Home > General > Affiliate Marketing Forum

---

Pop Optimization Mini Case Studies - 1 (48)

---

(This is part of a series - please see here for details.)


Campaign Setup:

Vertical: Sweeps

Geo: SG

Traffic Source: PopAds

Targeting: All phones and tablets, all OSs, wifi+carrier, general categories (no adult), all connection types and speeds; also, started dayparting right off the bat to limit test budget by only running during 6pm-12am local time (this has become a standard thing to do for me as of late)

Bid: After I've selected the targeting above, I clicked on the "Summary" tab to check the average bid - I don't remember what it was exactly but it was 0.001x. I set the bid at 0.0022 so I would get high enough traffic quality to gauge promise of the offers.

(Note: I could have set up 3 camps at staggered bids like I often do, and picked the highest ROI bid to continue my testing, but when I'm testing lots of offers and geos I get lazy and just test one above-average bid. However, this way I could only see whether the offer converts or not - to accurately gauge how likely it would be for the offers and landers to make a profit, I'd still need to test staggered bids at a later point.)

Number of Offers: 2 very similar offers, from different aff networks. Payouts are $8 and $12.

Number of Landers: 6 landers ripped from Adplexity (basically I just put the "prize" name in the search field to get lander results, then picked ones that looked really different, downloaded them and fix the "errors" that resulted from the ripping - no modifications aside from that)




Optimization:

So I gave it a $40 budget just to test the waters. After the budget was exhausted I checked stats and found 1 lead made for $12. So I gave it another $40, and made 2 more leads as a result.




I compared the offers using my new favorite tool:

https://win-vector.shinyapps.io/CampaignPlanner_v3/





So I decided to cut the inferior offer.


Next I checked Offer -> Lander stats:



3 leads made by 3 different landers. Nothing to do here - will need to collect more data.


Next I checked OS data. Since I'll only be working with one offer, I can just check OSs for that offer to see more accurate ROI data:



Android is obviously performing very well. I could choose to keep running all OSs because I haven't spent enough money testing them yet. But because I don't particularly feel like spending too much time on this geo+vertical right now (my focus is on other verticals), I want to get green fast and scalp quick profits. If you're testing a lot of offers and need to build up cashflow, or if you're working with a limited budget in the first place, picking the most promising traffic segments to optimize would be a good strategy. So - I decided to just target Android from this point onwards.

An added benefit of only targeting profitable segments, is that it will allow me to bid higher to get more traffic volume.


Then I drilled down to "Mobile Carrier".



Nothing interesting here. All 3 leads were made on wifi, but then there's not been too much carrier traffic to tell how well the offers could convert on that.


Next I drilled down to Offer -> OS -> ISPNAME (this is a PopAds token - when working with other sources you can drill down to "ISP/Carrier" instead).



As you can see, SingNet Pte Ltd made all 3 leads, AND it's the ISP with the most volume. Again, I could choose to just target this ISP to reach green faster, and doing so will also enable me to bid higher to get more traffic volume. Alternatively, I could just leave everything running to collect more data, and cut non-converting ISPs.

Note that I could have just drilled down directly to ISPNAME and made the exact same observation, but because I decided to only run that one offer and only target Android, I drilled down to that offer and then Android to get a better idea on ROI. (Also keep in mind that the ROI may further increase after I cut landers and test staggered bids etc.)


I took a quick glance at placement data as well:



Even without drilling down to winning offer -> Android, we can see that there's no "leech" placements eating up budget. None of the placements have even reached 1x payout in spending so there's no need to even check calculators etc. Will leave them all running for now.



Next Steps:

So in Voluum I removed the underperforming offer, and in popads I selected only Android as targeting, and in the ISP section only chose to target SingNet Pte Ltd. Because I'm targeting the best traffic segments for this particular offer, and the ROI is out of this world, I'm changing the bid from 0.0022 to 0.0222 - yes a factor of 10 times just to collect data faster. I could have just run at the original bid to collect data slowly, but it would take longer to cut landers. Once I have a winning lander I'll clone the camp and test staggered bids, then keep all green ones running. I may be competing with myself a bit that way, but I find that having the same camp running at multiple bids can result in less performance fluctuation - although I don't have solid proof that this is actually the case.

