Home > Paid Traffic Sources > Push Traffic

---

Scaling Smartlinks on Push (52)

---

Hey guys

Really guys, this is some sort of a self-help post, but feel free to follow along and learn new things with me And it's not even only about smartlinks anymore. Just push $$$ in general. But with inclination to $$$$, or maybe $$$$$. $$$$$$?

Lately I've been running Adult Dating Smartlinks on Push

I noticed some funny behavior when trying to scale these profitable campaigns to other traffic sources though and I'm not sure if that's related.

I tend to run these smartlinks in such way: If the volumes&bids are OK for a particular GEO (i.e. its fine to go with lower bids that are suitable for the average payouts with still sufficient volume & enough conversion % - propeller metric) I would tend to start from High user activity and once i get my funnel & zones optimized I would scale this funnel to other user activities and include broker traffic as well.

If the volumes&bids are not very favorable (i.e. bids are relatively high for average dating offers' payouts which are located inside the smartlinks AND/OR the volume in GEO isn't too high) I might start with all user activities and bid relatively low (place the bid according to the graph on Low user activity and then include other user activities as well).

I then set-up TheOptimizer rules to do the "dirty" work for me.

Ok. Now I got a smartlink campaign doing some profits daily, and I want to scale that smartlink to other traffic sources (I would do that as fast as I can like within a few days of seeing &x-$xx green)= I would then scale it to other traffic sources and *boom* (insert explosion here) campaigns start to tank, I wonder if this is related to the smartlinks optimization, as new traffic is coming in from other traffic sources it is in a way might be trying to adjust & optimize itself to the new traffic and is falling from the good funnel that was working for me in the original source?

I think I noticed smartlinks would be less stable though, even on the same traffic source, sometimes it would be related to lower payout offers starting to rotate in the smartlink for that particular day and sometimes it would be the same payout but just lower CR for them on that particular day. Maybe that's how the advertiser is making up for having no caps?

I tend to like split-testing smartlinks from different networks when doing a test in a new GEO but I usually always settle for a particular smartlink from a certain network within a few days, so I would then only run that smartlink already.

What do you think guys?

Btw, I tend to like smartlinks as they don't supposedly have any caps for their offers. Is it a no-brainer? I know you guys are usually against smartlinks... But this could be an easy setup to scale!



*twinaxe -zen mode now*

---

Yes, that could be the reason.

Often smartlinks work that way that you start sending traffic and they test different landers and offers with the traffic that you send.

Then they start optimizing for your traffic so that usually the performance increases.

When you then send new traffic from other campaigns or sources to the same smartlink it can happen that it basically has to start over because the new traffic can convert different so it has to run the tests again.

What you can do is to try to get separate smartlink tracking links for thedifferent campaigns so that you can keep the traffic separated.

Really man, hard to tell the reasons.

But it´s correct that smartlinks can be less stable than "normal" offer campaigns.

I personally would use smartlinks only for remnant traffic or backbutton traffic.

I prefer to have more control over my campaigns and one part of it is that I want to know with what payouts I work.

And on smartlinks if can happen that bascially every new conversion has a different payout than the previous one.

The single offers inside the smartlinks can also have caps but when it´s reached a different offer just gets rotated in so that traffic can continue.

---

Thanks alot for your reply I wonder what do you think about my setup of making the campaigns in general, as well as scaling?

Just thought I would ask a couple more questions while I got your attention :
- What x payout rules are you usually using for cutting placements on a source like Propeller? In other words, how strict are you with killing ?

- How fast do you take to test your campaigns? Because some affiliates tell me *oh this offer is running more than 4 days so it must be profitable (on spy tools...)* But then I think - hey, I might be running a campaign while just optimizing it for a week or two weeks (based on if its like not a totally garbage -90% ROI but if its like -40% ROI.. yep that's a minus)... because for a GEO like DE for example, with a 10$ daily budget and a decent bid , it can take a while to go over placements.

For example; GEO: UK, Dating SOI Payout: 2.4EUR, Bid: 0.033$, Low User Activity (+Medium&High but set based on Low) because High bids could be more expensive here, running for 1.5 weeks already just to optimize
http://prntscr.com/s0oyox (Sorry for not knowing how to post a picture here like the pros :P)

I would tend to cut them on just 1 payout in loss... or around 150-200 clicks & no conversions (with these SOI payouts, of course), and after 400 clicks, I cut all those placements that are less than break-even (I set all that in TheOptimizer so I'm pretty relaxed), I also do similar numbers for creatives, but I am actually less strict with the rules on creatives, because I feel the zones should come first, and a creative not profitable can also be slightly abit unprofitable because of it going through the shitty zones.
So yeah, creatives are abit more relaxed but in general its a similar rules like 1x payout in loss, 200 clicks & no conversions (but I may restart them if they get paused just because of the reason I said above ) In your opinion, creatives should be optimized less strictly than zones?

But of course I also place in my funnel 6-8 creatives and 3-4 landing pages and ideally 2-3 offers to test. (Just to mention that when I can see some greens for campaigns I am not profitable on the daily basis yet

---

Smartlinks are a shortcut and as such they are limited in profitability as you cannot properly optimize the funnel. That's why I always say they are a good choice in case you have traffic that you dont know what to do with or can buy at a VERY low price. In case you're serious about adult dating you need to focus on single offers, build the whole funnel for max conversions and buy the right targeted traffic.

I understand your motives, it sounds easy but everything that sounds easy in AM usually doesnt work that well

This doesnt mean it cannot work for you, but the price of the traffic will play a major role in the puzzle.

---

And here we go

Basically that´s how I do it as well, but when I see success I first try to get more volume on high activity and then test other activity levels as well.
I also include broker traffic from the beginning.

I personally always start with high activity, even when I have to run on low volume then.

Reason is that I only need it for the test to see if something is working or not.

For unknown outcome in a test I want to have good quality and don´t need much volume, for proven and good converting funnels I can live with lower quality as well and rather want to have volume for scaling.

It depends if I cut in test stage or when I run a converting campaign.

In test stage it´s not that much about "x times payout", there I rather try to get rid of high volume nonconverters to have the traffic distributed across more placements.
When the tests are good and I have a converting funnel I test the excluded placements again to check their quality.

In a running campaign then I am pretty strict with cutting the placements but there is no golden rule for me, it depends on several factors so that I basically have different rules of thumb for different setups.
These factors could be the offer payouts, where I run it, the available volume.

This is why I don´t really like to talk about such numbers or rules because what works in one situation doesn´t necessarily work in other situations as well.

