Home > Mobile > Follow-along Campaigns

---

It's just a numbers game - nomeus chasing green campaigns! (23)

---

After launching my first campaign grinded through The Mobile CookBook and decided to do a different offer but with a proper setup.

As AM suggested:
Offer - Sweepstakes (iPhone Giveaway)
GEO - AU
Payout - from 2.00 - 2.80 depending on the offer (Have not decided on the exact one)

Decided on 2 angels, Made 10 banners for one and 4 for the other. (6 more to go)
Will get everything approved and set up the campaigns tomorrow.

One question. If AM suggests offers that a lot of people are successful with, does that increase the probability that a lot of your conversions will be duplicates, and I might not get paid for them? What I mean is that a lot of people who signed up through my funnel might be also signed up with different affiliates funnel, right?

---

Spent the weekend on reading a lot of tutorials, follow along's, tips and tricks and I read a lot that sweeps don't really work well with banner ads. Found one specific follow along that tried exactly what I am doing and after couple weeks of testing author came to conclusion that these offers are not doing well with banner ads. But I see that a lot of iPhone sweeps offers allow only banner ads and AM suggests exactly those as the one which are doing the best. How do they do it then? And whats the best way to approach sweeps? Pop ads (most of the sweeps AM suggested are against those)?

---

You can try PPV traffic with an intermediate page. Most chances are the advertiser wont be against it and its probably better to run that way anyway (but ask your AM if you arent sure).

PPV (or Facebook) with a good LP is a good way to go about Sweepstakes offers.

Do some spying and see. You can also find such info on this forum.

---

Every offer is different, so you gotta check with the AM to make sure what is OK to push and what is notm, otherwise you risk not getting paid. Some will allow pops, some will only ask for banners ... you simply got to obey the rules.

Sweeps are pushed on pops a lot, simply because its easy to setup, so plenty of people use it as the first vertical and traffic source to play with. There is also a lot of volume available, you dont need banners ... sweeps are hot right now so you will see them everywhere. This doesnt mean they cannot work on banners, you just need to select the right spots and play with the targeting.

---

Does the offer allow adult traffic? If not then no dilemma at all.

I think adult traffic is not smart with this offer anyway. Think about it - you "browsing" an adult website, do you really care right now about groceries? :-)

---

Offer allows adult traffic. Canceled for now!

---

Thanks for inviting me to review your following-along nomeus - will do my best to provide suggestions.

You've already got some valuable feedback from matuloo and Oran. Limiting to pop traffic and non-adult (mainstream) is a good idea. Now you don't need to test banners. Once you have a good offer and lander, you can always scale back to display then.

Are all those low-bid camps? I usually suggest testing several bid points (low average high) to see which one will give you better ROI, so that you can do your subsequent lander and offer testing at that bid point to save on test budget.

However, if you feel that your Mob+Wifi camp is at an ROI you can tolerate, you could just use that camp for the rest of your testing. I don't know how long it took you to collect this data, but if it was over the span of half a day or more, then none of the other camps was giving you enough traffic. You would want to pick a camp where you're getting enough traffic and conversions to allow you to gather data faster.


I've verified using the split-test calculator that lander V2 can be cut. However:

1)"How to Cut Placements, Landers and Banners - Part 2" isn't for cutting landers - please see Part 1 instead.

2)One of the landers only had 2k+ impressions whereas the other 2 got 4k+ impressions each. You didn't run them at the same time did you? When running a split-test, refrain from adding candidates before one round of testing is over. Otherwise, you wouldn't be comparing apples to apples. When you add a new lander to the testing - the stats collected before that point cannot be used, because the conversion rate can be very different from day to day. For example, if you started testing 2 landers yesterday and added a new lander today, and yesterday was a really good day of the week where CR was high, and today is a bad day of the week where CR is low, then the test results will be unfair for the new lander.

Also, for certain offers, performance can be very different for wifi vs. carrier traffic. If would be a good idea to collect more conversions, then in your tracker drill down to "mobile carrier" stats to see how different the CR is between the two types of traffic. It may be a good idea to test them in separate campaigns later.

The same can be said about OS - the same offer can perform differently for different OSs.


I don't think you need to launch more camps (unless you have a good reason to). You have an offer that seems to be converting. I would suggest running your current 2 landers to see if you can cut one more, and then add those 5 landers in to test with the winner, to start a new round of testing. After that testing is done, test the current offer with a bunch of other offers (same type). If after that you're still not at least breaking even, I would try another geo.


Have fun - great follow-along so far! Looking forward to following your progress.



Amy

---

Nice! So the trend is that this offer is doing a lot better on carrier traffic than wifi traffic. It would be good to focus on carrier traffic for now. Although there's not much traffic for carrier, AU is still a pretty big geo, so by scaling to all major pop networks you can still make some profits.


