How to run campaigns on pop-traffic? (Algorithm) (22)

Hey, Guys, I wrote this article in my follow along, but as people often can not find it, I post this material in this thread.
I want to say in advance that it's only my opinion and my experience. I'm not saying that it's an absolute truth. And sorry for my bad english

Ok, let's go!

1. Choose a vertical.
First of all you have to choose which vertical you'll promote. As you may already know, the main verticals are: Installs, Pin-Submits, Sweepstakes, Dating and Adult. But which are the easiest for a beginner? For me they are - Antivirus pins, installs and sweeps. BUT! At this moment it is very difficult to work with installs, for several reasons:
- there are few good offers;
- there are fines from advertisers when you use aggressive landers. I did not get a fine, but a lot of my friends told me about them.
- you can't get a very big return on investment (ROI) with installs right now.

So I think when you’re just starting out you should choose Antivirus Pins or Sweeps (Win an iPhone 6S etc..). Why these verticals? Because you can use any traffic for these verticals (but not always, read the rules of the offer). It may be a grandfather who wants to watch old films, a man who wants to watch a porn, a woman who wants to read the news or a child who wants to play a new game. They are all our potential audience and we can monetize them with pins or sweeps.

2. Choose a GEO.
The main rule for me - TRY EVERYTHING!!! Don't spy in adplexity and watch what people are running. Don't ask your manager about TOPs, because often managers will tell you about offers which need traffic, but they may not be a top GEO or a top offer. I know, because I worked as a manager for 6 months But at the start you have to choose tier 2-3 countries. I recommend starting with: Asia (excluding VN, IN - as it's too hard to convert them now), Eastern Europe, LatAm (exc. BR, AR - as there are no good offers there right now) and Africa.

So how do I choose a new GEO? I just go to the traffic source's inventory or ask my manager about top GEOs and run them all. Of course, if your budget is not very big you can only choose 3-5 GEOs. And it's normal to see our ROI in the red after 3-5 days of testing. In such cases I just change the vertical. If I have an ROI in the red on pins, I can try Sweeps. Do not choose the geos for offers, choose offers for geo. If there is traffic, and people buy it, then it is possible for it to be converted.

3. Choose the offers.
Just one phrase - TEST F%#?ING EVERYTHING!!! A fresh example: I decided to test a new geo (SK) and I asked my manager if there were any good offers at the moment for SK. And he told me: "Bro, just try xxxx, xxxx, xxxx - those are the best right now. The others are shit.". So I tried them, and my ROI was -45%. After that, I decided to test all the offers I could find (there were about 10 of them) and after the first day of my new test I found a very good offer which gave me +367% ROI! My manager hadn't mentioned that one, and I hadn’t seen it in the tops or in spy-services. So what is our conclusion? Try everything. Try all the offers you can find. Do a split test of networks, because one offer can give you different results in different networks.

Very often, managers recommend us offers which receive heavy/a lot of traffic. BUT! It doesn't means that those offers will give us a good ROI, it just means that they’re the most popular. They are not the most profitable. If you have found a lot of offers for one geo (for example you can find a lot of offers in PL, ZA, KE etc.) you can ask your manager which of them are total shit and only add the good ones. But it wouldn't be a very objective test. What am I doing? I’m just setting up all the offers which I can in order to perform a split test. If I only find one offer, that’s fine, I'll only test one offer. But you need to read the offers description, because sometimes you may run five offers at the same time, but maybe only two of them accept android, one of them only accepts one carrier etc...

4. Choose a source.
For an objective test you should run your campaign in 2 traffic sources. This is because sometimes a campaign can give you -80% in the "X" source and +156% in the "Y" source. Which sources am I using? At the moment I'm working mostly with: Popads, Propellerads, Popcash, Plugrush, RevTop, Adcash, Exoclick, PopMyads, Redirect.com and Zeropark. But to start with I recommend choosing Popads, Zeropark or propellerads because they have a good traffic volume and your tests will be objective. They have a lot of traffic in each geo, a reasonable price, simple interface and the approval process is very quick. You can use their campaigns planner or ask a manager about geos and cost etc.