That´s also why I try to explain in the push guide how the things are connected so that you really understand the game so that you can come up with your own dynamic logic that you can then apply to different situations yourself.

But to give some examples let´s first take campaigns in European geos with average offer payouts of $1.20-$4 or so.

There I would cut rather aggressive like 1 payout in loss.
In the end I want to have positive outcome so when a placement isn´t converting after 1 payout then even the next conversion wouldn´t bring it into profit so I kill the placement.

Of course it happens that suddenly after 3 x payout in loss you receive 5 new conversions from that placement but I am no friend of what could happen, I rather work with the actual numbers I have.

When I run campaigns where very high volume is available I sometimes also cut earlier than 1 payout because then you often receive late clicks or more traffic anyway befor the placement is really stopped so that it approaches the 1 payout more or less as well.

On low volume however I sometimes let it run for a bit longer because that´s the only way to collect more stats so that it makes sense to let it run for a bit more to get a better impression of the placements longterm performance.

The we have the offers, for very low payouts I sometimes let it run for more than 1 payout because it doesn´t hurt much.
On the other hand vor very high payouts I cut it much earlier.
As an extreme example, I would run placements for $800 payout crypto offer until 1 or 2 payouts in loss.

And I use rules in Optimizer to re-activate placements with positive ROI when for example some late conversions happen that bring the placement back into green.

Rough test 1-2 days, this is enough for me to decide if I continue with the campaign or not.
3 days is absolute maximum but usually it shouldn´t take long to see if it has potential or not.

And with that I don´t mean that a campaign has to bring in high profits after 2 days, it´s just that it´s enough for me to see if the campaign has enough potential for my personal workstyle to keep it running.

I would decide on case to case and also make a difference between testing to see if it has potential vs optimizing to prepare it for scaling.

For testing even a week would be way too long for me and two weeks would also be too long for optimizing.
In that time you usually should have it trimmed down already to a winning combination so that you are ready to scale.

$10/day for a geo like DE is way too low.

That way you will never be able to compete and you need ages to gather enough stats.
Let´s say you start a campaign with 4 offers and 3 landers and average offer payout is $2.50.

Then the minimum test budget just to see if the whole campaign has chances to succeed would be $150.
When you see promising results the test budget should be around $300.

With $10/day you would be finished with your test after few weeks and then you would only be ready to optimize a bit before you can start scaling.
In that time other affiliates spend 100 times your budget, get results = stats much faster and snatch the conversions from under your nose.

---

Gold

That make perfect sense.

How are you optimizing to prepare it for a scale? I think that's an important point personally.

Yessir! So for more expensive GEO's one should approach with a bigger daily budget Take note, newbies.

Just wonder how you calculated that number?

Makes perfect sense.


I also wanted to ask what do you think about these statements?

So basically I also tend to approach creatives with 1x payout rules all the way starting from testing. I then would also sometimes swap creatives to similar to the better performing ones when they all get disabled based on the rules that I set but I still got zones to go through.

In a running campaign too, I would cut creatives based on 1x payout in loss rules, and once most of my 6-8 creatives get disabled based on the rules I then might also swap some of the excluded creatives with "new" creatives that are similar to the best performing ones. Are you doing this? Do you swap them inside an already running campaign or do you restart the campaign?

This is what my Pushground AM said about their creatives:
http://prntscr.com/s15ll8

---

Drill down to a good combination of 1 lander and good converting offers (mostly 1 but can be more as well when the situation demands it).

One thing I read here over and over again is that beginners overcomplicate the process alot.

Browsers or even browser versions, device brands, OS versions...

I can imagine that all such stuff can seem to be very important but for me it isn´t.

My different targetings basically are mobile vs desktop, 3G vs WiFi, OS and that´s all.

I also target stuff like specific carriers or browsers but only when it´s required by the offer because other targeting isn´t accepted.

A winner is a winner and converts basically everywhere, a losing campaign won´t turn into a winner when I have to optimize by stuff like browser versions or OS versions.
Then it´s simply no winner, easy as that.

For me it´s not interesting if a OS version or browser version has -90% ROI when these elements only send few % of the traffic.

If you exclude it or not won´t have a big impact on the total performance anyway so I don´t mess around with it and better spend my time on stuff that has more impact.

Disclaimer: This is only my personal way to run campaigns and I am used to do it like that for many years already.
Other affiliates have other work styles that can work just as good for them.
It´s important that everyone develops an own style that works best best for him, a copycat who doesn´t understand the game won´t survive longterm.
Again, this is why I try to explain things in a way that you understand the process behind it and not just show you that something works so that you don´t learn from it and could just copy it.

As a rule of thumb I have the formula

Number of landing pages x number of offers x average offer payout x 10 = Test budget

That formula is only true for completely new campaigns without previous stats.
When you have a good BL for a campaign already and only need to test new offers or landers you can use

Number of landing pages x number of offers x average offer payout x 5 = Test budget

For both formulas: When you don´t see any signs of success after 50% test budget kill the test and do something else, no need to waste money.

But same as with all other rules and numbers, it´s no universal rule that can be used for each and every campaign you run.
It´s just a rough orientation that you have something to work with so that you don´t run your tests by random numbers for each new campaign.
Especially in the beginning an easy repeatable system that helps to run all campaigns in the same structured way can help alot.

I judge by clicks or CTR when I know the "normal" performance of the lander that is used in the campaign.
When I then see that specific placements have huge differences between the average performance I also cut by stuff like clicks and CTR.

But to do so you need enough stats first to be able to find significant enough patterns.

When I don´t have enough stats I don´t cut by clicks because clicks are no currency, I don´t pay traffic with clicks and I don´t get my revenue paid with clicks.
In the end the only important thing should be the money.
Cost higher than revenue = bad, cost lower than revenue = good.

Clicks are just one of many numbers in the campaign and just by themselves they don´t say anything.

They only become really important when you set them in relation with the money you spend on them.

Here is a very theoretical example just to show a bit better what I mean:

Let´s say you run an offer with $2.40 payout and you pay only $0.10 for 200 clicks.
Why would you then cut a placement just because it doesn´t convert on 200 clicks?

Maybe it converts within the next 600 clicks, then it would convert 1:800 but the adspend (assuming numbers above) would be only $0.40.

For me the performance and thus the process to exclude placements is usually directly connected to the money I spend.

I pay money for the traffic, I get my revenue paid in money so this is the variable with the highest validity about the performance.

All other metrics and numbers can only tell you about the profitability of a campaign when you see them in connection with the money.

Long story short, you can cut placements by clicks but then you should have enough stats already to know the average performance of the campaigns elements to see rather reliable patterns.