Once I have a winning lander, I'll set up another camp to target all OSs and ISPs, and just run the winning offer+lander combo. This is because I haven't even given all OSs and ISPs a chance to convert in the original camp. I'll clone this camp at least twice and test staggered bids, then leave all green camps running.

Also - once I have a winning lander, I'll scale the winning offer+lander combo to all major pop sources, and target Android only to start. Will test staggered bids on these other sources, and then just cut placements now and then. If the camp on popads show that other OSs are profitable, I'll add those OSs to all scaled camps as well.



Other Notes:

-That offer I cut - I can only conclude that it's a loser offer when it comes to Android OS, but I haven't tested it enough for the other OSs. Right now I don't have an offer running for SG IOS traffic yet. If in the future I want to test more offers for this geo, I just may include this offer when targeting non-android OSs.

-It's important to note how important finding the right offer is! I've only tested 2 offers here because like I said, my focus wasn't on sweeps and I just wanted to set up a quick test because my AM recommended the offer so I quickly grabbed another one and just launched the test. After scaling, I should definitely mass-test offers.

This example clearly demonstrates what I mean when I say that a good offer will convert with an OK-lander, but not even the best lander will convert a shitty offer. Just using a few ripped landers is enough for testing offers to identify a good one. I could always test custom lander text/angles/themes, and/or variations of the winning lander, to further increase the ROI - because by then I'll have a high-ROI camp running on multiple sources, so the increase in conversion rate will result in big jumps in total revenue, which would justify the effort put into tinkering with landers.

-Remember I started dayparting since day 1 to get the most bang for my buck during initial testing. I will now remove dayparting and test all hours from scratch, and chances are all hours will be profitable in the end because margins are so high.

-And let's not forget that I should ask for a quality check and pay bump once I have a winning lander - because then I'll mostly likely have enough leads to ask for both.

-Most offers you test will not be such an obvious winner right out of the gate. This is one rare gem for sure. Thought I'd mention this so that you don't set an unrealistically high bar when testing your own offers - otherwise nothing would pass your stringent criteria. Most camps will require testing more offers and landers, and will involve some degree of cutting stuff before they can become profitable.


Although I haven't revealed the exact offers, you could probably figure it out based on all the other stats I've shown. I just didn't want to make it TOO easy - because then that would just end the fun for everyone within 24 hours.

Comments/suggestions/feedback/criticism good or bad - all welcome. If you'd like to see more mini case studies like this one please let me know as well and I'll try to find the time.




Amy

---

Just want to add that this is a CURRENT CAMPAIGN that is running right now. I just started testing 2 days back so these stats are fresh from the oven. Like I've mentioned, it wouldn't be difficult to find out which offers I'm running. You're welcomed to join in the fun if you like.



Amy

---

Thanks Amy for this great post. Just a quick question..

So when testing an offer like this, you always create one campaign (only) for both 3G and wi-fi (when allowed) traffic together?

Thanks.

---

Nice! A real time campaign case study!

Good luck to you vortex!

---

In this case, what we should do is keep testing different offers from different GEO to find as many Golden Offers as possible to bank the money. lol

---

Good question - it really depends on what your personal approach is and what you're wanting to achieve.

For this particular camp, I wanted to quickly test one of the offers because my AM highly recommended it - so I quickly grabbed a similar offer from another network and just set up a single camp at a single, above-average bid to see if the offer had any potential.

If instead, I was wanting to conquer the SG sweeps vertical+geo, I may have tested more offers and set up two camps to target 3G vs. wifi separately. i.e. Go all out hardcore.

However, seeing that SG isn't a big geo to start with, it probably wouldn't be worth the time to test offers for carrier/3G traffic separately. It would take the small amount of carrier traffic forever even to test a few offers.

