Home > POP / PPV / Redirect > Follow-along Campaigns

---

Journey to profit (53)

---

Hi everyone

I'm new to affiliate, and started half a year ago, I have run SOI on push, but the result is not good, also the budget is not enough to run. I paused my campaign, and now I'm back, I want to start from pop, I can get more data under the same cost compared with push traffic. (sorry for my bad english, correct me if I'm wrong)

My goal is test out a strategy to run campaign to profit which can support long terms.

basic info:

geo: IT (I have been targed on this geo all the time, the volume is big. but when I start this journey I'm thinking if I need to switch to another geo which competition)
tracker: bemob
aff.Network: clickdealer
landpageing server: gcp 1core 2G located at Frankfurt
CDN: cloudflare
Traffic source: Adsterra

Day1 12.19-12.20

offer: 2.4$, shopping card, soi
traffic: popunder
cost: 55$
revenue: 24$
roi: -56%
bid cpm: 0.5$, on traffic source, the minimum is 0.35$, recommend is 0.52 (Q: should I start with a lower )
landing page: 1 (I have this one on push, it has conversion, I wan to try, but looks like I shuld add more, so in the day 2, I add another one)

I setup two campaign, one for desktop, and one for mobile.
it has three offer, offer 1 is the target offer, offer 2 is back redirect to a push collect page and then go to a smartlink , offer 3 is bot test

• campaign 1:
  

it is desktop targed


the conversion(new) is event when user scroll the page.
looks like bot rate is about 34%
the ctr is a bit low 0.12%, the scroll rate is about 0.78%

• campaign 2:

it is mobile targed


bot rate is about 24%
the ctr is lower than desktop, conversion is less than desktop. (looks like desktop conversion better, so I paused mobile traffic)

Question:

• 1. should I splite desktop and mobile traffi at traffic source and tracker, I have two two campaigns both at traffic source and tracker，I've just heard that desktop and mobile bid price is different

• 2. the tr is a bit low, is this normal at pop traffic, I'm think if my landing page or nginx server has some issue

• 3. on my lading page server, it is just a static server, nginx hold the html , css and js file, is that ok. And below is cdn statistic data

• 4. the daily cost is exceeded my budget, and I only run the campaign from Italy's local time 7PM-12PM, so this campaign didn't cover the whole day, I'll struggling to switch to another geo, I havn't swith to another geo since I run campaign.

• 5. I find my back redirect not work well, for push collection, it bulk scrribe to onesignal an another traffic platform, but looks the scribe number is about 20.

• 6. there is about 20% data loss between traffic source and bemob, on the traffic source, it shows 11152 impressions , while bemob only has 88727 views, about 20% data loss, is this normal?

because Adstrra didn't support dedicated impress url to track, they use landing to track impression, so the campaign cost mod is CPV, 0.0005$ per view.
I turn to bemob support, they said it has many factors can cause this, maybe it's popunder, user close the page befor bemob tracked.
I read anoter thread, he said I can add an alert to page to attact user, before they close the page


Day2

today, I run above campaign 1 and campaign 2 for 1 hour, didn't see converion, and I paused the two.
I creat a new CPA campaign and add a new landing page to see if it can works, on Adsterra, it needs a verification stage, and the min budget is 50$, and I set price to 1.5$ per converison, which is 62.5% of payout.
This is the status


The ctr is still low, and I splite back redirect to smart link, not go to push scribe to see if the back redirect works.
I havn't run CPA model, the roi green, but the volume of traffic is less than yesterday's campaign. And the traffic campaign stage is test, I don't know what will happen after test stage is finished.

TODO:
1. I need to add an exit pop to see if it can increase the ctr.
2. Check if need to switch another geo.
3. finding more landing pages.

could you please give me some advice, I feel a little discouraged, but I have to carry on, I true believe this can be profitable, and I have inspried by jaybot's FA.



https://imgur.com/H4kZVpH
https://imgur.com/H4kZVpH

---

Hey Peng, great to see your follow along

About Geos, IT is very competitive and rather expensive.
Also you want to run SOI offers but in IT there are many good CC submits available with much higher payouts.
So in this Geo you have to compete against other affiliates who promote CC submits and often can bid much higher on the good placements.

This can be little bit challenging.
But -54% without any optimization isn´t that bad, especially when there is enough room for optimization.

Best is to start with a bid just a bit higher than recommended.
That way you don´t overspend but already receive decent quality.

