Home > Paid Traffic Sources > Mobile

---

The Lazy Follow Along to Mastering Mobile (33)

---

Hello mates,

How are you doing?

After a year of procrastinating and debating with myself to jump into paid traffic. I have finally decided to start my first mobile campaign on pop traffic.

The potential money I could earn doing SEO was a bit of distraction, even though I made my first six-figure profit in 2016, it's tough to ignore the digits that the paid traffic guys are making. So I'm.

My starting point is very low and newbie, I spent a few days trying to understand tracking, uploading landers, basics of programming, CDN, things you can pretty much ignore if you are doing SEO with wordpress sites. In a nutshell, getting the basics and the foundation straight before I could invest in data and potentially make a few bucks.

My game plan is to start promoting offers in Brazil and LATAM, so I can upload landers fast and decrease the amount of time spent on translations. Just like any skill, I believe that the faster I can get data, the faster I can learn how to manage and make campaigns profitable. And if I seat in this chair I'm in now for a couple of months, it's almost impossible that I won't make campaigns green, or maybe yes.

My budget is: $100 USD per day.

My First Campaign Stats:

Vertical: Sweepstake (Win Iphone 6S)
GEO: Brazil
Traffic Source: Pop Traffic (PropellerAds)
Offer Payout: $0.13
Bidding CPM: 2.02

I spent $20 and made only 14 conversions.

What I did, I ripped 3 landers and distributed the traffic between them. I messed up with a tracking, I didn't specify the CPM bid when setting the campaign. Since I couldn't find a solution to it, I calculated the ROI roughly >83%.

Here is the stats:


My question are:

1) Should I run the campaign again to get the data on the ROI and make further analysis or test other campaigns?
2) What is the min. ROI required to optimize the campaign and turn it green?
3) What is the average CPM bid in Brazil for the first testing phase, propellerads suggests (Min CPM rate is $0.50.We recommend to choose CPM rate $6.26 for the best result)?
4) Is the offer payout okay for that GEO?
5) How to track high converting or promising placements?

Sorry if I'm asking too much.

Thanks guys

---

Welcome back and cheers to a great start of a follow-along!

LATAM is a great choice! And I'm in complete agreement with everything you said in the quote above.


A couple of suggestions:

-BR has a lot of traffic volume, and if you're targeting wifi traffic, it will take quite a bit of cutting placements to make the camp profitable. Ideally you would want to use a really good offer+lander for cutting placements - otherwise it could get really expensive (as you've already seen). Other LATAM geos would definitely make for an easier start for a new person. But let's see how you get on with BR if you wish to continue.

Note: Just to make sure, you ARE aware that BR speaks Portuguese whereas most other parts of LATAM speaks Latin Spanish, right? Just want to make sure, because I've made that mistake before!

-You need to test more landers, and preferably more than one offers to make sure you have at least one that converts well enough to help you cut landers down to a winner. There's a rough guideline on running sweeps at the end of this post for your reference:

https://stmforum.com/forum/showthrea...l=1#post293497

One of your landers is ready to be cut - please follow this methodology:

http://stmforum.com/forum/showthread...Banners-Part-1

If you run a bit more traffic to the 2 remaining landers, you should have a winner soon, but because your current ROI is so low, I would suggest to just start another split-test instead. Take the 2 remaining landers, and add another 5-8 landers, and also 1-2 other offers that have similar payout, and start another test. The split-test calculator assumes all conversions to be worth the same, which is why it's important for the payouts to be similar. The more similar, the more accurate the results.


1)No need. Just approximate by assuming all your traffic is costing $2CPM (which is what you're bidding). Check your OS and Mobile Carrier stats to see if any of them is remotely close to being green - I'd say -30%ROI or better. You can check OS and Mobile Carrier separately, and also in combination, i.e. OS > Mobile Carrier (drilling down to 2 levels).

All you need to have a profitable campaign, is an offer+lander+[at least one major traffic segment or a few smaller traffic segments] that's doing 30%ROI or better (after all, your time is worth something, plus you have business costs to cover!) A "major traffic segment" is one that has enough traffic volume to provide you with enough profits to justify your time in running the camp. Examples are android/ios, phone/tablet, a major mobile carrier, even a browser. I was just talking about this concept in another reply earlier today - for your reference:

https://stmforum.com/forum/showthrea...l=1#post302985

2)Great question! I would say your Offer>Lander>[major traffic segment] will need to be doing at the very least -30%ROI or better (usually 0%ROI or better), before you spend money on cutting placements and implementing dayparting to give the camp a final push into green. Make sure you test enough offers and landers until you have a great offer+lander combo, and THEN cut unprofitable traffic. Don't take an offer+lander that's doing very negative ROI and try to make your camp green by cutting placements, as it can get expensive, and you may be left with no traffic after your cutting spree.