A lot of new people think that by bidding higher, you'd be losing more money in exchange for more traffic. This is often not the case. One important thing you need to know, is that the best placements are in high demand - when people see that they convert well they'll bid higher to get more traffic from them, which will drive up the price. So when you're bidding really low you'd be getting traffic from low-quality placements - ones with bot traffic or traffic that otherwise convert badly. Not to say bidding low is always bad, because sometimes the low cost can offset the low conversion rates. You wouldn't know unless you test different bids. This is why I always suggest to set up at least 3 camps (exactly the same in targeting) at different bids - low, average, and high - to see which bid does better.

Personally, for a camp that's already showing promise, I like to set up 5+ additional camps with different bids to get an even more complete picture of the bidding landscape. It's up to you how many you want to set up. For a geo like AU there should be enough traffic volume to set up quite a few camps. Just remember to set a low daily and overall budget for each camp to avoid any surprises (e.g. $5-10/camp).


Since you now know that carrier converts better than wifi, you can just leave that mob+wifi camp for now.


Gaming can work for TH on pop, but then of course it would really depend on the particular offer.

A better approach would be:

1)Test multiple offers from the start.

2)Test several different bids.

3)Drill down to OS -> carrier -> offer to see if any of the combinations is even showing promise for any one offer.


For something like gaming, direct-linking may be enough to give you an idea on which offer at least has promise. However, to be able to even get a rough idea when you drill 3 levels deep (OS -> carrier -> offer), you'll need to run a lot more traffic than just the few dollars you ran, to gather enough conversions to see any sort of trends. I DO agree with you though that your offer may not be a very good one. This is why testing multiple offers would be a good idea.

Some people just pick an offer and try to make that offer work. A better way would be to pick a geo and vertical that you know is a good combination, then test a variety of offers and landers to try to make that traffic work. For a high traffic volume geo like TH, all you need to do is identify one major traffic segment (e.g. one carrier and/or OS) AND one offer that does well for that segment, then scale that to multiple networks, in order to make money.


Keep up the great work!


Amy

---

Hey Amy,

Awesome content! Thank you Amy, I will keep everything in mind on my next campaigns. I kind of want to stop this campaign because there was a lot of errors in the process and it is hard to repair whats been broken since the beginning. Ok Ill do my best to repair this campaign

---

March 17th - ROI +0, Lets Add Some More Campaigns

First of all, to Amy, In the previous post I was in a hurry so I was not grateful enough. Thank you very much for taking time to read everything I have written. I have a question about multiple bid bidding. How close should those bids be? If my current bid is 0.008, the average bid is 0.0337 (tried that, a huge waste of 20$)? Like 3 campaigns @ 0.005, 0.008 and 0.013 for example? Or should they be further from each other?

Campaign #1

This campaign is actually breaking even right now. If I could just double the CTR, then that would be more than awesome and I could get it profitable. Then I could increase the bid and maybe

I added ZeroPark traffic in the mix just to see what happens, and after two days, it is braking even too. Should try to double the CTR and then it could be more green than now.

Tried increasing bid on PopAds, but that resulted in the same amount of CTR and the NET result was worse than on lower bid.

Campaign #3

Decided to try another campaign.
Offer - iPhone Sweep
GEO - ES
Payout - 0.8$-1.2$ (two offers from different networks)
Traffic - PopAds

Did a lot of errors in the beginning.
1. Didn't double-check the offers links. 1 conversion was not registered.
2. Didn't separate WiFi from Cellular campaigns (since the last WiFi campaign performed much worse from the start)
3. Set the mid bid too high and got 70% of my initial testing budget get eaten in 2 hours.

Enough talking, here are the stats:


Landers:


For my next testing round, I created two campaigns WiFi and Carrier with low bids, (I will try mid bids a bit later), cut Lander#3 forever and Lander#2 for now, maybe I'll test it again later. Added one more to the mix.

Need more data on offers to decide on which is better.

Thank you for anyone reading this journey, I hope I can provide value to those who are in my shoes and just starting out! Learn from my mistakes!

---

When a campaign is breaking even, while the CTR is still important, but you're literally a few blacklist away from a profitable one.

Remove placements that are doing horrible.
Remove ISPs/Brands/OS/browsers etc that are doing horrible.

You instantly now have a profitable campaign

---

March 18th - Confusion and Poor Friday

I dropped that AU campaign purely because I could not get more traffic for a good price. But thats ok.

Now I have a -50% ROI campaign!

Campaign #3

After two days and 40$ later I saw that PopAds websites below quality 5 have eaten half of the budget and haven't brought any conversions, so I cut those off. All the other placements where actually green! Hoped that today will bring some good news. Boy was I wrong!


