Home > Paid Traffic Sources > Push Traffic

---

How Important is BlackListing placements in Push? (24)

---

Is it mostly about the adcopy & icon or is it about mostly blacklisting placements?

I've heard that blacklisting placements doesn't do much in terms of improving ROI. So is it mostly about switching up the ad copy frequently?

---

I’d say it’s: ads > landing pages > sources (placements), in my experience I rarely blacklist placements, however I run most of my campaigns on CPC

---

Ah I see. Thanks


Sent from my iPhone using STM Forums mobile app

---

Maybe it's not as crazy as with pops for example, but I still blacklist quite a lot.

I would suggest you to check your placements at least on weekly basis, there might be some "junk" that just eats up your $$$.

---

When I blacklist bad placements (let's say with rules like 2 x Payout and no conversion = Block OR 4 x Payout & no ROI = Block) I often see that campaign is doing even worse than before. I'm not really sure how to manage that to make it better than worse. Maybe blacklisting not everyday and not so aggressive could be better? What is your experience guys? Because not removing bad placements is wrong as well.

---

I'm even more aggressive than you (especially with higher payout offers) and it worked well for me.

It shouldn't be worse though, maybe the problem is with the offer or funnel itself?

---

Hmm, probably you're right. Or maybe my bid is too low and when I blacklist some placements then my overall CTR is lower = I lose bid battle (as I run on CPC, not CPM model).

---

I wouldn't call it not important. Placements, meaning the site the user subscribed is less crucial then in pop for instance. You can't compare it. One day a certain placement can perform absolute nothing and the next day a surge of awesome users can sky rocket your ROI. You need to focus on where the users are coming from, meaning user age, that's your main optimization tool here
Good luck mate

---

Is there a way to blacklist within your tracker or traffic source?
I can't work out how this works.. or are you literally just writing down on pen and paper "[x] traffic source doesn't work with [y] offer" ???

---

You need to have your tracking in order, so pass the placement/zone token id to your tracker. Then you check the stats in your tracker and blacklist the under-performing ones directly at the traffic source.

You don't do this on a "traffic source" level, as any traffic source has dozens, hundreds or even thousands of placements and their performance can vary a lot. So you don't want to cut out the whole source, just the placements that are not working for you from whatever reason.

---

That makes sense.
Legend.
Thank you man!

---

In most cases i block sources that send many clicks with low LP CTR. For me i always get negative roi on them. But then the ad copy and LP plays a higher role on the conversion.
@matuloo, @erik :

What's the average LP ctr we can expect for T2 geos, like EU ? Im getting like <10% lp ctr and its low for me but i dont know if its the average

---

It depends on the country. In EU, I see LP CTR between 10% - 20%.

Sent from my SM-G965F using STM Forums mobile app

---

Varies for me quite a lot too, I've seen LP ctrs between 5% up to 40% on push traffic in EU geos. Usually, with higher volume placements, the CTR is on the lower end, with some specific placements that have good traffic, it can go way higher. But I'd say 15% is about the average, give or take a % or 2.

---

LP CTR also depends alot on the vertical you run, what kind of landing pages you use.

For example when you run sweeps it makes a huge difference between advertorials vs spin and win.

So in the end a general CTR can be anywhere from 2% (for Crypto advertorials for example) to 50% (for easy sweep landers).

---

I'm quite aggressive too. It doesn't seem to be reasonable to spend like $50+ per target to see if it converts or not. So I'm blocking my targets around 1/2+ of the cost. My offers are usually $15-40, so at $10-20 spend (I'm running cheaper countries so there should be enough clicks).

---

I already told about it in another post but I am also rather aggressive with blacklisting.

In my campaigns I never optimize by browser or OS version or such small stuff.
So when I have a working funnel the only element I optimize is by placements (in Pops and Push).

My goal is always to run my campaigns on profit so basically I can cut all placements that are 1 payout in loss.

In some campaigns I am even more aggressive like payout = $3.5 and I blacklist all placements with $1 in loss.
But such rules I only use when the campaigns are running for some time already and when I have enough (at lest few hundred) conversions to judge the campaigns longtime performance.

---

OS version is basically the only thing I'm using for optimizing. Sometimes connection type too, but it's rare. There can be pretty crazy differences between versions. Like, Android 9 is doing 200% ROI while Android 8.1 is doing like -400%.
So I guess everything really depends on the offer and need to optimize what's relevant to your offer.

---

Sure, when there is such a big difference it can make sense.

But I not only check the ROI but also the volume an element brings.
So when the biggest elements (in this case OS versions) are profitable I don´t even care about the other ones.

When the OS versions that bring 85% or 90% of the traffic are profitable I let all others run as well, even when they are on loss.
Same goes for browsers or device brands or other such stuff.

Because often you have the situation that the more elements you cut and the tighter you target the less volume you receive for the whole campaign.
And often you also receive less volume for the profitable elements when you target too tight.

I set my targeting when I create the campaign and then (usually) I don´t touch it anymore.
Only thing that I then change are placements.

---

The minimum ROI you will always achieve is -100%, it's not possible to have an even lower ROI

---

That was exaggerated intentionally

---

Stop reading my minds guys :P

Oh? But certainly Android Webview... and sometimes OS version--

Right! I was thinking the same thing. But lately, I've been wondering... does it affect the volume the source will send--

I could sit back and let you guys talk shop all day

---

I cannot confirm this with a direct response from a network admin, but I feel like it does. Whenever I tried to optimize to much, I ended up with no traffic at all. And the campaigns where I was willing to take all the traffic were rewarded with the most volume. And I'm not talking about the total volume as that is naturally higher, but specific traffic segments that I was also targeting with the optimized campaigns... the one with low or zero blacklisting was getting more of it.

In the end it does make sense, the networks want to sell all the traffic, so they might have systems in place to reward those who dont try to cherrrypick... but again, I cannot really confirm this with some official statement.

---

In my experience when trimming the fat and pausing bad sources interferes with scaling up it means the camp would've been a dud anyway. Bad offer, market mismatch, slow offer servers in geo, failed lander angle, etc. I will never care about engagement metrics.

Native sources seem the most punitive with campaign pausing and it seems to depend on the RTB algorithm they're using. So you need to use/experiment with other methods than a hard pause in my experience; like a low bid, cpv/coefficient surf to stay in the (rtb alg) game, etc.

If I have positive roi out of the gate (rarely true RON but still pre-optimized for obvious things) I will still trim the fat by blocking worst performing sites/zones/pubs/widgets where I can. Then that gets saved to our list library for network/geo/vertical.

---

Home > Paid Traffic Sources > Push Traffic