BR though is a bit of an exception - because it has so much traffic volume, you can probably cut a lot of placements and STILL have enough traffic left over to profit from. In spite of this though, you should still test offers and landers until you're doing reasonably high ROI before relying on cutting placements.

3)Gosh - don't bid at the "for best result" bid unless you know you're running an offer with high profits margin and just want to hog the traffic.

Don't get me wrong - it's great to test bids and even high bids (in fact I encourage it - bidding high can trigger traffic from high-quality placements you wouldn't otherwise get traffic from). But for initial testing of landers and offers, I'd suggest bidding a bit higher than the first "bend" in the traffic graph:



The first bend is at $1.14 (I'm assuming you're targeting wifi traffic?) so maybe bid 1.22, 1.33?

4)The payout is only a part of the equation. I've made profits from $0.03 payout offers. Offers are set at certain payouts for good reasons usually. Low payout offers tend to convert better than high payout offers (of course this is a generalization and there will always be exceptions). It's true though that when you have a high payout offer that converts well, your profit margins are more likely to be higher than if you had a low payout offer. But I wouldn't pass on an offer just because the payout is low.

5)You'll see placements stats in Voluum, and also at the traffic source.

You can approximate costs as I've mentioned above based on your bid CPM. Or if you really want to be accurate, you can export placement+cost data from the traffic source, and export placement+revenue data from Voluum, and combine them in a spreadsheet:

https://stmforum.com/forum/showthrea...kly-amp-Easily

I do have a spreadsheet tool you can use to cut placements accurately, but that would be overkill for pop sources. Instead, using rules of thumb would be better:

http://stmforum.com/forum/showthread...Rules-of-Thumb

For example, I like to cut a placement when it's in loss by 2x payout.


Never apologize for asking questions! Two reasons:

1)You're paying for forum membership to get this kind of help!

2)Chances are other newbies have the same questions, so they'll be learning from the discussions as well.

I'd like to thank you for taking the time to start a follow-along and for being transparent with campaign details. This is the best way to get detailed feedback, and at the same time will help other beginners that may be reading your follow-along as well.

Looking forward to your next update!



Amy

---

can't wait to see your progress!

---

This will be a great thread to follow. Good luck with everything!
-Zach

---

I'm running propellerads and sweepstacks too. In south africa. What is LATAM?

---

LATAM stands for Latin America.

---

I'm back with more launches huhu...

The next thing I did was rip 6 more landers and distribute traffic among them promoting one offer:

Traffic source: PropellerAds
Vertical: Sweepstakes Win an Iphone
Payout: $0.18

Here are the campaign stats:


Cost: $10
Revenue from Affiliate Network: $1.02

From my manually calculations of ROI, the best lander doesn't have an ROI above 73%. I will keep launching more campaigns until I hit a reasonably ROI that will make optimization possibly profitable.

I'm starting to become more flexible with the process of campaign launching but I still have issues with tracking.

1) What can possibly be the reason Voluum is not showing impressions?
2) Why is Voluum not tracking ROI
3) Do you guys have a specific analytical formula you use to decide how much budget you spend in every campaign you launch?

---

Voluum's not tracking ROI because it doesn't know what you're spending per visit: you normally need to run a separate spreadsheet for this, as Vortex mentioned.

Analytical formula: the usual rule is "get to statistical significance, then make decisions based on the ROI you're reaching". You can figure out statistical significance based on the posts Vortex linked to. As for ROI - that's where rules of thumb come in. Usually I drop a campaign if it doesn't do at least -50% ROI on its first test, but a lot of this comes down to developing hypotheses about why the campaign didn't work.

You can't set a fixed budget to test a campaign under normal circumstances - but you can set "stop loss" rules at each stage of the campaign testing process to avoid spending more on campaigns that are unlikely to succeed, and I recommend doing so.

---

Caurmen has dropped some very wise tips - I would suggest to read them 3 times.

Couple things to add:

-I think impressions are for banner ads, so if you're running pop traffic or not tracking banner impressions then the numbers would stay at 0.

-I wouldn't care too much about ROI at this stage assuming you're not losing a lot of money, because once you have a winning lander, you can take that and test all the similar offer you can find - that's usually when the magic happens.