So it would make sense to just test 3G+wifi traffic together, and only use offers that accept wifi traffic. Then if the camp becomes green, and I have it scaled to all major pop networks, I would MAYBE test carrier offers then, but ONLY IF the total amount of carrier traffic across all sources are even worth my time. If I choose to test carrier offers (of the one-click variety), I would most likely start a camp to target only carrier traffic, and if I find an offer that converts better than the current "general" offer (the one that accepts wifi), I would edit the tracker paths of all camps to reroute all the carrier traffic to this new carrier offer. Or I might even just set up new camps on all sources to target carrier traffic separately and bid higher than for the wifi camps, in order to get more carrier traffic volume.

There is SO MUCH we can do to optimize our camps. Yet, over-optimization can be a bad thing. I like to spend as little money as possible to test a wide variety of stuff in an efficient way (you need to have a system), pick the low-hanging fruit to optimize so I don't spend much time/effort to get green, and then scale quickly to all major sources, plus a few smaller sources that I know can give me higher ROI and more-stable camps that last longer.

But I digress - hopefully I've fully answered your question.



Amy

---

That's what I do: Once I have a good lander I mass-test offers, and find the best ones to scale.

For certain verticals, a single good lander can be used to test all geos.

Of course people can rip your lander, so you'll need to stay ahead of everyone else by testing more offers, and/or figure out a way to keep your landers from spy tools and competition, so that your custom/original lander will live longer.

There's always a way for a lander to be ripped and it's just a matter of time really, but if you can delay that as much as possible, and keep up with the offer testing and scaling, there's still money to be made.

Thanks for your comment.



Amy

---

Hi Amy, and thanks for this excellent post!

I'm interested in knowing more about hiding landers, especially from spy tools.
Do you have any tips or guidelines on doing so?
Thanks!

---

This series, although old, would be an excellent start:

http://stmforum.com/forum/showthread...s)-Part-1-of-3


Other than that, you'll just need to be creative and have good coding skills (or partner up with someone who does). I know you were expecting a more elaborate answer, but I'm not a technical person myself so I'm leaving this part to my friend who's a genius at coding, and I don't really even understand how he's doing it so can't tell you.



Amy

---

Thanks Amy, that's an excellent thread to start with!
I'm not a technical person either so I'll definitely look into hiring a techie to cook up something.

---

Very nice and useful thread! Thanks vortex!
It blew my mind how huge your CTR is... 42%!!! I thought for sweeps it was supposed to be way lower! I am averaging CTR of 2-3% in spain (i started testing the waters in that geo 2-3 weeks ago) for a low payout email opt-in iphone 6s offer.
Maybe i am having some problems with my landers, or maybe the difference in bot presence is very significative? I have heard theres plenty of bots in spain, but I have been cutting placements etc.
Maybe you are using a script like backbutton sends user to the offer?
Its even more surprising because you said you are using standard ripped landers from adplexity! haha
I know theres plenty of threads about improving CTR (have read a bunch and will read a bunch more now), but 40% difference in CTR is just too huge to not ask... Any theories? :P

---

Very useful case study Amy, thank you!! You posted a lot about how you test offers and landers, really awesome to see it in action with real stats.
Will you be updating this?

---

Hi Amy, you mentioned "errors" in the landing page when ripping it. Are you able to explain what errors these are?

---

Yup I was only using ripped landers - only corrected errors that resulted from the ripping. Didn't even bother adding additional code snippets. Nowadays most landers you rip will already contain all the standard bells and whistles anyway including backbutton, vibrate, sound, etc.

As I've mentioned, my experience is not in sweeps. However, based on what I have observed so far, lander CTR can vary by a LOT from one geo to the next.

The absolute worst I've seen is AU. I've followed some follow-alongs where members were running AU sweeps and were getting such horridly low CTRs - like < 1% even, that I asked them to send me their landers to run tests myself, to verify there wasn't any technical issues with their setup.

Therefore, I'm assuming that by running in SG, that you'll get similarly high lander CTRs.



Amy

---

I might, but no promises unfortunately.

Sweeps really isn't my current focus. I was just testing some offers for a case study I was doing. As bad as it may sound, due to lack of time, I'll likely end up letting my current budget run out and leave it at that.

The main focus of this case study was to show the optimization process, and not to flaunt how profitable I could get it to be. To that end, the original post has served its purpose.

Thank you for the encouragement - everyone's kind comments will motivate me to post more mini case studies like this one. Much appreciated!