About creatives, I don´t really consider them that strong when I calculate the test budget because you often have the situation that the trafficsource will favor the creatives with highest CTR so that they will send the most traffic anyway.

When I then see that a specific creative is performing very good in my campaign but doesn´t get enough volume on the trafficsource I would test it in a separate campaign to compare what is showing better performance for me.
That way it doesn´t need to compete with creatives with higher CTR on the trafficsource and I can compare the "real" performance stats much better.

I won´t check it because you just pasted a link to the screenshot

Yesterday I just posted a very easy tutorial how to insert images properly, you can find it HERE

Ok, I checked the screenshot nonetheless but would be great when you could insert the images in the postnext time.

But...

You didn´t ask a question about the screenshot, you just say this is what your Pushground AM says about it so I am not sure what kind of reply you expect now

---

Golden

Of course it is good for everyone to find their own style @twinaxe, but its also good to model the mind of an expert to save oneself going around in circles chasing things that don't work Yet, still everyone should do their own due-diligence and find whats working for them, but just knowing how the successful do it can help a ton

Yep I was going to ask you whether you tend to optimize/kill any other misc. stuff other then placements+creatives+landers+offers

Did you mean here clicks from the traffic source or clicks through the landing pages?

The reason I have these rules setup as clicks is because they basically usually represent 1x payout in loss for me, so 130-200clicks & no conversions with the SOI's I'm currently running would usually be around enough $ in loss to kill the placement. (Since TheOptimizer doesn't offer x payout rules so far).

This is just never a ratio I get with my current offers, of course it depends on the type of offers an affiliate running, just personally for me, the SOI's are usually sufficient enough to kill after 130-200 clicks & no conversions.

The reason I have set a rule of 400 clicks & <0 ROI = kill because after that amount of traffic to a zone/creative I expect to be running only a profitable zone/creative already. Is that correct thinking?

Well now I finally figured how to place images to STM like the pros do, so I'm all set to rock&roll

Would you listen to this suggestion from Pushground's AM?

So basically what I am asking is how you treat creatives Do you cut them by x payout rules same as zones, in all campaign stages (testing, preparing to scale, scaling)?
Or do you restart/swap your creatives at any point in the testing stage, or then at the running stage of the campaign? Because sometimes all my creatives get paused while I'm still in the testing stage of the campaign.

That's a pretty good idea.

I wanted to ask you how you treat killing campaigns and how do you differentiate a dead campaign from a normal dip in the performance? Especially in the case of European SOI's lets say.
Since a push campaign can be rather unstable, I found that a campaign can be -20 to -10$ for 3 days and then suddenly jump to a 60$ profit day.

---

Sure, especially in the beginning it can help alot when you have at least some rough proven rules to work with.

But everyone is different so it´s best to understand the whole stuff ASAP so that you can develop your own style that works best for you.

In my opinion this is a very important step in the process to become a successful affiliate.

Not really, when I create a campaign I set the targeting and usually don´t touch it anymore.
I don´t care if some OS versions or browser versions are in loss because it doesn´t make a difference in a good campaign anyway.

It´s only important that the biggest elements that bring the most volume are profitable.
It happens that your volume on your whole campaign decreases when you exclude too many elements, even on the remaining targets.

And when a browser or OS version brings only 2% of the traffic it doesn´t matter if it´s profitable or -100% ROI because it has no big impact on the campaigns performance.,
Better keep it running to receive more volume for the profitable elements.

Both, but again - to find such patterns you need lots of stats already from a longer timeframe.

In my opinion it´s still better to work with payouts instead of clicks.

And talking about Optimizer, do you use it in testing stage as well?

Because this is something I don´t do.
Testing is for me all manually and I only use Optimizer on working funnels for scaling.

Sure, this ratio was just a rather extreme example to show why I prefer to work with adspend instead of clicks.

Adspend is directly connected to the payout, it´s the same unit.
Clicks become only really relevant to the performance when you connect them to adspend, otherwise they are just clicks - not more and not less.

I am working on getting the push guide part about creatives anyway, will the next one to be published.

There I write about such stuff anyway

Generally it´s not wrong what she says but I do it a bit different

When you run on higher scale the campaigns usually become more stable.

When such hefty swings happen the first thing to check is if there are new placements that send volume.

From my experience the normal behavior of a converting push campaign is rather that after some time the volume slows down.
When the performance drops alot it´s mostly because new non converting placements send high volume or there is an issue with the offer.

---

True. Get it right, kids.

Yep I set TheOptimizer right from the start basically, why not doing so?

Can't wait to read it, when is it coming out?

I just think I need a better system of determining when a campaign should be cut right now, and when to keep holding it through the reds, to avoid unnecessary losses. Since I'm not running very high volume campaigns at the moment What would you suggest doing?

For new volume I always have the usual Optimizer rules ready to take the non-profitable zones down

I just read your massive GIVEAWAY of an offer yesterday, which is great and I wanted to ask, when picking offers, are you asking the AM for recommendations or do you just look for the best combinations of good lander/good niche + payout + bids + potential volume for the targeting?

While I keep running Push, I got some more questions coming in for you you are awesome!
- How do you determine the point in which you can scale a campaign to more traffic sources? I find it sometimes hard squeezing volume out of Propeller so by the "test fast" notion I would then quickly switch to more sources after seeing solid conversions, running on all user activity already and an optimized (or mostly optimized ) funnel already. I try to keep to minimal amount of traffic sources as possible at the start to save testing budget.

- After you would scale to other traffic sources, would you start with high user activity first on the new traffic source (if you have the targeting option) and lower bids? Or should we just scale as fast as we can?

- How do you go about creating BL/WL? Do you take BL/WL only from the profitable campaigns or would you take data from campaigns that converted solidly and had green zones but didn't quite make it to being profitable?

I have a spreadsheet on which I prepared many BL/WL for different GEO's (I know you are saying combine Blacklists at the start, but I already have solid data from previous campaigns from multiple GEO's )
and I would prepare/include information for User Activity, Bids, and Date added



- Do you collect BL/WL from Low user activities as well or just from High User activities?

- When I already have Blacklists for multiple GEO's is it better to make separation between tiers (T1, T2, T3) when running to GEO's which I haven't previously collected BL data on?
Should we update our BL/WL for separate GEO's every few weeks or to keep it constant and unchanged?

- Do you add big sources (like on Zeropark, Pushground, etc) into global traffic source Blacklists?

---

@twinaxe,

What do you do when Higher user activity in the GEO you want to run doesn't have enough volume on Propeller Ads, let's say?

---

Sorry man, somehow I didn´t see your previous post.