Put it this way: To have a successful pop campaign, you need a good offer and a good lander (and good traffic, but I'm only talking about your funnel). So you got to lock them down and it's more efficient and organized to lock them down one at a time. So you either lock down a good lander first, then use it to find a good offer, or the other way around. Right now you're in the first stage of locking down a good lander. If you don't run to statistical significance you won't be able to proceed to the next stage where you take it and test offers.

Of course, if it's looking like all your landers suck, you may want to replace them with a new batch (of course the more landers you test, the more you're minimizing that possibility - having 6 landers in rotation should be good enough for a start). Same for the offer - if you're suspecting that it's a dud, then by all means replace it. This is coincidentally why I always suggest to use more than 1 offer from the beginning to improve your chances that at least 1 of them will convert well enough to give you conversions - you need conversions to help cut down landers. Asking your AM for offers that a lot of other affiliates are getting conversions for (as opposed to having all the conversions coming from a couple of super-affiliates) would be good.

How the above applies to your situation: If you keep testing landers until you reach a desired ROI, it may never happen, because right now you're limited by that one offer.

So you deal with things on RELATIVE terms for now instead of absolute - test and find the best lander first, THEN when you're testing offers you can evaluate on absolute terms, e.g. based on ROI and profits.

Hope that makes sense! If it doesn't, please ask for further clarification.

Looking forward to seeing more stats! You're making great progress here, and testing more landers like what you're doing is always good.



Amy

---

Sweepstakes in Brazil is a great choice. Had some nice profits there.

Focus more on the offer instead of the landing page. The offer has the higher leverage. The offer makes or brakes your campaign. The lander just need to be good enough. Instead of testing 6 lander and 1 offer. Test 6 offer and 1-3 Lander.

Ask your AMs for recommendations, live spy and check adplexity which offers the other guys are running.

Check which lander is doing best internationally (competitive Tier1 GEO). Chances are high, that this version will work well in Brazil too.

Sending carrier traffic to a leadgen with $.2 payout is a bit of a waste. You can try setting up a second funnel with Voluum rules for carrier traffic and test a pin submit. They have higher payouts.

---

Hello guys,

Thanks so much for your output.

I'm not sure if I really understand what @sebastian is saying.
'
My question is 1) When you say "Test 6 offer and 1-3 Lander", do all offers need to be in the same GEO, and should I run all of them on the same campaign and rotate offers in Voluum, or for every different offer, I need a separate campaign on Voluum.

2) I'm having issues with Popads, I took the winning lander and tested on a new offer in the same geo. There is a message that says "Your campaign is currently throttled to spending not more than $3 per minute. Make sure to have more than $15 remaining in both budget and daily budget to avoid throttling.", even though I have my throttling disabled (I don't know what it is).

Results:
Offer payout: $0.18
Traffic source: Popads
Targeting Android

ROI: -50%

---

Yes. 3-5 offers per campaign in one campaign. You need to test lots of offers to find a winner.

Either your account budget is too low, or the remaining budget of your campaign is too low.

---

Hello there,

Things have been slow on my end. I haven't launched any campaigns since my last post.

If things go slow I go broke. I found out that the majority of my time is spend in fixing landers so I brought on board someone to help me with that so I could buy more time.

I don't know if it's too early to hire a part timer, I'm just praying it is .

So back to campaign launching, I went to adpexity and ripped the landers from PROs who are apparently killing at specific geo (Sebastian_R made a video on that).

I ended up with 5 landers. Ran some traffic and distributed them to 2 offers recommended by AM.

Here are results, I cut the worse, and found lander 5 and 3 to be the best performing. I will then take one of the landers (5) and slip test among other offers.

1) I don't know if that strategy is solid and whether I understood well how I should go about testing? Please let me know.

Here are the stats from Voluum
Geo: BR
CPM: $1.35
Traffic: WIFI
Offer: Win an Iphone 7
Device: All mobile
Traffic Source: Propellerads
Traffic Cost: $30

---

Your strategy is solid.

However you're testing more landers than offers.

The offer has the higher leverage.

So either spend more overall to test more offers as well.

Or cut down a bit on the landers (2-3) and test more offers instead.

For this particular camp, cut lander 1 & 2.

Let lander 5 & 3 run till you have significant data.

---

According to the peakconversion calculator, landers 1, 2 and 4 are ready to be cut.

It's looking like landers 3 and 5 are going head-to-head - in this case it wouldn't be productive to wait for stat sig for them.

I would definitely suggest to move to the next stage: Use both these landers to mass-test offers.