Amy

---

That's a good question! I'll try to write a separate post on that when I have a chance. Please stay tuned.


Amy

---

Def looking forward to read more on this!
Thanks for the CTR answer =) Now i feel better about it haha.

---

As promised - here's my post on the topic:

http://stmforum.com/forum/showthread...Ripped-Landers


Right now it's only got 4 fixes in it. I'll add more over time, and hopefully coding experts will contribute to the list as well!



Amy

---

Hey Vortex, thanks for the case study.

I see you set initial budget 40$ and thats really small for 8$ and 12$ payout offer. If offer would not convert on 40$ you would not have accurate data. Would you continue with spending 40$ more or would you stop your campaign? For decent acurate results you would have to spend about 200$, no?

---

It really depends on your offer testing strategy, and stuff like whether you're using a proven lander.

You could test a lot of offers, and set a lower budget for each, find the low-hanging fruit to optimize.

Or you could test fewer offers but more thoroughly.

It would also depend on how confident you are that the offer will convert. For example, if you're really confident because the offer came highly recommend by one of your buddies who's running it right now - then you'll want to test lots of landers and bids etc. to make it work.

When you're testing a ton of offers, being absolutely fair to each offer by dropping $200 on each may not be the most practical approach. You may decide to spend less on oeach offer and just test more offers. When looking for low-hanging fruit, I tend to test wide and allocate lower budgets to each offer.

And if you you're using a proven lander, then you really don't need to set a high budget at all. Whereas if you're still testing landers, you only need one lander+offer combo to work in order to profit, so in that case you'd want to set a higher budget definitely.

Thanks for your question!



Amy

---

This is really a good one! Thanks and look forward to reading next mini case study.

---

42% is completely insane, and its the average of all the ripped landers! I take it all of the landers had a backbutton redirect to the offer, counting each back button click as a lander click? I just now turned the plug on a SG campaign, and had CTR's ranging from 1-4% (back button went to another lander though, not the offer). My landers were ripped but optimized for speed since most ripped landers are really slow. Clearly there's an elephant breathing down my neck, and I just can't see it...

Do you find back button linking directly to the offer and not to another lander to be better? Also, do you usually prefer to run multiple landers when testing offers?

Wonderful post btw.

---

1-4% CTR?? Something must be wrong!

Let me give you the 2 landers that have performed the best so far for this camp:

https://www.dropbox.com/s/02w8e5t4sy...nders.rar?dl=0


You can calculate the CTR from the stats shown in the screenshots (EDIT: Please see CTRs from the new screenshot posted in the next post below). My setup is nothing special - just using amazon cloudfront as CDN. So if you run these same landers and don't see the same CTRs, there must be something wrong with your setup. We'll take a look at that later if that turns out to be the case - please start a separate thread and I'm sure more tech-saavy members than myself will be able to help you with trouble-shooting.



Amy

---

QUICK UPDATE ON CAMP AS REQUESTED:

I just did a quick check on stats - nothing special. Camp is still profitable but there's too little SG traffic on popads. Maybe I'm not bidding high enough - I haven't had the time to test staggered bids or anything. And I haven't had time to scale to other networks either. I didn't even remember to add more budget to this popads camp after the second budget ran out...


Anyways here are the stats so far:




And in case you didn't read the last post - here are the 2 current-best landers for anyone interested in playing with them:

https://www.dropbox.com/s/02w8e5t4sy...nders.rar?dl=0


I'm really not doing this offer justice. My focus and energy is elsewhere. But hopefully some of you will use these landers to test more SG sweeps offers and find a good one to scale to all major pop sources. I've only tested a couple of offers in total. It's highly unlikely that with such a small net, I've caught the best offer currently available for this vertical+geo.

Good luck!



Amy

---

Hi Vortex,

Awesome case study! I got a few questions tho... I tried replicating your popads campaign with identical targeting and bids, however the traffic estimate says i'll be getting none. I've even increased the bid to .09cents and still the same traffic estimate. Thoughts

---

LITERALLY none? Or very little? If the estimator isn't showing ANY traffic, please include screenshots of your popads camp settings and I'll do a comparison and let you know.