Just do it the way it works best for you

Optimizer can´t see which landers and offers are doing good or bad so it can´t do anything there.

That´s why I do the testing all manually and use Optimizer then when I have a working funnel to kill bad placements.

Working on it, was very busy recently witn another thing.

When there is not that much volume available I would stop the campaigns earlier.

It´s always a matter of potential of the campaigns.

When there is enough volume it can be worth it to keep them running for longer because then even after some optimization there should be enough volume left to make good for the losses rather fast.

But when you lose for example $100 on a campaign that can only make $10 profit a day it would take 10 days just to get into green and then you still wouldn´t make much there.

In the end you just need to run enough campaigns to gain experience, then it becomes easier to judge such things yourself

I don´t have THAT one best way to get offers.

Sometimes I ask my AMs if they have good offers for a specific targeting that I want to run, sometimes my AMs message me and tell me they have a good offer for me.

Sometimes I just test offers from newsletters or top offer reports, sometimes I test new offers from advertisers I ran successful before.

Sometimes I just grab random offers for a targeting I want to test, sometimes I test fresh offers with no history yet to be one of the first to promote them in case they are killing it.

I would say a good part of what I test and what not depends on my experience so that I can gauge a bit if it´s worth it or not.
And with that I don´t only mean if it´s worth it in sense of if I can run a campaign with profit.
For me personally it´s only worth it when I can run a campaign with profit and enough $$$ to be worth my time

I like to scale from easiest to hardest way to do so.

That means that I first use what is proven to work already so that I have to work with least unknown variables.

So I first scale on the original source with exactly the same targeting and funnel as I run in the original campaign just with a higher bid.

Then I run same funnel and targeting with lower activity levels.

Next step would be to run broader targeting when the offer allows it, for example when I ran on mobile only I would also test on desktop.

When I see that several scaling campaigns on the original source work I scale on other sources as well.

It depends what you are willing to pay and what you are willing to risk.

Starting with high quality and scale slowly can safe lots of money when you can use the stats from your campaigns for later campaigns.

I often want to have resultst as fast as possible so that I often run different targetings, different qualities and different source all at once.

But this is probably not recommended for beginners because you can also lose few hundreds a day pretty fast with it so you should be sure that it´s worth it and that you can make the money back that fast as well.

Usually only from profitable campaigns, at least for BL.

But when I see that there are placements that are doing very good and bring in many conversions in a campaigns that is not profitable overall I would use them in a WL as well.

Much too complicated the way you do it.

Just create an excel table for each trafficsource and for push also for mobile and desktop, for example Propeller Push Mobile and Propeller Push Desktop.

In these tables use the first column for global BL and the other columns for country specific BL.

That´s all

I collect there as well, what works on low usually also works on higher quality.

But on low activity/older user freshness I often prefer to run WL with proven placements/subsources.

I don´t separate that much but I do separate between really high Tier geos and really low Tier geos like India or so.

It depends, from my experience the lists for pop traffic can last longer than for push traffic.

I could imagine that it has to do with the nature of the placements.

High volume websites on pops they can be stable converters for long time.
When you have "Click allow to continue" pages to collect push subscribers these domains can rotate pretty often so that placements also can change faster.

You mean the subsources inside ZP, Pushground, Evadav and so on?

Yes, I treat them similar to zones/placements.

When there is not that high volume on high activity it can be worth it to test on high+medium activity without risking to lose much money because of a sudden traffic boost on lower quality.

---

Thanks alot for your reply

Wouldn't you still kill zones & creatives by x payout rules in the testing phase though?

Yep I agree

Would it be nice to still compare bids for each GEO for a Blacklist? like you said once, you create "baskets" of different WLs at different bids because if a subsource is not profitable at a certain bid area it doesn't mean it can not be profitable at a different bid.

Would you consider converters for a BL or only profitable placements for a profitable campaign?

I actually didn't mean here subsources, I meant here SOURCES. Do you add them to a BL/WL?

Replying to a previous thing you said :

Why do you think it's so that the traffic source decreases the volume of an enabled source?

So basically when you want to test a certain GEO but there is not enough volume at your testing bid you would go down the "ladder of quality" but try to keep quality as high as possible to stay as efficient with testing as possible, am I right?

If you are at the testing stage and your creatives reach certain rules like 1x payout at loss, or 10x payout and no conversions (just an example of when I don't care about excluding creatives that much and I am thinking in the "twinaxe-zen mode" ((yep that's a term))), would you exclude the creatives? What if you're early on in your testing budget but you already excluded all the creatives? What to do now? Kick dirt?

Another thing, how do you deal with determining the best working funnel while being in the testing phase but duplicating a campaign to raise a bid for example or refresh creatives?

---

No, for testing phase I have very dynamic rules for cutting placements deoending on the campaigns.

What you should do is to exclude non converting high volume placements in test phase and when you have a working funnel you re-test these placements for quality.

About creatives, I don´t use "x payout" rules for them because often you can´t really predict the creatives adspend.

Creatives with high CTR receive more volume and spend more, creatives with lower CTR receive less volume and spend less.

There I rather decide case by case.

Of course you can] do it and it definitely makes sense but then you need more stats already to have enough data for

1) the different geos
2) the different bid levels inside these geos

So at the beginning I would keep it as simple as possible to work good and effective with the stats you already have.

The quesiton itself doesn´t make sense

I would never consider profitable placements for a Blacklist

But I guess I know what you mean, when a placement isn´t converting at all = 0 conversions I would put it on a BL for other campaigns/bids/activity levels as well.

When a placement is converting but not good enough to be profitable in your original campaign it can be tested again with other settings so that it´s maybe profitable then.

Sorry man, now I am confused

How shall I add whole sources to a blacklist?

Please enlighten me what exactly you mean with it

Trafficsources want to sell all traffic, not only the good traffic.

When I target for example UK, Mobile, Android I receive all traffic inside that targeting.

Now let´s say "UC Browser" for that targeting shows bad performance across all campaigns on a trafficsource.

Many people would exclude the browser then from their campaigns because it´s not converting.

This means that the trafficsource wouldn´t sell traffic with targeting UK, Mobile, Android and UC Browser because everyone excludes that browser.
In other words, the trafficsource won´t make money on that traffic.

When I the see that UC Brwoser in my campaigns is at -100% ROI but is only sending 2% of the traffic I would just keep it running because it doesn´t have big impact on my performance anyway.

The trafficsource however sees that I buy the "bad" UC Browser traffic and in return favors my campaign with more volume to the other elements.

It´s a pretty rough explanation but I hope you get what I mean.