Like Sebastian pointed out, the offer is more important. The lander needs to be decent too, but there are only so many popular landers you can rip and test at any one time, whereas there are normally a lot more offers you can test. Plus the difference in CR between offers has the potential to be bigger.

After a round of offer-testing, you can drill down into OS, Browser, Carrier (if you're targeting that along with wifi), Device etc. I would suggest to only cut the major losers at that point, because there's a ton of BR wifi traffic - so you'll be able to afford to cut a lot of placements (instead of a major targeting segment) to get green and STILL have enough left over to make good profits.

Then, start cutting placements in loss by >2x payout. After cutting out the bigger losers, increase your bid again and cut another batch.

There are lots of ways to optimize placements, but this is one of the simplest.

Have fun!



Amy

---

Hello there,

I have been launching campaigns almost every day.

But I have not seen anything promising yet.

One of the campagins I was running, I took 2 winning landers, and mass test with 5 offers in Brazil. The ROI was around -80%. I then took a look at the major traffic segments saw nothing close to green and not even greater than -30%. I'm going to test some landers in Indonesia and see what results I can get I will post my results here.

I have been wanting to launch PIN submits campaigns but I was confronted with issues when mass testing offers.

I have set up on Voluum 9 offers almost similar payout, anti virus PIN submits. However, some offers only accept certain carriers and other don't. How do you go about this?

Lets say:

offer 1 > accepts X Carrier + Wifi
offer 2 > accepts X, Y Carrier + Wifi
offer 3 > accepts Y Carrier + Wifi
offer 4 > accepts Y Carrier only
offer 5 > accepts X, Y, Z Carrier + Wifi

How do you set it up?

---

You can use flows or path campaigns with rules in Voluum in order to route the traffic to offers that accept certain carriers. I would also recommend to stay away from wifi traffic completely, when promoting PIN submit offers - the CVR from that traffic is usually VERY bad.

---

For a big geo like Brazil, if you're running wifi, you'll need to invest into cutting placements aggressively. Here's a tentative plan:

-Test at least 5-10 landers to come up with a decent winner (which you've done), then test a bunch of offers to make sure you have a decent winner (this is subjective and relative, but obviously the more offers you test, the better the ultimate winner will convert).

-While bidding average and running your best offer+lander combo, cut placements that are in loss by 1-2x payout.

-After you've gotten rid of the bigger losers, increase your bid again and do another round of placement-cutting. Alternatively, instead of raising the bid, you can leave the original camp running at green, duplicate the camp, raise the bid on that, and cut placements (you can either cut placements based on loss for this new camp, or cut all the placements that have been sending good traffic volume in the original camp, so that the new camp will only trigger traffic from the higher-quality placements).


Like matuloo said, you can set up tracker rules to handle this:

Default -> Offers 1+2+3+5 (all the ones that accept wifi)

Rule -> If Mobile Carrier = X, rotate offers 1+2+5

Rule -> If Mobile Carrier = Y, rotate offers 2+3+4

Rule -> If Mobile Carrier = Z, go to offer Z

However - it would be better to just set up one camp for Carrier Y to see whether it has potential, and also set up a wifi camp for the rest of the offers, just to see if they'll even convert. If the conversion flow for carrier vs. wifi is the SAME, I would normally test the offers on wifi first to find a winner, and then test the winner offer on carrier traffic. This is because an offer that doesn't convert at all on wifi, will likely not do too well on carrier traffic either - this is a generalization of course.

BUT, if you have an offer that has a different conversion flow for carrier vs. wifi traffic - e.g. 1-click for carrier and MO for wifi - then it would be worth it to test the offer using both types of traffic, because the simpler conversion flow can mean a significant difference in CR when compared to the more complicated conversion flow for wifi.

In general, carrier traffic is significantly more expensive than wifi traffic, so it would be a good idea to set up separate camps to target each, so you can bid appropriately for each. You could even set up separate camps for EACH carrier.

If you try to target both wifi and carrier in the same camp, you'd either be overbidding for wifi traffic, or underbidding for carrier traffic. And you'd always get a ton of wifi traffic with just a little bit of carrier traffic no matter how high/low you bid. So whenever you have reason to test each type of traffic thoroughly, set up separate camps to target them separately.

Hope I haven't made this more confusing that it needs to be!



Amy

---

I tested antivirus PIN in Brazil
Popads:
Android:
Smartphone and Tablets
Wifi Traffic
CPM: $1.1

Lets call it camp 1:

Ripped 3 landers and test 7 offers that accepted wifi as MO flow:

ROI= -100% (I don't know if this proves how terrible PINs are with wifi traffic)



I will send carrier traffic to the offers, just for practice purposes.