Is it because of Friday? You cut half of what was draining you money, and the next day its even worse! I guess thats what affiliate marketing is about!

And I am really confused now, little help needed. I cut one lander that was doing the worst. Now I am left with these three.


Although significance calculator says that it is not yet clear which is the winner, it is very tempting to cut #3 just because of the poor CTR. For some reason I think giving those views to #1 or #2 would bring my ROI a lot higher. Am I right, or should I just give them more money to drain? Or maybe add one that is totally different from these three? (these are totally different from each other)

And what about the weekend? Should I continue my test or is the weekends such an anomalies that I should pause and wait for a Monday?

Thanks for reading, and happy weekend you all!

---

Haha - not grateful enough? Hardly! Your gratitude is felt and reciprocated - thanks for being a valued member. And you're welcome - helping members is my job and is something I enjoy doing. And it's always nice to see someone start a detailed follow-along for the benefit of everyone.


There's no definite way of choosing bids. The only suggestion is to not choose bids that are too close together or else they would be competing against one another for traffic from the same placements. So what would be considered "too close"? Unfortunately I can't answer that question either...

The bids you chose look fine. The fact is, the more bids you test, the clearer the picture you'll have of the entire bidding landscape. And it's hard to say which bids you should choose. If budget allows, and if you're sure you have a proven offer, by all means test more than 3 bids - 5, 7, or even more!


CTR isn't the most important metric. CR is.

Having said that, if you can increase CTR without negatively affecting CR, that would be a great thing!

For a ton of advice on how to improve your landers - click here and scroll down to the landing page section, and check out some of the threads listed.

alfiss makes a good point about cutting stuff now that you're breaking even. Alternatively if you want, you can keep optimizing landers and testing offers to get higher ROI and THEN cut stuff. Personal choice really.


Based on your stats, you can pause the mid-bid camp and just use the low-bid camp for testing. The 2 landers with 0 conversions are ready to be cut. Just be careful about adding new landers - finish one round of testing before starting another. Once you add a new lander it starts a new round of testing, and you can't use data collected from before anymore.


Looking good! Should see green very soon!



Amy

---

Those were yesterdays stats Todays stats were a little different.

But what about the traffic, when I choose low bid like in this example? I would get like 6000 impressions and after cutting the unprofitable placements in the best case scenario I would be left with 3000 impressions and get like 5 conversions and that would be just 5$ before traffic costs. That doesn't sound much. :/ I guess I need to find offer that nobody is doing.

---

Right now your aim is to test and improve your funnel (i.e. find a good offer + lander). As long as you're getting traffic AND conversions quickly enough for you to test offers and landers, that's all you care about at this stage. Once you know which lander + offer will give you good ROI, you can bid higher to get more traffic, as well as scale to other sources.

If the testing isn't going fast enough for you, you can choose to run at a higher bid to collect data faster, but of course the price would be to run at a higher loss. Alternatively, you could look for another traffic source that has higher traffic volumes for your target geo and do your testing there instead - assuming the traffic quality there is at least as good as your current traffic source.



Amy

---

March 24th - Confusion, Confusion and some more confusion

So where do I start.. Affiliate marketing can be very very very unpredictable. No stability what so ever.

March 21st
Trying new offer. Just wifi test at popAds, since the carrier performed very poorly. 5 landers, 2 offers.


Some promise, but not much, worse than before, even after cutting quality 1-5 that was not doing any good.

March 22nd
Since I had some budget, decided to try out ZeroPark with the same landers and this happened:


Popads suddenly just dropped dead. Not getting any traffic and aused at 8.00PM without any conversions. My mistake. Zeropark avg bid was just too hungry and decided to leave just low bid.

March 23rd
Since zeropark performed really well, decided to try to make it work. Took out one of the landers that did not help. Next day:


Zeropark much worse than before. Traffic costs magically doubled. Lander success pretty much even, so need more data. Looks like PopAds woke up and now I have some data to work with.
Zeropark cut one more lander, made 2 variations of the two best ones. Lets see how that goes.
PopAds just two landers (over a course of 4 days) was working well, so decided to put that those variations in here too.

March 24rd
Since I kind of feel good about PopAds decided to duplicate campaign to a bit higher bid, to get more traffic. And today this far have the worse results ever. I really hope it gets a little better today:


Seems like the more you try to improve, the worse results happen. You get a lot less traffic, it gets more expensive after some success, and results just drop.

Campaign #4
Week started OK, so decided to get one more offer going - Antivirus Pin Submit in tier 3 geo. 9 landers. 2 offers (waiting for approval from the second network).
Since budget allows me, tried to do 4 campaigns, two at ZeroPark two at popAds, here are the results this far:


Its evening in the GEO, so I guess results might change a bit, but the general feel of the offer is there.
PopAds avg bid campaign run through $15 in an hour, but actually showed some promise. Not too bad. Good to know that there is ton of traffic available.
ZeroPark avg bid is workable too. Will see how that ends.
For the low bids, not much data yet. Have to wait until the day ends.

