Mobile CPI Follow-Along (Simple) (20)
Hello everyone,
I'm a big fan of simplicity and keeping things neat, clean and organized. As such, I'll try my best to not over-complicate things. I've been researching mobile like a mad-dog these past few weeks. A few days ago I launched my first campaign.
Basic Info
CPI
Traffic Source: TapIt!
Geo: Indonesia
Campaign Template: Mobile Cookbook (mostly by the book)
Split testing between two offers and between direct-linking and a landing page.
Results after first day
Basic newb question: Do I cut placements based off of clicks or conversions? As I saw zero conversions, does this mean I scrap everything?
As I can see from your stats - try to work on increasing your CTR before you move on. 0.5 is too low
Hey mate, good start ! 
I'll try to keep things simple and clean as well, so here we go:
1. The CPC is way to high right now, to make low payout work you will need a < $0.01 CPC, or if the CPC is higher reallllly well converting traffic which I guess will be an issue when scaling
2. When a placements performs terrible CTR-wise (like one banner has half of the clicks and there is no obvious reason why it would be better converting) cut it at around 3 times spent for low payouts. It's better to make decisions based on the conversions but that's not really an option when you are just getting started
3. I am not very familiar with Tapit!, and every GEO is different, but 0.56% CTR seems very low. Check the placements if some of them has a terrible CTR with high spent. If you did everything right with different banner variations I assume it's not the banners fault.
4. Something else is wrong, otherwise you would see at least some conversions at $60 spent!! Check your whole funnel again and make sure everything is right. For low payout apps DL is easier in the beginning because you can't mess it up like you can with a LP.
Lastly, even if you want to keep things clean and simple, I would share a little more details about the campaign, funnel, ... what and how you are running. It's not really a secret at this point and the help you're going to receive this way will be way more important than keeping it behind closed doors.
Originally Posted by mareki
I thought I did a decent job with getting a wide variety of ads: I ripped a few, had a few professionally designed and designed some myself. The ONLY thing I can think of is the language. I DIDN'T translate anything. I read in a few places that its better to not translate in Indonesia, so perhaps I was wrong.
I'm thinking I should get some translations done, run my campaigns again exactly as before and see if anything chances.
Originally Posted by kepe95
1. Lowering CPC. I have a few guides on improving CTR but I'm coming up short when it comes to finding how to lower CPC. Anything to point me in the right direction?
2.-3. I went through some of my ad reports and it looked like a few were gathering a HUGE amount of impressions but very lower CTR. Other placements had a much higher CTR but not enough for statistical significance. Looks like I need to cut the fat!
4. Yea, it has taken me a WHILE to get my head wrapped around tracking. I watched Zeno's vid on tracking, read through multiple tracking guides and still came away with headaches, lol. I went through my tracking pretty meticulously, and tested it myself. Looks good on that end.
Thanks for the reply.
I was split testing this particular offer with the same offer from AppFlood and I noticed a few days ago that a) I could not login into the advertising platform and b) my AM manager nor seemingly anyone at AppFlood could find the time to respond to my e-mails. So, I just decided to drop the AppFlood offer from my campaigns. Not sure what's going on with that...
I'm still splitting between direct-linking and landing page. Like I said before, I took a lot of care in creating this campaign but it looks like I'm doing something MAJOR wrong. Just need to find what that is.
Keep hammering away.
So, like I said before I know I'm doing SOMETHING wrong. Let's look at the most integral part of this equation. My offer.
I switched to a new traffic source, ZeroPark and kept my lander the same. Some interesting results ensued. First lander was in English. Second in Indonesia.
376 visits and 46 clicks on my English lander and 1,825 clicks and 0 clicks on my Indonesia lander. Wow!
Still have yet to garner a single conversion.
Should my next move be to try a different offer???? I am thinking of staying with ZeroPark for a time being for two reasons:
1. I feel like landers are probably my weakest area and ZeroPark's PPV style forces me to improve my banners
2. ZeroPark and
Where most of the clicks in the Indonesian campaign from only a couple of targets?
You need to optimise first before thinking too much about low CTR/CVR... with PPV you often get a big chunk of traffic from a very low quality source and need to blacklist asap.
Originally Posted by zeno
I also started running this campaign in Brazil, but I have been getting NO traffic. I've spent less than $1 and this is with a bid place of between 2-4. I'm not sure what's going on here? Looking at the available traffic number it looks like around 50,000. Just not enough traffic in Brazil for this particular traffic source??