---

Yes!


Would be best to select carriers X+Y+Z, because there may be additional carriers (many countries have more than 3 mobile carriers).


Yup! Carrier traffic can be expensive alright!

In cases like these, I would suggest to set up staggered bid camps - one at around the average bid as suggested by the estimator, and one that's more reasonable - such as 0.002-0.005.

I just checked propeller's estimator and bids there seem pretty high as well for these carriers. Judging from the popads estimator's distribution charts, it seems that quite a few people are bidding really high. We don't know if they're offer owners or affiliates though. There may very well be one or more offers that are converting well, therefore driving up the prices.

If you go to Adplexity and specify popads + the 3 BR carriers + last 7 days + "Received most traffic", you'll see quite a few antivirus landers. So there may be some good AV offers available somewhere - you could do some research there.




Amy

---

Yes Pins with Wifi doesn't work in most cases.

Pins in Brazil are tough as well, the carrier regulation is pretty strict there. Big T&Cs on the landing page.

---

Hello Gents,

Thanks for your valuable replies. You guys are awesome.

A quick update. I have launched campaigns on a new GEO that is giving me better results. Maybe it is because I'm becoming better with analyzing data from launching campaigns almost every day. I'm also sending remnant traffic to afflow (WOW I didn't know I could ever do that)...nevertheless, I'm enjoying the process.

I took my best lander and mass tested offers, there is this offer that gave me +20% ROI, for one carrier and Samsung Models. However, when I whitelist the Samsung model and the carrier, it then gives me an ROI of -20.

Question 1) Why did this happen? Is it because the traffic segment I used to make that decision didn't reach statistical significance? It had over 10 conversions.

Question 2) When you optimizing a campaign and see that a traffic segment is promising, do you create a separate campaign (with a unique URL obviously) to analyse it separately?

Question 3) In GEOs where aggressive placement cutting is needed, is it possible to take a campaign from -70% ROI to green? How much change can cutting placement impact on the campaign ROI?

---

On question 1) - over how many days did you gather that data?

Conversion rates fluctuate day to day, so it's generally a bad idea to assume a trend based on a single day's data.

---

1)Like Caurmen said - it could just be fluctuation. Also, it's common for conversion rate to decrease as you run pop camps longer. Sometimes it starts to happen within hours, sometimes days, and if you're lucky, weeks. But it will happen - pop camps typically don't remain profitable for too long.

2)You can do this 2 ways - whitelisting like you did and then continuing to cut placements, or just blacklist the segments that are in loss by so much that cutting placements probably won't make them green. However, unless you're running in a high-volume geo+carrier, whitelisting one model or brand would probably not leave you enough traffic to be worth your time. Pick MAJOR traffic segments to focus on, i.e. segments that have enough traffic to justify your efforts.

3)That would really depend on the particular situation.

I would start by looking at placement stats:

a)Check out the placements that have made conversions - are they in the green? How many placements are in green? How many are in green with 2 conversions or more? Being in green with 1 conversion does not mean much because it could just be a "lottery" conversion - however if you're seeing a bunch of placements that have 1 conversion each but they're all in green then it's still a good sign. Being in green with 2+ conversions is a more certain sign that the placement will end up being in profit in the long run (relatively speaking anyway since pop camps don't usually live very long).

b)Traffic volume of green placements - if all the big placements are in loss and only a few small placements are green, it may not be worth it to keep the camp running - unless all the small placements collectively will STILL give you enough traffic to profit from.

c)How is the traffic distributed over the placements? If there's one placement or a couple placements that are big and in grave loss, that you can cut to immediately increase ROI by a lot, AND you see quite a few green placements that still have good traffic volume, then it may be worth optimizing further.

Generally-speaking though, unless you're getting a TON of traffic from a big geo/geo+carrier, I would think twice about taking a camp from -70% to green - because by the time you get to +30%, you may be left with a trickle of traffic making you a few dollars a day, after you've spent hundreds on cutting placements.

ALWAYS try to improve your funnel (i.e. offer+lander) to effect the largest changes in ROI, and then cut all the unprofitable stuff including placements. Even for high-volume geos/geo+carriers where heavy-cutting of placements is required, you'd definitely want to be using a good offer+lander combo to do the cutting if you don't want to lose your shirt in the process. I learned that one the hard way when I was a newbie, trying to use Exoclick RON to test banners for DE Wifi. All I can say is OUCH!