For testing it can also help to run on another source that supports SmartCPM (and also cost tracking).
Testing on SmartCPM can save you lots of money compared to running on fixed CPM.

What are the 403 and 934 Conversions (new) in your tracker?

Yes, you definitely should split test it.
Bids can be pretty different, CTR as well.

Better don´t mix it and create separate campaigns.
On mobile you can even try to splittest between carrier and WiFi trraffic.

Don´t worry, low CTR on pops is normal.
Especially in high Tier Geos.

Yes, for HTML/JS/CSS you only need a normal webserver like Nginx.

What´s the question?

Same as before, what´s the question?
Why the subscriber rate is too low or something else?

Yes, 20% is pretty normal for pop traffic.
On pops you generally have a higher trafficloss compared to other traffic types.

Don´t run your campaigns that way to start and if you see no conversions in 1 or 2 hours you stop them.
Before you start the campaigns make a battle plan

Calculate how much budget you need, if possible split the budget over two days, create the campaign and let it run.
When after 50% of test budget there is absolutely 0 success stop it.

But 1-2 hours usually is way too early to decide anything.
Performance can fluctuate alot between different daytimes or days.

That´s why it´s always better to let campaigns run over a longer timeframe to get a better impression about the average performance.

---

Hi Sebastian

Your reply really motive me and give me confidence to start from pop.

I forget this important things, I'll split test carrier and wifi.

403 and 934 is the number of user scroll on the landingpage. When the user scroll landing page, I postback it to tracker, I can calculate how many user have scroll activity on the landing page, do you think will it be helpful to judget a placement? I just postback it and havn't use it.

this is my campaign only run 5 hours, it didn't cover the whole day, so I can't judget if a placement is good or bad, as the ctr is different at different day time. if I have to get all day's ctr data(at least two days), it will execeed my budget, because Italy's traffic volume is huge. So I am a little hesitant to continue this offer, but you said it has enough room for optimization, I think I can continue to optimize it, at least 4 days.
And from my push campaign's experience, I noticed from 6PM to 11PM, it has more traffic, and converions for mobile user. So I want to test mobile pop at this period to save budget, do you think if it feasible. And for desktop user, I'll test run from 9AM to 12PM to figure out what is the best period.

Yes, I find the subscriber rate is too low, from the tracker, it has 1367 click, the push collection page is a bit aggressive, it will redict 5 times to ask user to subscribe, and then go to smartlink.
do you have any idea about this, I can't figure out the reason, so today I split half of the back redirect traffic directly to smartlink.

---

Day3:

Adsterra cpa campaign:
this campaign is still in test status. I splite another smart link to back redirect traffic, but no luck, 0 conversions. And I learned on things, in the bemob, multiple offer must have the same Affiliate network when enable multiple postback, otherwise, the conversion will mixed, like below:



And below is the total 3 days overiew



this campaign is still in test status in Adsterra, because the test budget is 50$.
looks like CPA is a safte way to test, but for now it isn't scallable, the traffic is not to much compared with cpm mode, I'll wait what will happen when the test stage is finished.

I restart one push campaign (the same offer, landing page), just because that account has some balance, it is a white list, and only have one placement, so can't be scalled. And from my experience, it convert well at night, convert bad at day time, below is continous status from Saturday 6PM to Sunday 6PM





during this period the roi is from +30% to -13% and 0%.
I export bemob raw data to a dashboard so it can be visualized.

the conversions are distributed over a certain period of time， is this will matter in a scallable campaign?


totaly, the result is not good, the next things is cut down the placement,
I find Amy has a good article about this, https://stmforum.com/forum/showthrea...Banners-Part-2 and https://stmforum.com/forum/showthrea...Rules-of-Thumb

Let the cpa campaign runing to get more data, check above links and learn how to cut down and make a clear rule and write script to automatically.

---

Day 4:

buy traffic from propellerads, create two SmartCPM campaign, splite carrier and wifi, yes, the recommend price is different. And propellerads policy is strict, they didn't allow icon for the offer, I have to cut down two landingpages. the roi is bad.



and CPA campaign at Adsterra is still in test stage, have one conversion, but there isn't too much traffic, I guess it's maybe I need to increase the bid price.