If not, just ask

The question would be "Do I really want to run the campaign or not?".

When I really want to run it I have to deal with the available things.

The main reason why you should always test on high activity/new user freshness is the quality.
Another reason is that you can control the volume much better doing so.

We can say that during the test the combination of high volume and lower quality should be avoided.

When there is not much volume on high activity and you want to run the campign it can help to test on high+medium activity (or a bit older user freshness on other sources) so that you receive enough volume to make it worth it but still don´t risk to lose too much money because.

The combination of high volume and lower quality isn´t given there anyway so that it can work to test it on a bit broader targeting then.

Again, for creatives it´s every dynamic and can change from campaign to campaign.

I can imagine that this is probably the answer you don´t want to hear but this is really something where you just need to get some experience yourself.

The more campaigns you run the easier it will become to decide what to do with the creatives.

The problem is, when you are still testing you don´t know if the creative is that bad or if it´s the LP, the offer or the traffic itself.

Here are some approaches anyway how you can deal with creatives

- When all creatives receive more or less same volume and don´t convert I just keep them running together, when nothing works it´s probably the funnel/offer that isn´t good enough
- When all creatives receive more or less same volume and only some of them convert (several conversions, not just one) exclude the non-converters and just keep the converting ones running
- When all creatives receive more or less same volume and they all convert good keep them running together to reach statistical significance to know which ones to keep and which not.
Work from bottom to top = don´t keep the best ones running but rather eliminate the worst ones
- When you have creatives that have lower CTR and receive less traffic = less adspend and then you have another creative in the campaign that doesn´t convert but has higher CTR and more volume then exclude it when it spend much more than the other creatives, no matter if the other creatives convert or not.
When the other creatives convert it´s better to stick with them anyway but even when they don´t convert it´s better to receive traffic from few different creatives instead ofonly receiving traffic from one non converting creative.
This increases the chance to find a winner
- When you have creatives that have lower CTR and receive less traffic that don´t convert and then you have another creative in the campaign with higher CTR and more volume that converts just keep the converting one running
- When you have creatives that have lower CTR and receive less traffic that convert and then you have another creative in the campaign with higher CTR and more volume that also converts run separate campaigns.
One campaign for the lower CTR creatives and one with the high CTR creatives.
That way they don´t compete against each other in the same campaign and you can better track which is working better for you.

Never happened and probably won´t happen

First and most important task is to find the best combination of LP and offer.
That´s more important than creatives (of course only as long as the creatives are related to the funnel and not completely crap ).

So in this stage I usually don´t raise bids or duplicate campaigns.

When it then have a working combination of offers and LP and when I try to find the best creatives it can become a bit tricky.
In the tracker you can only see the creatives ID but not what the creative really is about.

I like to prepare an Excel table then with several columns for info about the converting creatives.

One column for campaign name/targeting, one column for the creative title, one column for creative description, one column with info about the icon and if a big image was used or not and one column about all creatives used in the campaign like "icon: abc and xyz, big image yes/no" or "icon: abc and xyz, big image no" and so on.

That way I can see pretty good which combinations work best.

It also helps to see if a specific element only works when there is no alternative to splittest in a specific campaign, for example when there is only the same icon in all creatives in the campaign or if the element is also working good when it´s splittested in a campaign (2-3 different icons for example).

Btw, you have great questions

---

Thanks for the Golden answers I messed up some of my previous questions but let's get to it.

What is the reason you come back to retest the high volume non converters after the funnel is working?

I think I didn't quite get you there Yes creatives with higher CTR will receive more traffic, but both of them can reach "x payout in loss"? Hopefully you can clear that up.

Got it. Keeping it simple is good

Crap, I wrote BL instead of a WL I meant to ask "Would you consider converting zones (but not very profitable) for a WL or only profitable zones from a profitable campaign?"

Sorry there

I simply meant here a whole source like "82" in Pushground, (when there are 82-256, 82-532 <-- those are subsources, which, I got that you treat subsources like normal zones on Propeller).
Of course you also have these on Zeropark, EvaDav, or other ad networks as well.
If the source is not converting, or vice versa - converting really well, do you do with them anything as far as WL or BL goes.

Yeah that makes sense I will now laugh at my friend that optimizes browsers and then complains he is not receiving as much traffic as he used to

What is the reason that you are looking for the best creatives over different bids?
Like a funnel, which you find at the best quality and then scale to other qualities to get more volume, creative is part of the funnel.
So why won't you find the best creatives at the best quality and then just scale that?

The reason I'm also asking, is because my creatives are very funny, they go all around the place, I can have 6 creatives which are all quite similar with small variations (it could sometimes even be literally the same image but with 3 dots in the title instead of 2 dots, minor changes) and the traffic I would receive from Propeller for each would be a huge difference, so I find it harder to find "the pattern" of my best performing creatives.

Adding to that and to which also relates to creatives, would you launch several campaigns in the same niche (but different offers) with the same creatives direction or even exactly the same "best performing creatives" in the same GEO? Because I can launch a campaign with a bunch of offers in a GEO and because I don't run too many GEO's (only so many GEO's you can run good adult dating, in my opinion) I feel tempted to add more tests to the same GEO with different offers or flows etc but with the same creatives and same BL, similar bid, (devices too ofc) etc just because I don't like wasting time and I want to test more.

If you got one best performing creative, will you run the campaign on just this one creative? (And lets add a condition to the question: the other creatives are all unprofitable or barely breaking even, and they are very similar variations of your main creative, just for some reason, they are unprofitable )

That sounds interesting, but I don't quite get it, similar to when I heard about Einstein's theory of relativity for the first time

---

@twinaxe Would you rather test 4 different offers for a specific GEO or 2 good offers with 2 different landers for each?

Does this look OK for Tracker numbers on Push? (As far as visits vs unique visits go.. or clicks vs unique clicks go..)

(bemob)

What daily budget would you start with for a European SOI? I know you can't give hard numbers... but can I find an approximate direction For example 1/3 of testing budget (which is Avg. offer payout x Number of landing pages x number of offers x 5 w/ a blacklist or 10 for raw test).

Btw, I missed your webinar HERE! Is there anyway to get a replay of it?
EDIT: I just bought the replays and I'm watching them now, haha.

---

Sorry dude, I somehow missed your posts.

I will reply in detail tomorrow when I am back on my PC.

---

@twinaxe Would you say bid "regions" for GEO's are pretty much the same across all traffic sources? Sources which has user freshness.. or not.

I know you say its worth testing a funnel on high user activity with a higher bid. I can accept that statement when you are running let's say FR and you bid 0.024 instead of 0.02;

How about when running on very cheap Tier 3 GEO's? Where you need to play with bids of 0.001 - 0.005 lets say.