5. Make a lander.
The main rule for me - STOLEN LANDERS ARE SHIT!!! You need to create your own landers if you want to make a good profit. But what should you do if you're just starting out? Well, you can use Adplexity or Adsxposed and download landers from it. But only at the start! And before you run them please clean them (delete static parameters from Adplexity) For example, when Adplexity's bot copies the lander you can see "Your LG-3451 is infected". And people still run these landers! WTF? It’s absolutely essential to check the code before using it! Just write this instead: <?php echo $_GET['brand'];?>-<?php echo $_GET['model'];?> or some other script. The same with double-dates etc. I'll write a dedicated post about landers soon. So, you have to choose 3-5 landers with different angles and do a split-test, because you can't know exactly which lander is the best without testing them, right? After 10-15 campaigns you can choose your best performing lander and translate it. That’s what I do.

6. Create a campaign in a tracker.
You have to create the campaign in your tracking system (I'm using Pixelk), then add all your landers and all the offers you can find to the spilt-test.

7. Create a campaign in a traffic source.
Ok, You've already chosen which offers to take, your landers, sources and you’ve made a campaign in your tracker. You have to create the campaign in the source. At the beginning with I set the widest possible target range:
OS: Android, iOS, Win.Phone, BlackBerry.
Devices: Smartphones, Tablets.
Connection: 3G+WiFi.
Carriers: All (Only if the offer accept certain carriers and it's a 1click offer)
Browsers: All.
BID: This is the most interesting thing. For the test I set an average bid of +20%. If I can't see it, I set: $1 CPM for Asia, Latam and Africa and $2 CPM for the EU. And after my first day I change my bid (either increasing or decreasing them). But you gotta know that if your bid is higher than the others, you will receive higher quality traffic. But please don't set $10 CPM. Change the bid gradually. If I see a lot of traffic and a bad ROI I can decrease my bid. If I see a good ROI, but little traffic - I can increase my bid.

8. Running.
How much do we need to spend to carry out a good test? For the objective test you should spend: (number of offers)*(number of landers)*(average payout)*5. For example: you have 5 offers and 3 landers, and payouts are: $5, $3, $5, $4, $7, you should spend: 5*3*((5+3+5+4+7)/5)*5=360$. But that’s too much even for me, I usually spend: (average payout)*10 for each test.

9. Analytics.
How do I do the analysis? I do it in this order:
1) Offers. I choose the best 1 or 2 offers which gave me the best ROI and run only those.
2) Landers. I choose the lander which gave me the best ROI and then create 3 others on its base, using different texts, colors and buttons etc... and set these 3 new landers and my basic lander to the split. After that, I repeat this cycle until I definitely have a good lander. Optimization of the lander can give you an ROI of +50-150%.
3) Devices (Tablets/Smartphones)
4) OSs and versions.
5) Browsers and versions.
6) ISP (Carriers)
7) Categories
8) Sites.
As you can see, optimizing the sites is the last thing that I do (I turn off only the sites which are giving me -100%). Why? Because after these 7 steps, a site which gave you -76% at the beginning could give you +120% following optimization!!! If you turned off this site right from the start, you wouldn't have any traffic from it after following your 7 steps of optimization). I make ten payments on each site before forming an opinion/making a decision.
Of course, if you see that some option is giving you -100% - turn it off immediately.

Good luck! And remember the most important thing - the more you spend on the tests, the more accurate they will be. Do not be afraid of a negative ROI, virtually any campaign can be made profitable.

P.S. If you have any questions, you can write to me here. If you want new articles, write to me on FB with suggestions for future topics: https://www.facebook.com/mrbraun31

Man! This is great info - Thanks for the detailed guide.
I added you in my contacts.. hope we can chat a bit..

Thanks for the algorithm, mrbraun! I'm trying to hit green on pops and it's been challenging especially with geos.

Awesome, thanks for all these informations !

This is a fantastic guide and deserves to be in its own thread!

Thank you MrBraun! Hats off!


Amy

I've been meaning to ask you, how do you approach the ad-block thing on Popads? Do you do both, non-adblock, or just ad block? I don't quite understand what this is.

Also, do you prefer pop under? or pop over?

Spacibo

Originally Posted by xesturgy

I've been meaning to ask you, how do you approach the ad-block thing on Popads? Do you do both, non-adblock, or just ad block? I don't quite understand what this is.

Also, do you prefer pop under? or pop over?

Spacibo

Thanks for the info!! Your posts are great, really valuable stuff

Mr Braun - Thank you so much for this - SUPER helpful and great to see how you break it down.

I have one question regarding #8 Running

Hey, thetailend!