And back redirect traffic to monetizer, got 50 subscription.



for how to cut down placement, Amy's fomula is complex, I don't know how to translate to code, can I just black a placement based on cost = half/payout, when visits number reach 1000?





s

---

get a christmas gift from cdn, landing page url has received a phishing report, and one of the landingpage is blocked, lucky, only one of the the url is blocked, after checked url at https://transparencyreport.google.co...wsing/overview

thanks this thread: https://stmforum.com/forum/showthrea...gged-by-google.

---

Hey Peng, I didn´t forget the thread.

It´s just little bit chaotic right now because of christmas

I will reply with more details when I have some rest.

---

That sucks. Happens often, but kinda randomly.

If it were just the CDN, remove the page (or copy paste it to a new folder), tell them 'Thanks for bringing this to my attention. It's standard lead gen form, but it seems there was a script on the page that we were unaware of. We have removed the page entirely and will continue to monitor the situation' Or something to make them happy.

You may want to check your page against https://sitecheck.sucuri.net/ and see if anyone else is blacklisting your site. If so, your camps will get flagged and paused soon on most sources (propeller, ZP, ExoClick). If that happens, the easiest thing to do is simply buy a new domain and point everything, or move everything over. If you don't want to do that, as long as you don't have a google flag, you can contact most of the sites (e.g. Norton) and give them the same message as above and they should review and remove their flags.

Good luck!

---

Thanks for sharing how to reply to cdn, I havn't meet this before.
https://sitecheck.sucuri.net/ is helpful, I've add it to bookmark.

---

Just a question, did you receive an email or so from your CDN service about it?

Because that never happened to me and I run pretty high volume.

Of course I often had domains flagged but I never received an email from CDN about it.

Forget this, it's not worth your time.
I always buy cheap domains in bulk.
They cost only $0.80 or so.

Then I already add them to my CDN and DNS so that DNS propagation has finished and that are ready when I need them.

When then a domain gets flagged it only takes few minutes to change the domainand continue with campaigns.

---

Yes, I got email from cloudflare, they ask me to remove the page, and they will review again.

for domain name, will it be better to use .com domain? I find that .xyz domain is the most cheap, and can save a lot cost if buy multiple domains.

---

I've got complaints from S3 before so it does happen. From what I can guess, it comes from an actual complaint, not a bot doing site checks.

What's worse is that this can affect the whole chain as the complaint will also get forwarded/doubled to your sources. So keep an eye out.

As for domain name, I do not notice a big difference. The only thing is some sources and publishers will block anything that is not a .com domain. You can see which pubs these are on ExoClick. but I honestly don't bother, .xyz .club .site should all be fine.

I only use a .com domain for FB camps for superstitious reasons (and anything you're running on FB should be as clean as possible, no scripts or anything so it shouldn't get flagged anyway)

---

Day5

Adsterra cpa campaign is active, and I increase the bid price, 80% of payout, want to get more traffic, but todya morning I find I forget add budget to the campaing.

create a cpm campaign, target mobile, wifi. 20$ daily budget, start the campaign at 7pm italy local time. roi is -64%


pause propellerads campaign, need to remove logo in the landing page, I have no experience of PS, need to find some similiar image to replace it.
I need to try more traffic source, so today register admaven.

---

Day 6

yesterday, Adsterra cpa campaign has few traffic, due to I set a wrong total budget, no converison.
another mobile wifi cpm campaign, roi is too bad.



from the above table, looks like I should block these top placement.

Now, I'm write a tool to retrieve every placement's metric,(visit, click, ctr,....), and try to auto add these placement to blocklist at traffic source platform. Hope I can finish it today.

---

Day7

yesterday, I spend some time to write that tool, and until now, it can basically work.
So I didn't pay attention to the campaign, just let it run, and get data.
3 campaign, 1 cpa, 1 mobile wife, and 1 desktop, from 6PM to 12PM at target geo local. All conversion is from desktop, mobile wifi has no conversion.



About this tool, it now can update blocklist for Adsterra Campaign, my idea is auto add some placement to blocklist by fetch it's cost, conversion from bemob, basically, it can work now.
But I can't figure out an rule. What I think now is block it when no conversion and cost reach half of payout. I'll test this rule today to see if it works. I don't know if I'm go to the right direction, but maybe it deserves to test and test.

---

I see that your CTR is very high.

70.46% on desktop is very unusual.
Do you have some scripts running on the lander that could increase the CTR?
Or maybe your click URL on your backbutton?

I just ask because this could make it harder to interpret the stats.

About rules, unluckily there are no strict rules that you can use for all campaigns.