And let's say its a network that allows only changes of 0.001 in the bid (not 0.0001)

Then for High user activity if you want to test at 0.003 (let's say BR on Evadav) but you will not get almost any volume. But then if you want to raise the bid to get more volume on high user activity, you need to raise the bid to lets say 0.005 (that's about 70% raise in bid). So for these would it be better to test at slightly lower bid as well as user activity? Maybe on all traffic?
Or what's important is to avoid that 30+ days traffic (anytime traffic) - in order to avoid too much swinging in shit traffic.

---

Note: when I say "user activity" I just mean user freshness all accross the board, I just use the term from Propeller cause that's what I got used to.

---

On which sources can we run aggressive to? Like facebook icons and stuff.

---

Ok, here we go

The reason is that you exclude non converting high volume placements during tests so that you get your traffic from more different placements.

When you receive 80%-90% or so of your traffic from only few big placements then these placements have too much impact on overall performance.

When your traffic comes from many different placements then each single placement doesn´t have big impact on the overall performance, that way it´s better to get an impression about the "average" campaign performance.

But during test stage you can´t exclude placements because of bad quality because when you don´t have a working funnel you don´t know which elements are good and which are bad.

So when you have a working funnel then you should re-test the placements to really check their quality.

Yes, both of them can reach x payout in loss.

But mostly the high CTR creatives will keep running and the lower CTR creatives won´t receive traffic anymore or only very little volume.

That´s why I don´t care about these creatives then but when I see that the low CTR creatives are converting good I would test them in a separated campaign where they don´t have to compete with high CTR creatives.

It depends, when a placement isn´t profitable at XYZ activity/freshness with a CPC of $0.xx then it also won´t be profitable with same targeting at higher bid.
Then you could test it on same targeting with lower bid.

It´s probably my personal workflow

Basically I first try to find a good offer/LP combination and when I have it I test different creatives if needed.

Creatives are sometimes a work in progress that can take some time.

It´s rather a bit of finetuning when I have a good funnel (LP/offer) already.

I don´t really check for such small changes, I rather compare combinations of title & description and icon & big image.
Additionally I make notes if I tested only with big images, only without or with and without in the campaigns.

Are we talking about stuff like running sweeps and then different offers so that creatives would only differ by product?

For example "Click here to win an iPhone 11", "Click here to win an Amazon Gift Card", "Click here to win a Samsung S10" and so on.

But anyway, just create different campaigns.

In the user decides where he clicks and where not and in best case a user can even convet on several campaigns.

Yes, when I have a strong creative I rape it left and right and sometimes run it for months without changing it at all

Ok, let me try to explain better.

When you run campaign A with only icon A and campaign B with only icon B then you can compare results which icon did a better job.

But the user had only one icon per campaign so it was shown to all users.

Now we have campaign C which split tests icon A and icon B in one campaign.
There you can see then a more direct comparison between those icons.

They were both shown in one campaign so it can give a clearer picture about the performance when one of the icons was clcked much more or converted much better.

I would test what looks promising.

When I have 4 different good offers for same product that I can run on the same lander I would test them, when I have only 2 good offers and 2 different landers for each I would also test it.

I wouldn´t decide by such metrics, I would purely decide by the potential it has

Dude, what shall I say?

So far these are only numbers, no idea if it was desktop or mobile, what geo, what TS and all such stuff.

It´s impossile to tell if it looks ok or not but the unique clicks seem to be far off for some campaigns.

But again, hard to tell what´s going on there just from the numbers.

You already have the formula, great.

Then just try to split the total budget across 2 or 3 days to get a better impression about the performance when it runs for a bit longer.

Cool, I hope you can learn a bit there.

I hope I didn´t fuck it up completely, I was really nervous when we did it because it was a completely new situation for me

Well, when we talk about bid ranges from $0.03-$0.010 or so there´s not so much room anyway so of course they are somewhat similar.

But many sources have their own ways to bets work with them so I would rather decide it source by source.

I know the problem with minimum bids for lower Tier countries too good

Nonetheless I would still run it with $0.001 higher at high activity/freshness compared to lower bid on lower activity/freshness.

Unluckily I can´t tell any specific sources because I don´t know if they would be ok with it but just ask your managers.

On some sources you can even run super aggressive when they have specific channels for it

---

������ What a great thread! Thanks guys for sharing it!

I would try to give my 5 cents based on our experience as a traffic source and as a mediabuyer.

For huge countries we recommend to start with low freshness, 0 to 2 days, as @twinaxe said, it's a good way to not only have better quality but also better control over the budget. About the bid, it seems logical to think that lower freshness will be more expensive, the truth is that it depends on the supplier and the existing demand, at pushground we do provide the win rate by user_freshness so you can figure out quickly if you are or are not overbidding.

We have seen that user_freshness is one of the most relevant metrics that boost the quality (not just conversions). Having said that, once you have a winner combination (creativity-prelander-offer), we strongly recommend you to open it to more user_freshness.

Offer is king but we have seen campaigns increase their ROI by 100% by just changing the angle of the creativities. The recommendation that Gina gave you is right, but allow me to go a little bit deeper. We at Pushground send the same weight of traffic to all the creativities when you start a new campaign, once we start receiving conversions, system starts learning and then gives different weights to each of them but never pausing them completely.

We recommend to pause them manually when you see the system is allocating less than 10% of the traffic available, it means that our algorithm considered this creativity a Zombie.

Having said that, if you copy from other ads or if you use always the same creativities and just cloning them, your creativities will get delivered less and less. Push publishers normally hash the texts/icons/images in other to identify and prevent duplicates thus not burn their users.

We have seen a lot of success with the following rules: (1 offer, 2-3 prelanders, 5 creativities)

If payout < $3,
pause campaign if 20*payout and ROI -30%
pause supply if 10*payout and ROI -30%
pause site_id if 5*payout and ROI -30%
If payout > $3, same rules, just divide it by 2.

After this, we recommend starting micro bidding automatically at supply and user freshness level.

About blacklists and whitelists, in push we have seen that it's good to keep an eye on the site_ids that can ruin your campaign, but more importantly the supplies. There can be big performances difference between them. Because of this, we have tried to simplify this at Pushground by adding another quality layer which allows to set your campaigns "to start" or "to scale". This essentially acts as a premade global whitelist to allow you to easily test new campaigns on a smaller selection of higher quality traffic.

As @twinaxe said, normally bad feeds are bad in general, and good feeds are good.

I hope this information provides some extra insight.

---

Boom Boom Boom!

