With my new better angles (cashvertizing, and Finch's 7 deadly sin angles) Should bump cpa down significantly. After reading about angles my initial banners look like shit.
To do:
Finish the last 30 banners
figure out how to archive all this data. Anyone have an excell they use that they could share?
Yeah roll out your new stuff and kill off the old if it's simply not working.
Rolled out the new stuff.
50 banners. 100 daily spend. bid 1.5x payout.
And its only buying <20 worth of traffic a day! This was targeted w/s as in my previous experiment it had the lowest cpc.
However this time my cpa is like $9, instead of the $2 before, and I am not getting as many conversions oddly.
I have blacklisted a bunch of placements where cpc is 3x average and I have spend 2x payout...
however my conversions are coming from things where my ctr is insanely small, like 0.05% vs the average for the campaign of 2%. Of those few clicks on these low ctr placements, often >20% convert, but limited sample size of course. My only conversions are coming from placements where there is no way they can be profitable. wtf. I should of course still kill them though, correct?
Just feels like I am not getting enough conversion data to make any decisions. I have never managed to reach -50% ROI yet, which may mean I should try another offer? Zeno, in this case what would you do? up bids more to see what happens and spend more per day so that I get more data? or try a new offer?
Are those 50 banners inside 1 campaign?
I would definitely not do this. 50 banners with a $2/day budget each may be clashing with the delivery systems. Run fewer banners and give them more breathing room in terms of budget, else you may unwittingly starve them of any traffic.
Thanks zeno! Been doing that the past days.
One issue is my new-sized creatives (ie 300x250), I started with the 1.5x payout target bid, however I do not seem to be driving significant traffic. Some camps on this size I only win<1% of bids. is there an upper limit to how much I should up my bids? Almost at 2$ now and still only a trickle of traffic.
first profitable day. 
Home > Mobile > Follow-along Campaigns