There can be too many exceptions to have just one general rule.

But especially as a beginner it can help alot to work with some rules of thumbs so that you have something to rely on.

Later when you have more experience you often find your own style how to run campaigns.

So what I recommend is to have one rule to use for calculating the campaign test budget and then another rule for blacklisting placements in a working campaign.

As a rule of thumb for new test campaigns where you don´t have any stats available I like to use

Number of landing pages X Number of offers X Average offer payout X 10 = Test budget

Example:

3 LPs x 4 offers x $0.80 average payout x 10 = $96 test budget.

When I don´t see signs of success after 50% of the test budget I stop the test and better test something else because then it´s mostly not worth it.

About rules for placements.

You can only blacklist placements for bad quality when you have a working and converting campaign.
When you are still testing campaigns you never really know if the placement is bad or the lander or the offer.

In the test stage blacklist non converting placements when they spend most of the test budget.
At this time you don´t blacklist them because they are bad, you only blacklist them because they send much volume and spend much of your test budget.

For testing it´s not good when you have some very big placements that have a big impact on the campaign performance.
It´s better to receive more diversified traffic from many different placements so that a single placement doesn´t have such big impact.

So in this stage you don´t need a rule like 1/2 payout or 1 payout, you just blacklist when they send too much volume and when you have a working funnel you retest these placements.

In working campaigns I prefer rules like 1 payout in loss -> blacklist.

---

Yes, it's not the actuall ctr, it's bot test ctr, so there are about 30% bot.

for this general formula, do you mean test multiple offers (even they are different verticals) parallel and find which has the potential to profit? if my understand correct, I pick up 3 offers from Aff.network, payout is $0.80, and for each offer, I prepare 3 landers, so the total test budget for these 3 offers will be: 3 (LPs) X 3 (offers) X $0.80 X 10 = $72 test budget, cost at one traffic source.

Thanks for this suggestion, I completely agree.


This is a good idea, I noticed usually top 5 placements spend 70% of the total cost. It just show top 5 placement's performance, not the traffic source's performance!

---

This is gold.

I want to go over all my camps and redo all my BL now

---

Doing good so far, keep it up!
Is that grafana you are using to display the statiscis? That's some food for thought.
Is there an API you are pluging it into for real time data or exporting the results?

---

Yes, it is grafana, and bemob has api, so I just call the api, and write the data to influxdb, then create dashboard at grafana.

---

Day8

I test that auto block tool yesterday, roi is still red as usuall



just ignore the ctr, it's bot test ctr.

I tested auto block bad performance tool, I worked and block some placements, the traffic is distrubed better than before, but there are too many placements, it needs spend more many to verify the performance. Also I noticed desktop user is not activity then mobile user at night, I start from 7PM, set $50 to desktop campaign, $20 to mobile campaign, bid price is a bit higher than recommend.
mobile campaign run for less than 3 hours then stoped




Below is the placement blocked, but as you can see , there are 14 pages, 50*14=700, it's hard to test all of them, I'd better to switch a smal geo in order to test placement's performance.

---

Day 9

Today no update for campaign, I paused all, and looking for new offers and geo to test. And learned one important from twinaxe

This is the thing I need to think about and action.

---

Hey Peng,

Yes, in my opinion one of the most important things is to don't waste more energy and time than needed.

This also means to work as effective as possible, spend your energy on finding stuff that works instead of trying to do everything perfect from the beginning.

But I also know that this can be pretty complicated in the beginning when basically everything you do seems to be important.

---

D10-D14

Not too much update these days, I get 3 offers for NL and PL, get 7 landers, speend a lot of time to clean it.
And test one offer for NL yesterday, payout is $3.
splite two campaigns on PropelleraAds: mobile 3g and mobile wifi:



create 3 campaigns on Adsterra: desktop, mobile 3g, mobile wifi:



too bad.
Today I'll get another landing ready, and test both of them to identify the offer.

---

I read jaybot and heikoo FA, and find my campaing's CTR is too low, their campaign's CTR usually greater than 5% while my camp is less then 1%





Amy said, offer, lander, traffic, are the 3 main campaign elements. Am I missing some thing cause this low CTR?

---- Update ---
After check the metric from tracker, for yesterday's camps, I didn't select language, only 68% is Dutch, also 6 of 7 conversion are from Dutch language browser.

---

You can´t compare apples and oranges.