Thanks for the answers man

I meant like one creative ID, where that is the creative all traffic goes to. Would you keep that campaign running on just this one creative? (given that all other creatives were unprofitable or slightly break-even).


I meant the ratios of unique visits to non unique vists, unique clicks and so on. If you notice my unique visits are double the visits in some campaigns. I wonder if you noticed that this is normal?

Here are some better screenshots:




Notice the differences for Italy across different Traffic Sources:



I guess the nervousness helped you bring out the nuggets in yourself, since it was very info-providing and valuable

I actually meant all GEOs, FR, DE, UK, US, IT, ES and the list goes on

I got your point, but in the example I provided (and similar cases), the bid you set at higher activity doesn't get you almost any volume, I'm talking about a potential daily spend of 2-3$
(and this is dating = higher CTR)

Good. More findings for me to find (I mean it)




@pushground Thanks for the love guys.

How big of a change was it for the angle and what kind of verticals? I find this hard to imagine for verticals like dating & sweeps.

This is a very good feature I have a problem with DZ and EG geo's which I need a super low bid but the bid I've set should be fine according to the % of traffic & winrate calculations, like I sould receive 100x fold traffic of what I'm currently getting. I tried solving this problem with the awesome Gina but no use. Still the same issue?

Is this built into the algorithm? Or how do you decide that the creativities will get delivered less and less?



Thanks for all the love guys.

---

In the dating case, an example of two different angles would be: nice impossible teen girls and a mature divorced mums, we have seen better results on the second.

If you have already a high win rate by each of the supplies then look on your targeting, maybe you are targeting to narrow.
An important point is the frequency, when you are selecting once a day, you are limiting your traffic about 80%. Lower frequency normally means better quality but it worth to give it a try. Sometimes you need to fill the user screen to get his attention

I guess each traffic source has different approaches, but we know it's a common practice to not burn the users with repeated ads.

---

Great that you speak about dating vertical, maybe it´s good to mention that especially dating can be somewhat different to many other verticals when it comes to the push creatives.

For other verticals it´s mostly a good and catchy title to grab attention and a good dscription to get the user to click through.

But dating is a bit different, there it´s all about the images.
Dating is very visual so there I would focud more on the image that you use.

I often run push campaigns with icon only, without a big image, and it´s working good for most verticals.

But for dating and stuff like browser games I see much better results with big images because such verticals are more visual than others.

And I am with @pushground, I would also try to work with rather mature women compared to teen girls.

Maybe you can get higher CTR and even higher CVR on the offer with teen girls but mature women often result in better quality for the advertiser, this can be a leverage for higher caps and better payouts and can prevent from getting kicked from offers.

---

I agree on that man I recently started going more towards this angle in dating In the screenshots I've sent above its mostly the "young teens" style though.
Any opinions on my questions??

---

This is a great thread, packed full of info! I've enjoyed reading it.

One question though:
@pushground

So you train your media buyers to cut placements at 5 x Payout for offers under $3. That's interesting to me, the most I've heard of before is 2-3 x payout. Could you explain more why you do this? would this be different in any way because you are a traffic source and get cheaper traffic?

I'd love to know more about how you train your internal media buyers (guess it might be private, but no harm in asking!) hehe.

Thanks, Gav

---

Sure I have an opinion

It rather looks like it´s the other way round and the visits are much higher than the Unique Visits in some campaigns.

And yes, I know that you meant the ratio but still hard to tell anything.

I don´t know what is considered as a unique visit/unique click by Voluum.
Is it unique cookie per 24 hours?
Because I think this is how Binom is doing it.

But to be honest, I don´t use it unique visits in my stats so I probably can´t really tell much about it anyway.

What´s your click freq cap per 24 hours?

Because when unique is counted as unique cookie per 24 hours and you would have a click cap higher 1 per 24 hours and a user clicks twice or so on creatives in that campaign then such stats could probably happen.

Maybe it would be good when @Voluum could step in here and enlighten us about it

Sure, when this creative is doing best the of course I would keep the campaign running with only that creative.

---

It's actually Bemob And my freq/cap is 1/24 for clicks Haha do you think its possible to get refunds from TS for charges on clicks that didn't meet this restriction?




Have you noticed that volume is lower/CR is lower on Push when you're running on only one creative ID?

--

After you will find your winning creative, will your optimization stage/scaling campaigns (Yes I've got the structured approach from the Webinar ) have small (5-6) variations of that creative in the campaign just to try and improve the performance of it a little bit more?

--

Would say that bid regions for the popular GEO's on push are similar or different across different Traffic Sources (TS that offer User Activity targeting or the ones that don't for example as well)? Like would you usually set the bid the same across different traffic sources for the same campaign or do you find it can change a lot?

---

Oh, looks similar to Voluum

Really man, I don´t really think that this is the reason then.
At least not on such scale.

It was just a try to explain what could be happening from a technical view.

Apart from that it´s more or less impossible to really tell how these numbers could happen.

No, a good creative is a good creative is a good creative

Not really, I am not looking to improve the performance just a little bit more.

Then I rather spend my time on tasks that could have much higher impact on performance/profit.

Again, often the CPC range I use is only few cents so the bids are naturally somewhat similar.

Nonetheless I never take the bid from trafficsource A and assume that it´s working the same on trafficsource B.

I decide from source to source and try to find what is working best there.

---

So personally you don't look at unique visits/clicks at all?

---

Correct, I don´t look at it at all

---

In Voluum, a visit is considered to be unique when a visitor activates a campaign URL for the first time or activates it again but a cookie related to the original visits has died out (a lifetime of a cookie is 24 hours).

BeMob might be counting uniqueness differently - it's best you check this with them. I know of a couple of cases where their unique visits/visits stats looked very suspicious and in no way explainable - maybe it has something to do with the logic behind this metric?


Karolina

---

I see, it´s basically the same as in Binom.
@roiter123, you could also try to find out if BeMob tracks unique visits on a campaign base or for the tracker in total.

You run several IT campaigns on Propeller, when you use the same targeting it can happen that 1 user clicks your creatives from several campaign.

When the tracker then sees uniques for the whole tracker and not by campaigns these users would also seen as non-unique.

Again, I don´t know if that´s the reason, just trying to brainstorm a bit about your stats

---

Thanks a lot I'll check what's up.

Twinaxe what is the reason you target only native + english language on Push? I saw you say that on your webinar.

---

Because language targeting can increase quality.

---

Thanks for the answers I take mental notes
You are against changing bids in a running campaign, now if you were to start a campaign and your overall initial test budget is 100$, daily budget is around 30$, you start the camp and the first day gets 10$ spent, would you just raise the bid on the same campaign or what would you do?

