Hello,
Can someone help with this data? I feel I do the same mistakes every time I run a new mobile campaign.
It is an app install with payout $0.2
In my opinion, I would blacklist all these placements because spending has reached more than 3X payout for some placements and no conversions. For other placements, the CPA is too high compared to the payout.

Is it a good decision? Bad decision? Why?
Thanks
First step, look for all the outliers. That means looking at data across the 15+ tabs and blacklisted those targeting criteria that have had zero conversions but spent $1+.
Step two, look for any single super spenders that have eaten up a big chunk of your budget at a high than average CPA.
Thirdly you should bid lower, you should be seeing lower than 1 cent clicks with those CTRs.
1. The first place is to focus is placements, but that's assuming there are no huge outliers in the other tabs.
2. Yes.
3. 6x CPA is a bit agressive for a totally unoptimized campaign. Try 3x.
OK, since I have you here, I will try to get more help from you if you don't mind
and the readers might also benefit 
The campaign is paused now since my first post.
I know my cpa goal is too high, but anyways, with this data, what would you consider outliers? I am targeting one country, android OS and wifi.
These are my creatives data, maybe blacklisting the last two creatives?

Now the state data, I should blacklist Jalisco and Veracruz?

This is the OS version, the last 3 rows blacklisted?

Devices, none blacklisted yet, because spending is less than 3X payout for the non-converting ones?

And Category, I should blacklist IAB9?

The subcategroy is not recording conversions, I just noticed that in all my campaigns... They might have a bug.
Would that be a good approach? The CPA is pretty high though, but anyways I would like to get the methodology right at least 
Thanks
I don't think you have enough conversion data for the creatives to cull them very effectively just yet - data is from a short time frame and their CTR's are similar. Is the angle different between the creatives. You could go out on a limb and focus on the top 2 creative but I would probably focus elsewhere first and get a 2nd day worth of data on them to see if the trend holds up.
I wouldn't normally cut by state and in this case the data is pretty spread out. Might be something you'll want to look at later after optimising more influential things.
Same with OS version. Device is more important (and why would e.g. Android 2.3.4 >> 2.3.6?)
There are so many devices that the data is quite thinly spread. There are no huge outliers... Different devices will use different apps and the users might prefer different sites so it all kinda leads back to placements being the critical point of optimisation.
However, looking at your categories, arts & entertainment seem to be where it's at. Blacklisting any of these will wipe out a lot of placements. Hobbies & interests = low win rate (probably higher CPM needed) and no conversions. However have all these impressions in this category likely come from 1 or 2 placements? If so, might be wiser to kill those instead. E.g. if TuneIn Radio and Mexico FamilySafe are both in the Hoobies & Interests cat then they would account for most of the impressions... But if they aren't probably safe to just kill that entire category. I would kill placements first then watch how hobbies&interests responds and then blacklist it if it still performs poorly.
Society and Tech/comp = high win rates, low CVR, definitely traffic from many placements. I would kill them for now.
Assuming your OP image of placements is still relevant, I would: Kill Grindr. Kill IM+ android. Kill Android. Kill 4shared mobile. Kill Mexico_android_FamilySafe. Definitely get TuneIn radio the hell outta there. (all very low CTR...)
I've lost my train of thought, good luck.
Haha I think Zeno is the total opposite in terms of left brain, right brain compared to me.
He's left extreme I'm extreme right.
There is no clear "right" way to go about this, but how I would do it is...
1. Murder the two bottom creatives now.
2. Lower bid to 3x.
3. Blacklist tunein, grindr, ims android.
4. Make a seperate campaigns for mp3skull and tubemate.
5. Make new creatives, to get closer to 1% ctr.
6. Run and analyze data again.
Lol yeah we probably are... However your approach unarguably has a lot more experience/intuition behind it (vs just logical reasoning) and probably appeals more to a newbie! So I say go with that!
Thanks to both of you guys, really helpful to help me get a methodology of thinking.
So Lorenzo, you did not see any outliers from the other tabs and this is why you started blacklisting apps? If you had seen any outliers, would you have blacklisted the outliers or not yet?
1. I saw two of your creatives were crippling your campaign. They were the outliers.
2. I blocked those targets because I was paying attention to the CPCs. $3 CPC, .75 CPC, and .74 CPC.
3. Time doesn't mean shit. Always look at spending and conversions.
Derp
Yep I did mean CPC. 
Another chance because low CPC compared to others and it seems to have potential? Is that it?
If the starting bid was correct, 3X the CPA you would blacklist it?
Thanks 
Lol I must be in retard mode, couldn't even see they had no clicks. Lot's of good to and fro going on here! :-)


Split test 
I blacklist too.
Thanks Lorenzo and Zeno, this has been really helpful for me 
The LorZeno combo. Good luck with the campaign!
To be honest, I don't care about the campaign 
What I care the most is how to do it well for other campaigns 