You could only compare when you run exactle the same landers on same infrastructure on same trafficsources with exactly the same targeting and bids.
And even then you can have a difference in performance.

For example mostly Tier1/2 geos have lower CTR than Tier3/4 geos.
Desktop mostly has lower CTR than mobile.
Questionnaires often have lower CTR than spin and win.
And so on and so forth.

You can use other peoples campaigns as a benchmark when it´s the same stuff.

But there are so many variables involved that can lead to different performance or CTR.

And for example when you run SOI sweeps with questionnairews in DE on desktop pop traffic than 0.5%-1% CTR is rather normal.
Especially when the campaign is not optimized and the bad placements are not removed.

---

Day15:

run two campaigns on propellerads with 1 offer.



it start from 6:00AM to 12:00PM, and price mode is SmartCPM with recommend bid, ad delivery mode is distributed. looks like distributed mode can save money, and get the average performance of placement. Please correct me if I'm wrong for this setting.

when testing multiple offers on the same traffic source, I have two ideas, but don't know are they make sense? what would you prefer if the offers are in the same vertical

1.


2.



While clean landingpages yesterday, I find back redirect code will be blocked by symantec on windows, and a workaround is obfuscator that part.

---

Day16

Not too much update, I was interrupted by other stuff, and not get the change to update landingpage.
Yesterday I make a mistake, mobile wifi campaing on traffic reached to daily budget, and it stop at 5PM at geo localtime.



And today I'll continue gather data, and add another offer to current campaign.

---

The 2 ways to do this, that make the most sense to me are:

Method 1)

Campaign 1 = targeting Desktop, rotating all offers that accept desktop traffic

Campaign 2 = targeting Wifi, rotating all offers that accept wifi traffic

Campaign 3 = targeting 3g, rotating all offers that accept 3g traffic

This would be the preferred way, as this way you avoid overbidding/underbidding - as desktop/wifi/3g traffic all cost different.

Method 2)

Single campaign: Campaign 1 = Desktop + Wifi + 3g

With tracker rules to direct each type of traffic to offers that accept that type of traffic. If you're using Voluum for example, you can set up PATHs in this way...

Path 1 - IF Desktop, go to offers that accept desktop

Path 2 - IF Wifi, go to offers that accept wifi

Path 3 - IF 3g, go to offers that accept 3g

By targeting all devices in one campaign, you'd be overbidding and/or underbidding. However, on some traffic sources, when you target broad you may see more traffic volume and/or at lower prices.


Hope that helps!




Amy

---

Hi Amy,

Thanks for your detail guide, I'll test mulitiple offer by method 1.

---

Day 17

Yesterday, still red roi



the placement seems are evenly distributed than before, I'll keep it runing and get more data then optimise the campaing.



And I get another offer which seems works in another geo, I'll test tomorrow.

---

Tody I'm tesing an offer without landingpage, because it already have one.
I have a query about track direct link's ctr, I can only get visit number, can't get the click number, below is the flow:



And this is tracker's data:



It seems I can only optimize the campaign by conversion data, other than ctr and other metrics.
Could anyone give me some suggestion? Thanks a lot.

---

Day18



only 4 convertions, can't create a BL basedon this, but I found most of the conversions is happend 5PM to 11PM, so today I only run campaign at this period, hope tomorrow will get more data.

---

Hey there, sorry for the late reply.

Yes, you can´t get the click number because in such case there are no clicks in your own "funnel".

You can only "track" clicks in your tracker that happen on your own pre landers.

As soon as the user gets redirected to the offer URL they leave your tracking.
And when you direct link you still send the user to the offer URL, even when he then first gets redirected to a pre lander from the offer itself.
So you can´t really see what happens on the offers own pre landers, the CTR and stuff.

You can only see if a conversions happened or not.

Exactly, you can´t optimize by CTR.

I also wouldn´t spend too much time to drill down to stuff like browser versions or so.

Such campaigns you basically only optimize by placements.

Sorry but what are "most" conversions out of only 4 conversions?

I know what you want to do but running on prime time only makes sense when there is enough traffic left.

Usually it´s always better to run campaigns 24/7 and in case you see huge spikes at different times you can then run at specific times.

But 4 conversions are way too less for it.

So test it but when you don´t receive enough traffic in these few hours continue with the campaign running 24/7.

A good campaign will convert then as well, when it´s only converting few hours per day and giving you 3-4 conversions during this time it´s not worth it.