For a great test you need to spend (average payout * 10) for each landing page/offer combination, but I don't recommend to do it. It's too expensive. Just spend $35 for a good test.

Great stuff. Posts like these were the exact reason for joining STM in the first place. Very nice guide. I even just printed it out so I can keep it and read anytime.

I haven't run any pop ads yet, but I signed up to ZeroPark last Friday and waiting for the phone approval hopefully tomorrow. After that I need to dive into the world of creating landers, which I'm sure is fun !

Love your wise words in not sticking to the norms of ripping from Adplexity and only going off what your aff. manager suggests. I've asked for suggestions from my off. managers many times and never had one conversion. Probably because I'm a newbie he only throws me the junk that needs a bit of traffic !!

Great tips anyway, and it's exactly the way I prefer to work, i.e. go with it, test plenty, and see what comes up good.

Nice guide mrbraun. Nice. Thanks man.

Great post sir!

mrbraun - Great post!

I'd like to take this opportunity to tell you about Clkim and how you can turn any content into a pop ad..

I built Clkim 2 years ago when i wanted to monetize one of my sites with full page interstitial ads. I wanted to be able to test different feeds, optimize in real time and not have to pay crazy ad serving fees. more over i needed a solution that was unblockable by ad blockers. clearly i couldn't find any solution that did that - so what does any entrepreneur do in that case - you built it!

Fast forward 2 years - Clkim is now a leading link shortening and monetization solution. Essentially we enable marketers to turn any link into an interstitial ad they control - regardless of the origin and destination. This means you can now share content leading to sites like CNN on facebook (using your own branded domain with us) and monetize it with your own full page ads.

To make things more interesting we added a whole layer of marketing that enables you to redirect your traffic based on geo, mobile os and a/b test destinations.
To make things more profitable we let you setup your own full page ad campaigns and specify how they populate.

You have full control on the content of the ads, on how the interstitial looks and how it behaves (countdown duration, auto redirects, extra pop on skip ad etc.)

Again - i apologies if this post seems overly promotional, but the fact of the matter is that we have affiliates that have increased their links CTR by 35% (because the branded links provide more meaning to their end users) plus other affiliates that have doubled their conversions rates because they can now add super target ads to ordinary links.

i'm happy to answer any questions.

best,
Yoav

Great post Mr. Braun. Lit a bulb over my head ( including your other posts as well )

@Yoav - hi man, great product you have there and thanks for sharing. did something similar for myself on a gaming project, using DFP ( to cut costs even further )

It would be however very very interesting ( for other stats sharing as well ) to actually put that increase into context. a xx% increase of something ( from - to ) can be
very misleading since it's out of context.

Also, CTR alone out of context, is again misleading for most people since it's a vanity metric.

simple math: 0% original CTR increased to 0.35% is a 35% increase and is true till the end. however, means nothing unless correlated with other significant data.

Ya reckon?

@danielt - agreed. my bad!
35% increase CTR is the average increase our users see on their posts so if you were sharing a post and typically getting 100 clicks then with your own branded link you should be getting 135 clicks..
when you think about it - its not a very clever increase - when you compare this link: http://clk.im/SaSxm to this link http://shalev.co/PeopleAreAwesome the latter will get more clicks because it has some context - people understand its coming from me and expect to see something about how people are awesome.

As for the conversions increase - i urge you to try the platform 10 days for free (no cc required). judge for yourself how much of an increase you get with it.

Regarding DFP - i found are a few issues with it (my own personal taste):
#1 their tags get blocked by ad blockers
#2 the learning curve of loading a new campaign and managing the adspots was too long and was too much of a hassle
#3 it wasn't flexible enough to do what i had wanted it to do
there were more - but these 3 were enough for me to go ahead and develop this platform

got it

the DFP choice in my case was somewhat natural since I've been using it for years before this
and ran quite a few versions of it to display ads, but what you mentioned as issues are all true

thanks for the feedback

Ok - that makes sense but I can see how it would add up fast when mass testing.

Thank you!

One more question kind sir:

What do you do if one of your offers in your split test allows only certain carriers? (e.g. offer A, B, and C allow all carriers, but offer D allows on 2 specific carriers)

Would you run this offer in its own campaign so you can toggle the carriers in the traffic source?