Amy

---

Thanks a lot guys for the valuable knowledge in this thread. I will be very ashamed if I don't make this work.

Anyways, I'm joining mrpayne's challenge, for more details you can check it here:
https://stmforum.com/forum/showthrea...ero-to-xxx-Day

For the last two months, I have been launching campaigns on a consistent basis trying to master sweepstakes in Brazil and Indonesia, but couldn't find anything green or close to green.

Now I'm trying antivirus pin submits on a new GEO. I ripped 4 landers and cleaned them, asked for AM for best recommended offers.

Will share the results here, as the campaign is still running.

Thanks

---

Quite an interesting thread, a lot of deliberations and worthwhile conclusions.

Can't wait to see how Kelvonroy would do on Pin Submits, looking forward to campaign results.

---

So update #1 for the pin submit



Finally something green. That offer I drilled down to the best performing traffic segment, and best lander and offer combo.

For this campaign the offer payout is >$4.

1) Now another question, how to scale this? I haven't read anything on scaling for if you know any good threads, I will be very grateful.

What I did now, is increase the bid to see if I can get more traffic. I also launched two separate campaigns one lower and one higher bid to see which bid performs better. However, since I'm working with Popads, it is taking me ages to have my campaign approved sometimes 24 hours. Question 2) Do you wait for very long to get your campaigns approved on Poptraffic sources. If there is a shortcut please let me know.

I'm also trying to move the campaign to other traffic sources, still waiting for approval.

Question 3) Should I ask for a payout bump at this stage?

---

Awesome on the green!! The EPV and ROI are looking shweeet!

Testing bids is very necessary so is a great first step.

Regarding getting camps approved faster - once you hit around $12k total spend on popads, hit up support to ask for VIP status - all camps will be auto-approved instantaneously after that.

On other sources, once you're spending a healthy amount of money every day, you can request for a dedicated account manager that you can contact on skype. You'll get camps approved a lot faster that way.

As for scaling - yes scaling out to other traffic sources would be great - just take your best offer+lander+targeting, and on each new traffic source 1)test bids and 2)cut placements.

Before you scale though, definitely ask for a lead quality check to make sure everyone's happy first, before scaling. If they come back to tell you the quality is acceptable, THAT would be the perfect time to ask for a pay bump. "I'm looking to scale this big! But I have a couple of offers from different networks going head-to-head and can't decide on which one to scale. Could you give me a bump to aid me with that decision?" *wink*

Another way to get more value out of your traffic: Keep your current camp running, as well as other camps with staggered bids, and cut placements for each. Or, set up a higher-bid camp and blacklist in this new camp, placements that are getting good traffic volume in your current camp. Repeat for higher bids. This way you'll have each camp triggering traffic from new, higher-quality placements. So now you have 3 bidding/placement optimization approaches. Feel free to test one/all to see which one you like best.

Also - if you've blacklisted some targeting or placements while you were still testing offers and landers, you may want to test those again using your best offer+lander combo - some of them may be profitable now.

Lastly: Play with frequency. Duplicate your camp and set it to a higher frequency and play with bids. You'll get more traffic but lower ROI, so the overall result could still be higher profits. You can bid lower to compensate for the lower ROI as well - play with bids like I've mentioned.

Reminder: Don't spend time messing with smaller placements. Optimize the bigger placements and leave it at that. It would be better use of your time to focus on scaling, and on testing new stuff, instead of messing around with a ton of small placements.

Great job!!




Amy

---

A little update, I was able to make XX/day for a few days with that offer. But unfortunately, I was notified that the offer is paused .

Little did I have time to learn how to scale an offer to new traffic sources. But the fact that I had my first green signs on the tracker, keeps me motivated to test more and more.

Now scaling is my sticking point. I created accounts on more than 10 pop traffic sources, get them ready for when I find the next best offer.

One thing I have realized that caught me off guard was the fact that the URL gets flagged. Just when I thought I could let the landing pages run for days without changing the URL, this happened. I saw the ROI decreasing exponentially, thought it was an issue with an offer, but it wasn't. From the advice I got, I think finding a coder to design a script that checks the domain on virustotal.com, would be the best course of action.

Question 1) If you have a better way to deal with domain flagging, I would love to hear your thoughts. There is also some discussion that went on: https://stmforum.com/forum/showthrea...light=flagging

As they say, "an ounce of prevention is worth a pound of cure", if there is a way to prevent flagging from happening or extending the life cycle of a domain, please let me know.