---

Really appreciate your suggestion, Sebastian.

After read your post, I review that offer, it totally spend $150, have 18 conversions, the payout is $3, and it runs for 6 days. Due to the converions is too less, so I can't optimize too early. I'll drop this offer, keep testing and finding out a workable offer.

Day19-Day20


only have 1 conversion.
I'll drop this offer, I didn't block any placement, want to test the average performance for the . The result is telling me I should pause and testing new offer.

By testing this offer, I learned a lot:
- Split traffic source, mobile-wifi, mobile-3g, desktop.
- Don't optimize to early, otherwise it will cut down profitable placement too early and narrow down the traffic.
- A formula to test offer: Landers * Offers * average payout * 10 = Budget, and split this budget to daily, run at lest two days( But I didn't follow this, so I spend too much on this offer). And I think it's better to run multiple offers parallel to test one geo, in order to test performance of the placement. Unfortually, I didn't follow this too.
- Test T3/T4 first. I believe every offer can profitable, what I need to findout the correct placement to optimise. But for newbies like me, I should get more converions first, this is important, so T3/T4 geo is suitable to test without overspend . I have read this at other posts too many times. I know this is important. Unfortally, I also didn't follow this, I'm a little too anxious to see good result, so I run high payout offer, this is a mistake。NL and IT is not sutiable for me to test currently.

TODO:
- test t3 geo's offer
- pick a small geo, check traffic sources daily impression, make sure daily budget can cover 50% of them ( Please correct me, I don't have experience, this is just my idea, I just don't want over spend for daily budget while it didn't cover most of the placements)
- pick multiple offer, at least 3 offers
- get 2 landers for each offer ( for this, shoud I test direct link, especially the offer have an prelander itself?)

And for the idea to choose geo

let's take HU for example, on the propellerads, by select mobile, wifi, 1 impresss/day, it show Estimated impressions: 189,835, Optimal Rates: 1.5.
So the daily cost will be 189835/1000 * 1.5 = 284.75 $, obviously I can't do like this way and spend at this to test 3 offers.
is this correct to calculate and select small geo?

---

I can absolutely understand it but the reality is that basically the only difference between running small campaigns vs big campaigns, low volume vs high volume, Tier 4 vs Tier 1 etc is the money that is involved.

The process for testing and optimizing is more or less the same.

You have to do the same steps, no matter how big or small the campaigns are.

This means that you better start in low Tier geos with low payouts so that you can learn all steps involved without risking to lose too much money.

The knowledge you learn from these small or lower profit campaigns can then be used for bigger campaigns.

When you then start in high Tier geos with high volume, higher payouts and higher traffic prices you know how to run campaigns already so you should avoid making newbie mistakes that could cost lots of money in these Geos.

So for learning the stuff with lowered risk a combination of low tier + low payout + lower volume is very good.

But you also should be aware that such combination will never make you much profit.

Then you need to change at least 1 factor.

Low tier + low payout + high volume can work good (anyone remembers adult 1clicks in TH?)
High tier + higher payouts + low volume can also work.

To be honest, I never thought that way.

When you are in the learning stage it also shouldn´t be that important because you either

- have enough daily budget to buy all or most of the traffic

or

- your daily budget can only buy a smaller part of the traffic but then you just set the daily limit accordingly

On Pops you usually don´t overspend too much when you set moderate bids and get rid of the big spenders (high volume placements) in the beginning.

You can then test these placements later again when you have a winning funnel.

But in the test stage it´s better to weed them out first so that not only 3 or 5 big placements burn through your budget.
In the test stage it´s always better to get traffic from many different placements to have a more diversified traffic.

Best would be to find offers for same product (iPhone X, Samsung Galaxy S10 or whatever you want to run).

Then for landers it depends, I wouldn´t take fixed numbers for the amount of landers.
I would rather decide by how many different landers are available.

I am not sure what you want to run but when you want to continue with sweeps I would recommend to test spinwheel, questionnaire and gift box landers.

Forget about that calculation.

First bthing is you will never get all traffic for your targeting in one campaign.
Even when you bid high as hell.

The next thing is, even when the possible daily cost in your calculation is $284.75 nobody forces you to spend that much.

When your daily budget is $40 (just an example) just set the daily limit accordingly.

