It would seem that offers just don't convert (16)
I'll preface this by saying that this was my first campaign - I don't yet know what's hot and what's not.
I decided to run a Go2Mobi campaign on an Android utility app (reward per install). I set up 5 creatives, started with a low bidding threshold (first $0.2, then settled on $0.1) which I found throttled spend sufficiently to get enough data at a low price-point for the purpose of testing. The reward on the offer was $0.32 per install so I wanted to start with a relatively low CPC prior to campaign optimization.
So far the campaign has been run to a total of ~28,000 impressions and 256 clicks (CTR of 0.91% at a CPC of $0.02). The CPC here seems reasonable to generate profit. Having done a bit of research, it seems that Android click to install rates are on average about 10:1 which should mean an unoptimized $0.2 per install - nicely under the reward. However, in 256 clicks the offer has not yet converted once. I've run a test install of the offer to make sure that their conversion tracking was working properly and it is.
It seems that some offers just don't convert. Anyone have any experience with non-converting offers, or insight into expected click-to-install rates? I'm about to give up on this campaign as I'm pretty sure the offer is a dud, but I'm interested to hear what others think first.
Thanks
You might be right, and you might be not 
I bought a solid amount of clicks from go2mobi and I can tell you some things :
- There are placements/apps in the mix that simply won't convert no matter how hard you try. If you were "lucky" enough to get a large % of traffic from those, well no wonder you didn't see a conversion yet.
- Some bid levels work well better than others, hit the wrong ones and conversions won't happen.
- Some exchanges work better for some offers, maybe you tried the wrong ones.
It very well possible that the offer is a dud indeed. But on the other hand, I know there are situations where 256 clicks and no conversions don't mean shit ...
Did you use a landing page? Some offers convert VERY poor without it.
Did the traffic you buy match the required targeting? I know go2mobi isn't perfect when it comes to this, especially when it comes to carrier targeting.
I really don't want you to waste energy on something that has 0 potential, but at this point, it's too early to tell in my opinion.
Hi again Matuloo
Broken down by point, here are some of my campaign details based on your questions:
- I didn't use a landing page, the flow is banner click and the landing page is the Google Play app store (install to convert)
- The targeting I set up was device specific (whitelisted Android) - phone only, apps and website based placements (confirmed OS of clickers is Android in reports by filtering the items with > 0 clicks)
- Geo is DE (Deutschland), one of the regions in which the offer is available
- I used Smaato exchange b/c according to Go2Mobi's infographic, they run utility app ad traffic (Nexage is another one which I have not yet tried)
https://www.go2mobi.com/insights/eve...ffic-partners/
I'll un-pause the campaign and let it run at a significantly higher CPM bid and see what new data becomes available. Thanks for your insight here, again, it's much appreciated.
Continued running campaign to 500+ clicks, still no conversions (AM reported avg. EPC to be $0.01 presumably in an optimized campaign at conversion rate of a little under 3%). Maybe there's still work to be done here, and more data is needed.
Hey, salamander!
Nice to see your thread. So what I can recommend you:
- Why you're not using a LP? For me LPs works very good, and your ROI will be much better with it.
- Did you set up different creatives with different angles? Because for me to change the angle can give absolutely different result. If you
If you have 0.02$ CPC and already 500 clicks, and your payout is 0.32$ - it means that you did something wrong. I recommend you to try with LP and different angles. And what offer are you running? May be you can find another offers in this vertical and test them too using a split-test?
Ok so you're not using landing page and you are buying relatively low quality traffic with no targeting, what kinda banners are you using? Some generics to trick people to click or you are using some angle on them?
Direct-link can work if you use something like adwords traffic, but smaato will not cut it I'm afraid.
Another problem is your GEO, Germans don't install apps as wildly as people in Asia or Latam for example, it's harder to convert apps in more developed/matured GEOs according to my experience. People who are new to the world of smartphones are installing apps more willingly, and of course you can find more mobile-newbs in emerging markets than in Germany. On top of that EPC of 0.01 doesn't seem too hot for a market like Germany, it's more like what I'm used to see in India for example.
Next problem is the amount of placements you are buying from, if you look at the placement report, how many placements are those clicks from? I guess it's gonna be a lot of them. You're running with a low bid, so the chances are, some quality placements are most likely hidden from you and bulk of the traffic comes from placements nobody wants to buy from. I see you mentioned you will restart with a higher bid, did you?
I'm trying to help you see things from perspective, it's quite possible that the offer sucks, but what if you just buy from the bad placements because the good ones are too competitive? What if the problem is the absence of a landing page? I agree this doesn't look promising at all, but it's definitely possible to see 0 conversions after 500 UNOPTIMIZED clicks and it still doesn't have to mean the offer is bad. Even tho it very well can be Hope I didn't confuse you much.
Hi Mrbraun
My CPC was about $0.03 by the time I paused the campaign. The offer I am running is an Android app called Turbo Cleaner. Right now I have 6 creatives running and have hit 1 conversion after 1248 clicks. Can I ask you a bit about landing pages and why they improve ROI? Are you suggesting banner to landing page to Google play store? Or are you suggesting pop-up etc? I`ve started another campaign in a different Geo for this offer using native ads. CTR is about 4% as opposed to 1% so value seems much better, considering the ads are relatively cheap. I did do some optimization on the previous campaign but I don`t think I can make it profitable. Still a lot to learn though - if you could offer some insight into a landing page based approach, that would be great. Thanks.