Thats my guess. However I've noticed that in Popads.net and ZeroPark you can't actually choose mobile carriers, which makes things difficult. Are there any other traffic sources that are more carrier-targeting friendly?

Originally Posted by thetailend

Ok - that makes sense but I can see how it would add up fast when mass testing.

Thank you!

One more question kind sir:

What do you do if one of your offers in your split test allows only certain carriers? (e.g. offer A, B, and C allow all carriers, but offer D allows on 2 specific carriers)

Would you run this offer in its own campaign so you can toggle the carriers in the traffic source?

Thats my guess. However I've noticed that in Popads.net and ZeroPark you can't actually choose mobile carriers, which makes things difficult. Are there any other traffic sources that are more carrier-targeting friendly?

Will you optimize all 8 of these variables every time you pause the traffic to analyze? Or do you do 1 at a time? Or are the variables tested in batches?

Great post! What do you recommend, how many hits to buy before turning off non performing sources?

I have created a simple worksheet to help anyone calculate the necessary budget for this strategy.
It looks like this:


If you find it useful, grab it from here

thanks Mr.Braun for the details post and strategy

Originally Posted by mrbraun

Hey, Guys, I wrote this article in my follow along, but as people often can not find it, I post this material in this thread.
I want to say in advance that it's only my opinion and my experience. I'm not saying that it's an absolute truth. And sorry for my bad english

Ok, let's go!

1. Choose a vertical.
First of all you have to choose which vertical you'll promote. As you may already know, the main verticals are: Installs, Pin-Submits, Sweepstakes, Dating and Adult. But which are the easiest for a beginner? For me they are - Antivirus pins, installs and sweeps. BUT! At this moment it is very difficult to work with installs, for several reasons:
- there are few good offers;
- there are fines from advertisers when you use aggressive landers. I did not get a fine, but a lot of my friends told me about them.
- you can't get a very big return on investment (ROI) with installs right now.

So I think when you’re just starting out you should choose Antivirus Pins or Sweeps (Win an iPhone 6S etc..). Why these verticals? Because you can use any traffic for these verticals (but not always, read the rules of the offer). It may be a grandfather who wants to watch old films, a man who wants to watch a porn, a woman who wants to read the news or a child who wants to play a new game. They are all our potential audience and we can monetize them with pins or sweeps.

2. Choose a GEO.
The main rule for me - TRY EVERYTHING!!! Don't spy in adplexity and watch what people are running. Don't ask your manager about TOPs, because often managers will tell you about offers which need traffic, but they may not be a top GEO or a top offer. I know, because I worked as a manager for 6 months But at the start you have to choose tier 2-3 countries. I recommend starting with: Asia (excluding VN, IN - as it's too hard to convert them now), Eastern Europe, LatAm (exc. BR, AR - as there are no good offers there right now) and Africa.

So how do I choose a new GEO? I just go to the traffic source's inventory or ask my manager about top GEOs and run them all. Of course, if your budget is not very big you can only choose 3-5 GEOs. And it's normal to see our ROI in the red after 3-5 days of testing. In such cases I just change the vertical. If I have an ROI in the red on pins, I can try Sweeps. Do not choose the geos for offers, choose offers for geo. If there is traffic, and people buy it, then it is possible for it to be converted.

3. Choose the offers.
Just one phrase - TEST F%#?ING EVERYTHING!!! A fresh example: I decided to test a new geo (SK) and I asked my manager if there were any good offers at the moment for SK. And he told me: "Bro, just try xxxx, xxxx, xxxx - those are the best right now. The others are shit.". So I tried them, and my ROI was -45%. After that, I decided to test all the offers I could find (there were about 10 of them) and after the first day of my new test I found a very good offer which gave me +367% ROI! My manager hadn't mentioned that one, and I hadn’t seen it in the tops or in spy-services. So what is our conclusion? Try everything. Try all the offers you can find. Do a split test of networks, because one offer can give you different results in different networks.

Very often, managers recommend us offers which receive heavy/a lot of traffic. BUT! It doesn't means that those offers will give us a good ROI, it just means that they’re the most popular. They are not the most profitable. If you have found a lot of offers for one geo (for example you can find a lot of offers in PL, ZA, KE etc.) you can ask your manager which of them are total shit and only add the good ones. But it wouldn't be a very objective test. What am I doing? I’m just setting up all the offers which I can in order to perform a split test. If I only find one offer, that’s fine, I'll only test one offer. But you need to read the offers description, because sometimes you may run five offers at the same time, but maybe only two of them accept android, one of them only accepts one carrier etc...