---

Hi, there

I didn't give up, just suffering some tough things, I'm continue testing campaings today.
I get 6 soi offers for a new geo, the same vertical and when I check the creative links on the clickdealer, I found that there are multiple links for same offer, some have landingpages, some don't. And another thing is there are two offer with the same name, but different price, and one have landingpage and the other don't have, I guess maybe it's the advertizer is doing splite test.

Today I test 3 offers with totally 5 landers. For the first offer, I prepare 3 landers, and the othere two's landing page is from their creative links on clickdelare, because they provide two links, one of them havs landingpage. And I didn't find other landingpage at spy tool, so I want to test their landingpage.



The first offer is green, and maybe it's the first day to see green for an offer since I start this fa. And the second and third offer didn't perform well, because these landers is from clickdealer. I'm going the pause the last two offers if they still don't have conversions tomorrow.

And one more change is I change the deploy method, I use google cloud storage to host these static files, like s3 , I don't need to maintain a server. Is there anything that I need to take care when using this way?

---

Sometimes you also have the same offer as SOI and DOI which can result in different payouts.

I never used Google Cloud or S3 myself.

My setup is:

- Dedicated server at Hetzner (around €45 per month)
- Cloudflare for CDN and DNS ($0 with free plan)
- DNS Made Easy for DNS for my tracking domain ($60 per year)

The setup doesn´t cost much but can handle hundreds of thousands of hits per day.

And using Cloudflare is the easiest way I can imagine for CDN.

When my Push tutorial is ready I will also make a tutorial with more details about how to setup that infrastructure.

---

Not too much update recently, daily work take me much time, and spring festval is comming. So I test one ffer these days. I find that the offer which has survey prelading works better than directly (box + survey landing + offer > box + offer), here is 4 days data for 3 landers.
campaign: Android + Wifi



today I lead all traffic to the first lander.

Here is placement's data:


there is not too much convs for single placement, so I'm think I should not put which one to WL, in my opinion, I think if a placement get 5 conversions then I can put it to WL. This is a small geo, so there is not too much volume.
I have a question for when to put a placement to WL or BL, is there some common rules ?


-- update --

I forget one thing, last month one of the landing url is reported to cloudflare for abuse, and cf recovery the access after I remove the content. And yesterday this domain was deactive by namesilo without notice. And I'm trying to connect to their support to reactive the domain, but I think I won't use it as the landing page, and I need to prepare a way to switch domain quick one of the domain is deactived. And is there any other domain providers that you use?

---

Hey Peng,

Seems I missed your last post.

Are the questions still relevant?

Or do you have any other questions?

Let´s work together on more frequent updates in your FA again

---

Hi Sebastian


Thanks you, sorry too late to reply, I'm a bit busy on daily work. so I didn't update the post, but I din't stop run campaign.
I find three offers, there are the same advertizer, two of them has prelanding, below is the data from the begaining . After running some days, I pick the firt one, and lead all traffic to it.





Then I create another campaign with white list , the id is from above campaign which have at least 1 conversion. So basically, there are two campaigns, one RON campain, and one WL campaign.
Below is the WL campaign's data





And here is placement's data



some pleacements didn't have click even I bid at the recommend price, I'll create another campaign and bid more higher than recommend price to see if I can get more traffic.
I'll update the post more frequent and try to make the offer profit.

---

Hey Peng, thanks for the update.

Nice to see you´re making progress.

Check this out



Don´t only judge by the number of conversions.

You can see that the third offer with 7 conversions shows the best performance.

Same in your tracker, there it has only -18% ROI whereas the others have -31% and -43%.

So what you could do is to remove the worst offer and split traffic equally between the two better offers.
Then you can check which one is converting better.

Just from your numbers it seems that the third offer would have made more money for you when you would send all traffic there.

It´s good that you want to create a WL but 1 conversion per placement is too low.
This could also be just a lottery conversion.

I would say the absolute minimum to consider aplacement for a WL is 3 conversions but I usually only take placements with at least 5 conversions.

That way you can be more confident that they didn´t just get a lottery conversion but that they show a stable performance.

---

Here is the last 3 days update:

I splite test two offers on two campaign. (mobile android + wifi)

RON campaign:



WL campaign:




looks like the performance for the two offers are the same, but I can't judget it by 8 convs, will let it run.


The current issue is the traffic volume , from the traffic source, it says Estimated impressions: 101,463, Recommended CPM: 2.058.
but I can't get too much traffic even I bid at 6 for WL campaign. The price model is SmartCPM. I'm thinking to try CPM for WL campaign.
I know that the recommended bid is changed time by time, do I need to change my bid based their change via api? It's a bit confused to me.