One of the mistakes I'm probably making is not testing more landers. As soon as I find the winner, I become complacent and never attempt to improve it.

Question 2) What are the other variables of landers for antivirus PINs that have the highest impact to slip test.

Lastly, thanks for the help guys, I'm learning a lot and enjoying every bit of progress since I started this follow along.

---

I hope I won't get eggs/tomatoes thrown at me when I say this, but to date I haven't gotten a single domain name flagged (since I started doing AM a couple years back).

I think the main reason for that is I host them on the same IP as one of my sites, which is an authority site in Google's eyes - it has tens of thousands of pages, and was making 5-figures/month in Adsense earnings in its prime.

Google must be thinking that if a site is hosted alongside an authority site, that it must be a quality site as well.

However - the risk is that Google may catch on sooner or later, and penalize my authority site by association. That's the risk I'm taking right now.

So if you have a legit site, you can do the same - just be aware of the risks stated above.

Just test more landers in the beginning to find the best, then just run the winner and focus on scaling.

If you're really reluctant to do something, try to spend your time doing something else that can achieve the same goal, that you enjoy doing more (or hate doing less).

Testing more landers is one way to improve your bottomline, but so is testing offers, and scaling profitable camps to more sources.

So if you detest testing more landers, then just test more offers and focus on scaling more.

Having said THAT: Having a decent lander is still important. The offer is more important, but the lander still needs to be decent or you'd be at a serious disadvantage. This is why I suggested to test more landers in the beginning before complacency sets in. If you want to test fewer landers, then at least spend more time on optimizing each one for load time, and to test each one thoroughly to make sure everything works (e.g. all the js/jquery interactive stuff) and is displayed properly on all devices (browserstack is your friend).


Here's an exercise you can do to answer your own question:

-Go to Adplexity and browse as many AV landers as you could (select "received most traffic").

-Make a list of the different themes you see (e.g. google message, android message...)

-Make a list of the different features that you see (e.g. count down, virus scanner, vibrate, audio alert...)

-Make a list of the different ad text angles that you see (e.g. 2-step, loss of privacy, loss of data, slow phone speed, risks of device "bricking"...)

Then lock things down one by one. Example: First, rip maybe 2-3 landers for each theme, and test those to see which theme seems to be doing the best. Then take that theme and test ad text angles. Then take the winner and test features.

If testing landers isn't something you enjoy doing, then just pick a couple/few features/angles you really believe will make a difference and leave it at that. Trying to be a perfectionist when doing something you enjoy the least can lead to inaction. ANY testing would be better than no testing at all.


Likewise! Thanks for taking the time to document your journey!


Have fun!




Amy

---

Hello,

Time for a little update.

I'm still only earning low $XX/day and it's been more than 2 months that I'm consistently running campaigns.

I have some aged domains that were part of my PBN when I was doing good numbers with SEO. I will use it and see if solves my flagging issues.

I'm still getting new domains flags almost every day. I need an immediate solution and now I'm a bit afraid of letting the campaign run while I'm sleeping, if the URL gets flagged I will not be able to change to new URLS.

Yesterday, I launched another campaign in this new geo that I'm currently working on and found a good offer. I actually found 2 and I'm running them together like a battle. If one performs better during the day, I leave it running and stop the other, and vice versa.

Here are the results for yesterday:


I have also noticed that offers tend to have conversion variation in different hours of the day. I think I need to read more about day parting.

I'm still running those two offers with positive ROI. Asked for a bump but advertiser requested to go over 50 leads a day to get a payout bump. Now, my sticking point is scaling the offer to other traffic sources. I have tried but the ROI is just terrible. I will try other traffic sources to see if I can get better results and a payout bump.

Today, I'm running another offer that is promising here are stats:


I found the best combo (lander+traffic segment+offer), the offer 1 and offer 2 are running only on carrier traffic which is more expensive and limited maybe if they convert well on wifi, I might be able to scale it to XXX a day, pray for me hehe.

My other issue is cutting placements.

I see from the forum that some people cut placements based on the CTR. But I find it can be a hit or miss sometimes, due to the scripts and bots that results in a skewed CTR data. I once blacklisted the placement according to CTR cut off using the rules of thumb calculator. I then took the same placements and sent traffic to them in a separate campaign. Here are the results:


The placements that I once cut, where then profitable. In fact one placement got two conversions.

Question 1) When is it safe to cut placements based on the CTR?

I'm starting to enjoy this whole process of launching campaigns but I find myself very limited in the amount of campaigns I can launch. I will have to figure out a way to doing things faster and launch faster which means more offers.