Btw the above numbers is an example of a sweeps camp I launched yesterday in the US, I took mobile - high activity only with a bid of 0.036$, however I did take a BL from an FR camp.

---

That seems quite low isn't it?

Below is the traffic chart after the BL and all targeting applied & all other information, current CTR is at 0.5% (reminder: this is High user activity, for creatives I used standard sweeps creatives like @twinaxe used in his sweeps webinar)









The creatives I used (there are 4 in total):




Ask about the reason of low volume from PropellerAds, and they will give you all kinds of reasons that don't help you run the campaign profitable

---

Update: 2nd day spent only 3.6$

---

Yes, you could try to increase the bid.

For US Mob there usually should be enough traffic available.

But don´t forget, the bid is only one thing.

It´s also important to have a high enough creative CTR to compete with other people.

Otherwise your creatives won´t receive enough traffic = low volume.

About your creatives:

I would use a different icon in the first one, there are better message icons out there.
And when you make such a red circle it´s also better to put a "1" in it so that it looks like it´s 1 new message.

The image also doesn´t look that attractive for me.

For phone sweeps images of a hand holding the phone often worked good for me.

Then there is no real CTA in the description text.

And although I use emojis in my creatives as well I always try to use approptiate emojis for what I promote.
I would never use a smily with hear eyes for a sweep offer where I tell the user that hiw new iPhone has arrived.

Such emojis are better for dating.

The second creative doesn´t look that bad but it doesn´t tell anything about what you promote.

You could also try (1) New Message
Click here to win an iPhone 11

or

You have been selected
Click here to win an iPhone 11

Oh, and I always put the emojis after the text.

I don´t know if it has any effect if they are in fron or after the text, never really tested it.
But that´s how I do it

In the end it´s all about testing.

A higher bid won´t result in higher volume when creative CTR is too low.

---

Actually the other part of my creatives were like you said



Still the volume is very low, can you suggest me?

---

First test a bit different creatives to try to increase CTR.

If that doesn´t help try to increase the bid.

---

Refreshed the creatives and gave them "a spoiler and a twin-turbo", but I am low on faith.

Really, based on the Traffic chart it seems like it should have much much much more volume. 3$ daily spent? really?

---

True, it´s indeed very low.

But you shouldn´t forget, US is also one of the most competitive geos to run.

You have to fight there with many really experienced affiliates, with high volume advertisers and stuff.

And the numbers that you see in the estimator are not the traffic that is available for you, it´s the total traffic that then is split between all the different people who run that traffic.

Nonetheless you should be able to spend more

---

Hey @twinaxe While I'm sorting this shit with the low traffic volume, I wonder if you pass tokens of which traffic source you're running on to the advertisers/networks?

I know it can be very good for optimizing the offer for quality with the "masters of the universe" (AKA the offer owners/affiliate managers).

I've been doing that shit for almost every offer I run from the very beginning and it can be quite counter-productive. So just wanted to hear your take on it

EDIT- An example if I'm not clear : lets say I'm running on Propeller then S1=prop, if I'm then also running on EvaDav then S1=evadav. Maybe I'm doing that shit very non productive so-to-speak.

---

UPDATE: I improved the creatives and raised the bid to 0.04, I still got a poor 5$ spent yesterday Maybe propeller jinxed my traffic?

On the other hand, I took the same campaign and copied it 1 to 1 to EvaDav (although already optimized to 2 converting offers from 4), - Blew 20$ easily without a single SOI conversion Zones are looking no way near optimization and so for a test stage campaign they look fine on EvaDav (getting traffic from as many zones as I can and not from 1 - 2 sources):


However, worth mentioning that on Evadav I ran the test campaign without a Blacklist, though I kept the bid of 0.036. For Propeller like I've mentioned I raised the bid to 0.04 (in the same campaign ID, without duplicating), how high can you go for SOI sweeps?
Yes, on Evadav my user freshness was also 1-3 days (which equals to "High" on Propeller).

---

When I did run camps in US I had similar situation.

Not sure why but bigger BL = no traffic from Propeller. I think if you'd remove bad placements you'd get much more traffic but of course it's riskier and more likely campaign won't be profitable. But I experienced the same situation on Propeller when I was running US. It's hard geo to run on Propeller, at least for me.

---

Why can it be counter productive?

Btw, I never send the trafficsources clear names, I use the ID they have in my tracker for it.

Usually I send it to all offers, for most networks I have it in the URL template in my tracker anyway.

Here is an example: &s1={trafficsource}&s2={clickid}&s3={campaign}

This means that I send the trafficsource ID in s1 parameter and my trackers campaign ID in s3 parameter.
I like to send the campaign ID additional to the TS in case that I run several campaigns for the same offer on a trafficsource.

In the end I don´t see a problem or negative consequences doing so.

When I only run low volume to the offer it not from interest anyway, when I run high volume I try to work closer with the advertisers/networks anyway so it can help for a better and longer lasting cooperation.

Good possible, I also heard that sometimes there are black magic masters and evil sourcerers doing their work

Well, seems that the issue then is rather on Propellers side.

Hard to tell why exactly you have such low volume there, maybe a mix of hard competiton, too low CTR and bid.

As high as it´s still profitable.

Better try it with 1 day freshness first.

---

Really man, my creative CTR is so poor, I don't know what can I do with it. Yes I launched the US campaign on Evadav too as well as on Pushground, Result:

Propeller = 0.5-0.7% CTR (1-3 days user freshness)
Evadav = 0.15% CTR (1-3 days user freshness)
Pushground = 0.15% CTR (1-2 days user freshness)

Bid is the same 0.036 for all traffic sources.

Result:
Propeller: very low spent with initial promising results only at the beginning.
Evadav: High spent but -80% ROI, seems like still despite my poor ecpm (from poor CTR) they are sending me traffic but its shit shit traffic. I know you said good offers will convert on shit traffic but maybe this traffic is very - shit (said in Russian accent)?
Pushground: very low spent, similar to propeller but even lower CTR.

Initial good conversion but further poor performance can indicate about bad position in auction? Therefore bad CTR, or bad bid.

I am VERY NEW to SWEEPSTAKES so I have no ideas how to improve my CTR, I mean, my creatives are all the standard stuff affiliates are running (what I mentioned above).

---

@pushground

Can you enable Payoneer for me

---

@roiter123 Shall we still work in this thread on your campaigns or shall we completely switch over to your new FA?

---

Switch it over. This thread is only for reference now.

---

Home > Paid Traffic Sources > Push Traffic