There isn't much traffic for this geo on popads. You're definitely encouraged to test elsewhere - I haven't scaled ANYWHERE due to lack of time.

The aim was to show an example of the optimization process, and not recommend the geo or vertical - I may have just hit on a good offer.



Amy

---

There was a internal system error with popads and I reached out to them and it should be fixed by now

---

Ah - that would explain some things wouldn't it? Thanks Simon!


Amy

---

Very useful case study Amy, thank you!! !

One quick dumb question, why you didn't put Windows OS into white list?

Is this because the offer only accept mobile submission? As if you included Windows you will get more traffic and when people having proper keyboard they are more tend to finish those long forms?

Awesome stuff really!

---

Thank you very much for reading! And it's a mobile offer thus no windows OS.



Amy

---

Amy, first and foremost - nothing is more impactful that a thorough case study, many thanks for the insights.

I had two questions.. the first regarding optimizing only the best performing metrics. I have accounts with several pop networks but have only worked with about 4 of them thus far. To illustrate a specific example, if only Samsung/LG phones were in the green and there was enough volume in that geo to support going after them - are there many pop sources that allow for that specific of targeting, such as PopAds? Or, what is a better way to approach the optimization for scalability - at the OS level? Same could be said for ISPNAME.


Thanks!

---

That's a great question!

First of all, I try to identify the bigger traffic segments instead of picking the smaller ones. For example Android or IOS, or a major Carrier. That way, you can ensure there will be many traffic sources that allow for your desired targeting, that you can scale your camps to.

Secondly, if you are wishing to scale to a traffic source that doesn't have your desired targeting, the alternative would be to set up tracker rules to reroute the "other" traffic elsewhere. For example if you are wanting to target a specific mobile carrier, and the TS doesn't have carrier targeting, then you will need to somehow figure out a way to monetize the other carriers, either by redirecting to a traffic monetizer like afflow or ytz, or by testing and finding other offers that perform well for the other carriers. The end goal of course is that the campaign overall will need to be profitable.

Thank you for your question.



Amy

---

Hi Amy, a question on your methodology here insofar as setting up your Popads campaign targeting:

For initial testing (sussing out if the offer's viable), do you tinker with "Primespot", "Quality", or "Adblock" settings?

Also, what's your philosophy on popunder, vs popup, vs tabup? Obviously it would be most thorough to set up a campaign for each, but do you think it's reasonably safe to just pick one initially?
It could get expensive to test 3 bids @ popunder, 3 bids @ popup, and 3 bids @tabup -- but out of those 9 combinations maybe only 2 are viable, and we could easily miss them and conclude the offer's no good if we were only testing 1-2 permutations.

---

I actually leave primespot and quality settings at default. (As for Adblock - haven't had a chance to play with that yet since it came out.) Although, I suppose if you're wanting to see whether an offer will convert or not, then it may make sense to just target for example only Quality 10 placements. If the offer still doesn't convert on this high quality traffic then you know the offer's probably bad.

The only downside to doing it this way, is that of course the cost would be higher too.

Also, if there isn't a whole lot of traffic for your targeting, then it may make more sense to target broad even during initial testing, so you could get an overall picture of the campaign's profits potential. In other words, even if by targeting just Quality 10 placements, you see good results, doesn't mean you'll still see good results when targeting lower-quality placements. And for smaller geos, targeting just Quality 10 placements will likely not get you enough traffic to make satisfying amounts of profits.

As for popup vs. popunder vs. tabup vs. tapunder: Do remember that no matter what the ad format is, you'll be dealing with the same visitors who will react similarly to your lander and offer. So if your lander+offer sucks balls on popunders, the same funnel will probably not perform too well for the other ad formats either. I have a habit of testing popunders first because that's where most of the traffic usually is, and then would only scale to the other 3 if I get my popunder camp profitable.



Amy

---

very good post!!

looking forward to more mini case studies like this.

thanks!!

---

Hi Amy,
thanks for the study, real nice!