Most sources have a fair amount of BR traffic.
Where did the 50,000 come from? Their media planner? It could be your targeting that's doing it. Most BR traffic will be Android, a lot of it WiFi.
Finally.....some conversions!!
I have a few questions.
1. How much should I spend to get enough data to start cutting placements?? How many conversions? I noticed one of my landers was performing much worse than the other two, so I went ahead and pulled that. In the cookbook it says to cut placements that have 3X payout without any conversions. Does that apply to PPV as well?
2. Do I just need more conversions in general to start optimizing?
1. Spend at least the offer payout on a placement unless it a) shows terrible ad CTR and thus prohibitively high CPCs, b) shows bot like behaviour e.g. 100% or 0% lander CTR, or c) shows terrible lander CTR giving prohibitive cost per click to the offer.
Just come up with a benchmark rule and follow it. Rules will vary with traffic sources as many things change, but your process should remain similar.
2. Yes. Conversions are your most important piece of data. But don't go down with a sinking ship - if a specific lander, offer, traffic source or targeting parameter just doesn't convert, drop it.
Stats!!
This is so far with ZERO optimization. Split-testing between landing pages and offers.
Goal for today:
1. Go through after work and start analyzing and cutting placements where need be.
Some Lessons Learned:
1. Don't cut an offer too early! I had noticed early on that Offer B was seemingly outperforming Offer A pretty drastically. Waited for more data to come in and when I woke up this morning, Offer A had more conversions than A!
More stats. From the last two days. This is from a single campaign split testing two landing pages.
I think I still need more data. Most of the conversions come from one of the bigger placements. The ROI is ~50%. Other placements have far higher ROI but they seem to be quite small and I don't know if my data is significant yet.
Today is the day I FINALLY realized how important tracking is. In the back of my mind I always wondered what exactly it is for. Well, when you can sift through OS, carrier, device, browser data it comes REALLY in handy!
What I'm going to do over the next few days is this:
1. Let the campaign run for another day to accumulate more data.
2. Add another 2-3 landing pages and see if I can notice what exactly will increase or decrease ROI.
More stats:
Not bad. Making small progress.
Next up:
Continue to cut placements (I just started this).
Create variations of my LP's (they all have different designs but same angle).
Looks like it's all going well - and more importantly, you're learning a lot of very useful things!
So, after accumulating more data I ended up cutting my two worst performing landers. I then sat down and thought long and hard about split-testing the body copy. Came up with some ad copy, had it translated and then made two variations for 3 landing pages (including the one original). Then, I let the campaign run again and woke up to...
Some green!! I know it's not significant at all but it's the first time I woke up to green instead of red! Talk about surprising, LOL. My ROI has been steadily improving after cutting placements and making lander improvements, so the important thing for me to focus on right now is getting the order of optimizing everything right. If the ROI on my winning lander holds (probably doubtful lol) then I will start tweaking minor things, like headlines and images.
One thing I am little confused about is the cutting and optimization process.
1. Should I continue to cut placements after reaching 2-3x ad spend with no conversions?
2. What kind of target CTR and CVR should I be looking for?
So, I ended up finding a profitable campaign after doing some tweaks here and there.
Stats over the last 7 days.
I've been trying to figure out how to scale this. What I did after I was consistently at 25-50% +ROI was double my adspend, then double again. However, my CV% has been dropping steadily.
What I've learned so far:
Actually running campaigns is the greatest teacher of all.
Great work!!
Concerning your questions:
1. Not sure if I understand the question the right way. If placements have 2-3x payout spent with no conversion cut them. The same goes for placements which show no potential earlier, Zeno wrote a great explanation of those above. In general you can answer a lot of those questions by intuition/thinking smart or testing.
2. Concerning target CTR and CVR, it really depends on traffic source and dozens of other factors. I would just look if the ROI is good and you can make good money - without worrying too much about CTR and CVR if it's in the green. and of course always split-test stuff to see if you can improve it. Continue doing this for some months and your funnel will improve more and more.
Concerning scaling I'll let the pros answer, I guess it depends a lot on the traffic source.
Check for bot traffic. Not trying to get rude but tapit has a lot . Good progress ! Keep going !!!
Home > Mobile > Follow-along Campaigns