Again

Thanks for the help

---

Another update. First XXX a day revenue.

I'm still launching campaigns in antivirus pin submits.

I'm no longer running the offer I mentioned in my last update. Because the advertiser blacklisted me and wants backs leads worth $90. My AM informed me the other morning that the advertiser saw the lander I was running. It mentioned the carrier name and carrier logo. Those were my mistakes for not reading the entire rules for running a campaign. One of those newbie mistakes you know.

I tried to convince my AM to tell them it was an honest mistake. Good news is they unblacklisted me from running their offers.

However, once I started testing it, the ROI was terrible. I moved on to testing more offers.

Here are the results:


What I did here was:

I took two landers that were doing great for the last offer I was blacklisted and rotated with 5 offers. After I gather some data I used the tool [URL]https://win-vector.shinyapps.io/CampaignPlanner/[URL]. They say the rule of a thumb is to have a offer with 95% chances of being the best. I cut some corners here and with 90% I found offer 1 was the best.

I then stopped running the rest of the offer and kept offer 1. It ended up making my first XXX a day in revenue huhuuuuu. I'm pumped. Even though, it is not a lot of money, I can see some progress and that this paid mobile thing is slowly paying off.

I also setup another campaign cut placements from the previous one and lowered my bid. The results were great too:


This give me a better idea about the bidding landscape. My next course of action will be ask for a payout bump, cut placements and create a separate campaign with higher bid and possibly scale this campaign to other traffic sources.

Meanwhile, I will also be testing more offers, I now see how responsive my AM is and I have a good relationship with him.

Wait for more updates.

Thanks-

---

Woohoo!! xxx/day profits is right around the corner! Way to go!!

Can't fault you on your process - I have nothing to say about the decisions you've made. It looks like you have a good grasp of the process now.


Regarding AV: Nowadays, it's difficult to run antivirus on pop if you don't cloak.

Advertisers are one thing you could still get around - by asking AMs for only offers that allow aggressive marketing. The tough thing is finding pop sources that will allow aggressive landers.

If you don't cloak, you'd basically be limited to popads (which has limited traffic unless you stick to the larger carriers or wifi in the larger geos), and maybe zeropark and popcash - and not all will accept just any lander. e.g. Even Zeropark will reject the google logo.

I'm not encouraging the practice of cloaking - just stating an observation. If you want to keep running AV without cloaking, consider 1)running only offers that allow aggressive, and 2)sticking to popads, and maybe zeropark and popcash.


Any time you need to cut anything, you'll be faced with the same dilemma: When do I know I could cut it?

Cut early, and you could save more test budget, but may be cutting something that could turn out to be profitable in the end.

Cut late, and you'd be more certain about not cutting a potential gem, but it would take more test budget.

It's the same efficiency vs. accuracy argument I always bring up in every optimization-related discussion.

Main thing I want to say is, you'd do well to avoid making a conclusion based on a single occurrence - e.g. "I retested this placement I decided to cut out before, and it ended up in profits, therefore my cutting rule is flawed".

Don't aim to catch all profitable offers/landers/placements/etc. when designing an optimization strategy. The higher the accuracy, the more it will cost you. Sometimes, the amount of money you can save on testing will justify cutting potentially profitable stuff prematurely.

This is true in general and not just for CTR. As for your question "When is it safe to cut placements based on CTR" - I can't provide a cut-off, because efficiency vs. accuracy is a spectrum. You'll just need to find your own sweet spot.

You're absolutely correct in your observation that lander scripts have a big impact on CTR, and therefore CTR wouldn't be a good metric to use for cutting stuff. I personally almost never base cutting decisions on CTR alone - and merely included that in my spreadsheet due to popular demand.

Here though is a scenario where cutting based on CTR may make sense: If you're getting a ton of traffic from a ton of placements, and need to cut very aggressively to save your shirt, AND you've already made a good number of conversions using one lander - you could look at placements that are profitable and observe a general CTR range they fall under, then cut all placements that fall outside of that range. For example, if most of your profitable placements have a CTR of 30-60%, then you may decide to cut placements that have CTRs outside of that range. Only using stats from a single lander is important - without that common denominator you wouldn't be comparing apples to apples.


Exactly - pop is a labor-intensive traffic type. Camps are short-lived so to make substantial profits, you'd need to have an efficient process that would let you juggle lots of camps at once.

Great progress! Looking forward to seeing you hit xxx/day in profits real soon!



Amy

---

Home > Paid Traffic Sources > Mobile