As for RON campaign, 11650 visit comes from 1224 zones, I feel it's hard to test all of the performance.
I guess the way to test is find out zones which have conversions, then put them to WL campaign, and block it at RON campaign, and keep monitoring WL campaign's performance and blocked bad one. Buts this method seems too simple.
Because bid price also effect the performance, I find some zones only have 1 visit, but the CPM is 5$, 5 times than average bid.
I'm thinking if these zones has good performance, based on this, the traffic source has filtered good performance zones.
This is just my guess, I'll try it tomorrow and create a campaign to test these expensive zones.

---

Ok, good thing is that you see that your lander/offer combination is converting.

That´s great, now let´s check how we can get you more profit

Yes, keep them both running.

Is this the volume for RON or WL campaign?

Yes, on WL campaigns you can use CPM bids instead of SmartCPM.

As a general rule, when you run WL first check how your placements perform so that you can bid accordingly.
Create bid baskets for similar performing placements and run them together, don´t put all placements in one WL campaign.

That way you have more control over it and can get better results.

Dude, often the simplest things are the best.

Don´t overcomplicate the stuff

You see that your funnel is doing good in the WL campaign but to get more placements to run in WL you first need to run more traffic on RON to find more of these good placements.

---

It's WL campaign, I guess the reason is I put all placements in 1 WL campaign, I think I need to create bid baskets for similar performing placements as you suggested.


Yes, I'll keep in mind this.

I'll split the WL campaigns based on their performance and at different bids and see the results.

---

Great, keep us updated

---

Hello, this is the result after last time


RON campaign: just ignore 29, it's should be 5, other's are postback by the same id, I don't now why this happen



WL campaign



And after last update, I create another 4 WL CPM campaign, bid at 1.8, 2.5, 3.5, 5, according to their cpm on original WL campaign.
But I didn't get traffic, it's weired, so I restart orignal SmartCPM WL campaign.
And for the original WL campaign, I bid 10$, still can't get too much traffic

---

How did you chose your placements for the WL campaign?
How many conversions did the placements have when you considered them for the WL campaign?

When you took placements with low traffic and only 1 or 2 conversions it can be that these are only small placements that won´t bring much traffic in a WL campaign.

Can you post a screenshot from the original campaign with the placements you used for the WL campaign?
You don´t need to reveal the placements itself but it would be good to see how much traffic they received and how many conversions and stuff so that we can check it better in detail.

---

Here is the original campaign which I put into WL, there is only 6 zones which have more then 1 conversion

If the placements have one conversion and roi is positive, I put it to WL, I know maybe 1 conversion per placement is too low, maybe I need to wait more time because this is smal geo.


here is WL campaign data since it is created, my idea is based on this, create another WL campaign

---

1 or 2 conversions are not enough for a WL.
In a WL you want to have placements that are stable converters.
When you use all placements with at least 1 conversion and positive ROI you risk to put placements with just lottery conversions there that could cost you much money longterm because they only had a lucky shot but are no stable converters.

Then better run some more traffic to your BL campaign and collect enough stats for a real WL.

You don't do yourself a favor when you start too fast with no proven placements in the WL.

---

Hey Peng, are you still running the FA?

---

Hi Sebastian

Sorry I didn't update the post, yes I'm still running the FA.
the covid-2019 disrupted all my plans make me under a bit more budget pressure to test offer, I paused for some days.
I didn't get enough conversions from pop campaign, so can't create a good WL and BL.

Then I decided restart push campaign at 22, Feb. I learned a lot from the pop campaign, so I apply what I leared from pop to push campaign. Really appreciate you and others who give me suggestion.
I create a ron campaign for 3 offers, and it runs 2 days

baseded on the result I choose the first offer and keep it running.


for the ron campaign, I use the auto block tool to block bad performance placements, it saves a lot, and help to create a BL, the rule for BL isn't too strict at first.
I put placement to WL when it has 3 conversion, and create new campaigns, test different bids.


Below is the result from 22 Feb, until today, the offers are from two networks




And $39.33 revenue from monetizer


Thanks you again and I'll keep testing and learn how to scale it to other traffic source.

---

Thanks for the update.

Sure, stay safe and healthy.

And congrats man, you are in profit overall

---

Home > POP / PPV / Redirect > Follow-along Campaigns