I'm amazed by the CTR, I think something is off there (maybe the back button redirecting to the offer, a timer to the offer, a full-page click...) and that is not real (that's my experience. I had a timer to redirect to the offer on an AV one and it was around 40% ctr, the rest 1-5%)

Also, I think popup/popunder is pretty important too. With popup, the page has to load FAST, because the user sees it loading. And it's easier for the page to 'kick-in', if it's good. With popunder the user might not even see the page (if closes the browser after whatever was doing) or see the page N minutes later (thus the redirect to the offer on time not working very good).

After not being very successful on several tests, I think that putting yourself in the flesh of the user and understanding what's going on is pretty important, instead of just throwing shit to the wall to see what sticks. How is the network creating the popup? what kind of pages? what's the user connection like? popup/under? psychology of the landing/offer, etc.

Thanks again!

---

I don't really remember lander details - but as you pointed out, CTR can be inflated quite easily. This is why I'm always advising against cutting things based on CTR (banners, landers...) The exception is when CTR is so low that there's no way the math will work out to green. Another exception is for placements, where really high or really low CTRs can point to underlying issues - such as bot traffic. I've shared a couple of the best-performing landers for this camp in the post below:

http://stmforum.com/forum/showthread...l=1#post276677


Hm! Good point! Never thought about that! Thanks!


This is very true!

Although - you could argue that for pop traffic, throwing shit against the wall MAY be enough to get results. Everyone's ripping everyone else's landers anyways, so basically you could just rip a bunch and test lots of offers, find the good ones and scale. It's not real marketing I know, but unless you know of a way to keep your original landers out of spy tools, creativity and originality may not have as much of a place in pop...

But yeah I totally see what you mean. Sometimes we forget there are live people on the other end, and that we're dealing with more than just stats.



Amy

---

This is an extremely great guide! And also definitely very helpful for newbies like myself

I'd like to clarify something,

I am following my campaign creation and optimization as per your case study,

So currently I have 2 AV offers: A, B
Each offer I am testing: wifi, carrier

1) When someone asked how much should your budget be to test the offers, I understand you said in a previous post: "It really depends on your offer testing strategy, etc"
still as a general guideline, How much is a good amount to start testing the offers to see if they convert? 10x payout? (I'm doing AV offers in VN)

2) And do you recommend testing for wifi and carrier at the beginning of the campaign?
So it means I have 4 different campaigns now:

Offer A, Carrier
Offer A, Wifi
Offer B, Carrier
Offer B, Wifi

3) Also, what would be your best advice to me if I want to improve and get results from all the tests as fast as I can, so that I can start seeing greens asap? (Without burning up all my budget fast)

---

1)Assuming you're setting up 3 camps to test staggered bids, and assuming you'll be running 10-20x payout to each camp, that would be 30-60x payout in total, just to get a VERY rough idea. And this may or may not be enough to gauge potential of the offers - will just need to take a look at stats then and decide whether to run more traffic to test further, or not.

2)Just setting up 2 camps would be fine - a Carrier and a Wifi camp. You'd be rotating both offers in each camp. Running the offers in separate camps will subject them to different traffic (from different placements etc. etc.) which would make for an inaccurate split-test.

3)I'm assuming you're referring to pop camps? If not then the following strategy may NOT work well.

Pick a popular vertical that has lots of offers (look through top offers lists from multiple networks). Choose several Tier 3-4 geos to test at the same time (Asian, Latin American, African). For each geo ask for 3 offers recommended by AMs (that other affiliates have gotten good results with). For each geo, go to adplexity and rip as many different-looking landers as you can find for that offer type. For each geo set up 3 camps at staggered bids (low average high) and run 10-20x payout to each. Pick the highest-ROI camp to continue testing and cutting offers and landers. Once down to last offer and lander, mass-test offers using the best lander.

At any point if a geo looks hopeless, abandon it. Basically pick the most promising geos to optimize and scale.




Amy

---

Thank you so much! I've did some research,

and I discovered Sexy Ladies erotic content Mobile PIN in the regions of LATAM, Asia, South Africa

1) Based on your opinion, Would that be a good choice?