4. Choose a source.
For an objective test you should run your campaign in 2 traffic sources. This is because sometimes a campaign can give you -80% in the "X" source and +156% in the "Y" source. Which sources am I using? At the moment I'm working mostly with: Popads, Propellerads, Popcash, Plugrush, RevTop, Adcash, Exoclick, PopMyads, Redirect.com and Zeropark. But to start with I recommend choosing Popads, Zeropark or propellerads because they have a good traffic volume and your tests will be objective. They have a lot of traffic in each geo, a reasonable price, simple interface and the approval process is very quick. You can use their campaigns planner or ask a manager about geos and cost etc.

5. Make a lander.
The main rule for me - STOLEN LANDERS ARE SHIT!!! You need to create your own landers if you want to make a good profit. But what should you do if you're just starting out? Well, you can use Adplexity or Adsxposed and download landers from it. But only at the start! And before you run them please clean them (delete static parameters from Adplexity) For example, when Adplexity's bot copies the lander you can see "Your LG-3451 is infected". And people still run these landers! WTF? It’s absolutely essential to check the code before using it! Just write this instead: <?php echo $_GET['brand'];?>-<?php echo $_GET['model'];?> or some other script. The same with double-dates etc. I'll write a dedicated post about landers soon. So, you have to choose 3-5 landers with different angles and do a split-test, because you can't know exactly which lander is the best without testing them, right? After 10-15 campaigns you can choose your best performing lander and translate it. That’s what I do.

6. Create a campaign in a tracker.
You have to create the campaign in your tracking system (I'm using Pixelk), then add all your landers and all the offers you can find to the spilt-test.

7. Create a campaign in a traffic source.
Ok, You've already chosen which offers to take, your landers, sources and you’ve made a campaign in your tracker. You have to create the campaign in the source. At the beginning with I set the widest possible target range:
OS: Android, iOS, Win.Phone, BlackBerry.
Devices: Smartphones, Tablets.
Connection: 3G+WiFi.
Carriers: All (Only if the offer accept certain carriers and it's a 1click offer)
Browsers: All.
BID: This is the most interesting thing. For the test I set an average bid of +20%. If I can't see it, I set: $1 CPM for Asia, Latam and Africa and $2 CPM for the EU. And after my first day I change my bid (either increasing or decreasing them). But you gotta know that if your bid is higher than the others, you will receive higher quality traffic. But please don't set $10 CPM. Change the bid gradually. If I see a lot of traffic and a bad ROI I can decrease my bid. If I see a good ROI, but little traffic - I can increase my bid.

8. Running.
How much do we need to spend to carry out a good test? For the objective test you should spend: (number of offers)*(number of landers)*(average payout)*5. For example: you have 5 offers and 3 landers, and payouts are: $5, $3, $5, $4, $7, you should spend: 5*3*((5+3+5+4+7)/5)*5=360$. But that’s too much even for me, I usually spend: (average payout)*10 for each test.

9. Analytics.
How do I do the analysis? I do it in this order:
1) Offers. I choose the best 1 or 2 offers which gave me the best ROI and run only those.
2) Landers. I choose the lander which gave me the best ROI and then create 3 others on its base, using different texts, colors and buttons etc... and set these 3 new landers and my basic lander to the split. After that, I repeat this cycle until I definitely have a good lander. Optimization of the lander can give you an ROI of +50-150%.
3) Devices (Tablets/Smartphones)
4) OSs and versions.
5) Browsers and versions.
6) ISP (Carriers)
7) Categories
8) Sites.
As you can see, optimizing the sites is the last thing that I do (I turn off only the sites which are giving me -100%). Why? Because after these 7 steps, a site which gave you -76% at the beginning could give you +120% following optimization!!! If you turned off this site right from the start, you wouldn't have any traffic from it after following your 7 steps of optimization). I make ten payments on each site before forming an opinion/making a decision.
Of course, if you see that some option is giving you -100% - turn it off immediately.

Good luck! And remember the most important thing - the more you spend on the tests, the more accurate they will be. Do not be afraid of a negative ROI, virtually any campaign can be made profitable.

P.S. If you have any questions, you can write to me here. If you want new articles, write to me on FB with suggestions for future topics: https://www.facebook.com/mrbraun31