Just read your post Matuloo, so I think you`re definitely correct about Germany being a poor choice for this campaign - I`ve targeted another Geo now and have swtiched to native ads (at this point, I need to get my feet wet so to speak with different options, just to learn the business). I did restart with a higher bid and culled a lot of the placements which had a high number of impressions but no clicks - you`re most certainly correct about low bids yielding poor quality placements. I`m not done with this offer yet, I think there`s still a lot to learn, even if it may not be a profitable one.
Regarding the banners, they do not employ any trickery, they target 1 particular angle which is representative of the app`s offer. It could well be that my angle is poorly chosen as well. Just have to keep at it until I hone in on the specific issues.
So granted I don`t believe this is a solid offer, I still want to pursue a landing page approach to see how that improves the conversion rate, so flow would be banner->landing page->app store. It seemed counter intuitive to me at first b/c I was of the mindset that a longer conversion funnel would mean less conversions, but apparently that`s not the case.
Thanks for all the feedback.
Originally Posted by salamander
As for the landing page, it can definitely help the performance in many cases, so as a general rule in AM - you should always at least test them.
In order to get better results than with a direct link, the LP has to be more aggressive and "straight to the face" than the standard google play store lander. You need to push it on the edge - bold claims, scarcity claims - basically rave about the app as much as the advertiser allows - the basic principle is to make the user BELIEVE the app you are presenting is not just another app they can find in the store on their own, it is an exclusive deal that they should THANK YOU for showing them and it's only available HER and NOW, for the next 10 minutes.
In many cases you need to take it even further and resort to scare tactics and cloak it because advertisers and traffic sources don't allow this. I'm not a fan of this approach, but people are doing it all the time so I'm not gonna lie to you and pretend I didn't hear about it yet
This is the case especially with antivirus apps for example.Good luck
Yeah, I am using native ads with Go2Mobi (MoPub) - the CTR is about 4x better than what I was getting with straight banners, probably for two reasons (1, they look more like natural listings, 2, there is more space for up-front information than in the tiny banners alone). Yeah, I'm familiar with the aggressive landing pages and the advertisers (for the most part) do not condone it. I'm also not a fan of this approach as it is liable to burn bridges with Affiliate networks. I know exactly what you mean though, about the sense of urgency. I often ask myself who actually falls for these offers, but the proof is in the pudding. Utility apps seem like a hard sell, and I'm not looking for an easy way out, but I think there are better offers to run with. I'm about to try out a gambling app, and I think this has a lot more potential. The conversion stats are significantly higher than the utility app and it seems like something more people would be interested in. I'm getting ready to set up
Originally Posted by salamander
But, I still get contrary info from some of my affiliate friends, apparently it still works well for some ... so go figure
I'm not pushing apps myself anymore tho. The part of my biz that was about app installs is now focused on PIN submits, those are easier to run IMO, at least in GEOs that have not been optimized to death yet. But by all means try that gambling app, I heard good stories about something similar.
Cool, I'll give it a shot and let you know
Seems like it's time for a follow-up. I ran with the casino app campaign on a different traffic source (startapp) with a fairly disciplined approach. I like that they have a CPC bidding mode, which is nice for discovering high CTR placements although it is pretty pricey an option. In the end, the campaign was not a success, but I did manage to learn a few things - here's the overall strategy I took:
1) Generated multiple creatives for each ad type (rich text, full screen, banner, etc.)
2) Ran ads on all placements / connections with CPC bidding at the recommended bid
3) Identified converting placements and high CTR placements, blacklisted low CTR placements with over a threshold level of impressions
4) Culled ads which weren't performing
This basically left me with a somewhat lean campaign, I was then able to see which time slots were generating cheap conversions and started day parting for just those time slots. In the end, the campaign was not profitable, but it wasn't terrible either. I'd have to run it longer to get some final ROI numbers, but it's obvious that the near-profitable times are simply not generating enough traffic volume to be worth pursuing. My plan was to switch to CPM bidding once I had over a certain CTR (It ended up being anywhere over 10% CTR was more efficient to use CPM bidding), but I found that once in CPM bidding, impressions would start to spike (new placements with terrible CTR). Even in a pure whitelist test campaign, I found that impressions would spike and the CTR would plummet. I don't want to cry foul play here, but it was interesting to say the least.
My two main takeaways from this experience are:
1) Being able to throttle traffic is very important for testing the entire day, over the course of the week without things getting very expensive
2) Easier to convert offers (aka. cheaper payout) is probably a lot better for testing since you can get a larger sample of conversions with which to work - 1 or 2 conversions per day don't really tell me enough
That's it - time to go back to reading, and get ready for the next play. Thanks.
Congrats on the much more successful campaign! Looks like you're working with a pretty solid methodology there - looking forward to hearing about your next results.
Originally Posted by salamander
The network want's to make as much money as possible and the whole algo is designed around it. By knowing the CTR of your ads and knowing the CPM bids of your competitors, they know when it makes sense for them to serve CPC ads over CPM. This is usually the first few impressions per user. At certain point, let's say after the 2nd impression, the CTR decreases too much for CPC to be profitable for them, so they start to serve CPM only. Based on capping settings and CTRs, they can serve CPM along with CPC from the start, but at some point, CPM will take the lead and unfortunatelly the deeper yuo go with the impressions per user, the lower the quality - once a user has seen 10 banners, the interest is gone.
This is why I don't like when networks mix CPC with CPM, the best is to do either one or the second. I always prefer CPM personally, it gives me more control and more even distribution of ads too.
BTW : I'm glad your results improved a bit, keep going!
Thanks for the support everyone.
Home > The Newbie Zone > Questions and Answers