2) And is it considered Adult vertical? Because I saw on the thread below that Adult is now really saturated and near-impossible to complete
http://stmforum.com/forum/showthread...ticals-EXPOSED



Edit:
The reason I asked the question is because now I am currently running AV offers on POP, and I can see that the market is really saturated with all the same offers

And from my follow along, Attila actually advised me on my follow along >>> http://stmforum.com/forum/showthread...l=1#post285325

that running AV offer on pops is what everyone else is running, and it'll be really tough for me unless I have a huge budget (thousands!) myself

So my original plan was:

- I do not mind spending my first few hundred to test out AV offers to get my feet wet and really learn the whole flow of optimizing campaigns, etc

- However, I would also like to see some profits at the start as it will be very encouraging for myself, so I was wondering if you have any personal recommendations for offers in the tier3/4 geos and payouts below $1?

I have asked my affiliate networks:
-F5
-Adsimilis
-Clickdealer
-Avidi

And most of their offers are as such
Vertical: Sweeps, PIN submits
Geo: US, UK, DE, etc etc (not too many in LATAM, Asia, South Africa)
Payout: All ranging from $3/4 and above

Which I feel payouts are pretty high for a new person, and I've always read that we should start with low payout offers first

So I'd like to hear your thoughts on this, Ms Amy!

---

I've never run that particular offer - so really can't tell you. Why not just set up a test?

If it's an erotic content offer, then yes, you probably can only run that on adult traffic.

Adult IS saturated, but only for adult dating I think. That was what I started with, and it was tough then and could only have gotten tougher since then.

However, I'm still making money from adult traffic now, just not from adult dating offers. Stuff like video subscription offers, gaming, and antivirus do well on adult traffic. And if you have a larger test budget, penis enlargement offers (pills, extenders, manual exericse ebooks, pumps...) can do quite well also.



Of course Attila knows what he's talking about. But I'm suspecting that for a new affiliate like yourself who is not doing large volumes at the moment, it may be difficult to arrange for advertisers to provide whitelabel solutions that are customized for you.

As you've pointed out - pop traffic is quite competitive, and many people are running the same offers. This traffic type is also very volatile - one day you'd be at 300% ROI and the next day it may drop to negative. As a new affiliate without a significant edge over the competition, you'll just have to hustle harder and/or work smarter. This means you'll need to either:

1)Have a way to automate campaign optimization, and/or

2)Have a team of people working for you, and/or

3)Have an efficient testing, optimization and scaling process, and/or

4)Be willing to cast a wider net and test more stuff than other people - for pop this means mainly testing more landers and offers.

As you're a new affiliate, I would definitely recommend that you start with point 4), work your way towards point 3), and after that, consider points 1) and 2).

Better yet - once you've gained some experience buying traffic, MOVE AWAY FROM POP, and expand into mobile display, native, or facebook.

I'll see you on your follow-along thread!



Amy

---

Amy, amazing like always!!!

What do you think its the best way to use a back button redirection? Directly to the offer, to another campaign with different landers/offers?

Thanks!

---

One word: Test!

You can direct the visitor to the offer, to another profitable camp if you have one running for the same geo, or to a traffic monetizer such as YTZ or Monetizer.com/Afflow.

Some people set it up so that the visitor will never run out of offers to "back up" to - they'll arrange several camps in a closed loop. This is against the terms of some traffic sources though so be careful - people have been banned for doing this.


Amy

---

When you have a JS alert, the other JS script including the countdown shouldn't start until the user has clicked on the JS alert popup. Correct me if I'm wrong.

---

Great post. Learned a lot. Thanks a lot Amy

---

You're right - but then in that case, the popunder won't have any chance to lurk until it's discovered - the js alert will force itself on the visitor.



Amy

---

Hi Amy,

Been reading a few of your threads, really great stuff, thank you.

Quick question - your original stats are showing a ~40%+ CTR on the landers. Are you including the history/back button trick in those stats?

---

Yup I'm sure some of the landers I was using had the backbutton script in them! Backbutton has become a standard - it's basically on almost every lander you rip nowadays.

I shared the best landers of the bunch I tested, in this post:

http://stmforum.com/forum/showthread...l=1#post276676



Amy

---

Home > General > Affiliate Marketing Forum