Conclusion and some thoughts.
Affiliate marketing is superunpredictable. What worked yesterday, might not work today. Data and lots of tests are the only way to go.
To anyone who is just starting out, just as me, I would suggest, dont rush your decisions and dont assume anything. I thought that with this campaign #3 I wount try to rush things, but I still did, and that will make this campaign harder to repair.
Sometimes it is very discouraging to see some small promise one day, then make some improvements, and see your results shatter on the next day, but I guess thats just the part of the game.
I think I will stop my Campaign #3 through easter time and reevalute what I have done and decide on my game plan next week. Until then, have some more reading and learning to do!

---

I'm not seeing you mention cutting placements at all in this list - I'd definitely recommend at least keeping an eye on the placements you're seeing, and preferably cutting the real no-hopers. Vortex wrote a great guide on this here: http://stmforum.com/forum/showthread...Banners-Part-2

For example, the sudden drop in ZeroPark revenue could well be a result of a new placement suddenly surging up the spend chart on day 2, and pushing out a previously-profitable placement.

---

Campaign performance can be unpredictable, yup - I hear you!

There will always be day-to-day and hour-to-hour fluctuations. Your visitors will usually convert better when they're not at the office working for example. And you'll get traffic from different placements in different amounts during different times - because your competitors have dayparting set up and are also changing their bids etc. etc.

This is exactly why it's so important to focus on improving your funnel! Test more offers and landers until you can make enough of the traffic profitable. When you're testing several landers and rotating them all equally, the fluctuations mentioned above will apply equally to all of them - at least that's the assumption we can make here.

So by testing offers and landers and cutting using statistical methods, your funnel will improve, which means your overall conversion rates will improve.

You'll still experience fluctuations for sure! But with a bad funnel your CR may fluctuate between -60 to -80% ROI, whereas with a good funnel (after testing different offers and landers) you may get +20 to +40% ROI.

Therefore, I would recommend focusing on testing offers and landers, and not getting held up by the actual ROI. At least for the moment.

You've already tested a couple of traffic sources and several bids at each. That's fantastic! Now just pick the one that's giving you the best value and do more testing there to improve your funnel.

Also: Now that you have some conversions, it would be good to drill down into data to see 1)which traffic segments are giving you the most traffic, and 2)which of those major segments are converting especially well. For example, if you find that android and/or carrierX is giving you good CR and OK traffic volume, you could set up a camp to target just that OS/carrier, improve your funnel until it's profitable, then scale.



Amy

---

March 29th - From first sight of green to frustration in just couple of hours!

Following Amy's advice, decided to continue with just one, most promising campaign:
March 25th


I took 2 of the best landers (since they are performing very similarly) and made a variation for each and did a little test run on Saturday. These were the results.


Not bad. Not bad at all for sample Saturday test. Lets wait for the week to begin.

I changed campaign link on Monday so PopAds stopped my campaign (my error) until they checked the new link. So not much happened on Monday.

March 29th
Today woke up to a very promising start (about 2PM in target GEO):


And then for some unknown reason conversions just stopped and now this campaign looks like this (8PM in GEO):


Only thing that I have cut is one placement that had 0 conversions after like 7k views. I just try to work with landers to get close to breaking even.
Besides that, I get about 40% click-loss from lander to offer.
It is just frustrating! How can I improve anything if results from day to day are so inconsistent? One day from 5-8PM campaign get 10 conversions, next day it gets 0 in that time. What happened? More people are running this offer, so it is getting more and more competitive?

Honestly, I don't know where to go from here.. One part of me just wants to cut this offer and look for something new and start everything from the beginning with a different offer, other part of me wants to try to make it work, since I know that offer is converting.

I will look into this campaign, since I have more than 150k views, and see what I can save from it..

---

Looked at some placements, and it seems that there are certain placements, that get a lot more traffic for the bid I make. I would like to think that that is because not a lot of people target that placement, and that would mean that it is a crappy placement. At least thats how I understand that.
(last 3 day stats)


Drilled down by landers/os, landers/carrier, not much green there, so decided to put that placement on hold as well as 651501. Will see what tomorrow brings.

---

Yup - I agree that we should drill down to your existing data to take a more detailed look at what's going on, to see if there's anything to learn from it.


So if you'll provide me with some screenshots I could take a look at them for you. Please pick a campaign that has the most number of conversions, and show me the following data:

1)OS -> Landers
2)Landers -> Offers
3)OS -> Offers
4)ISPID -> Mobile Carrier
5)ISPID -> Offers
6)ISPID -> Mobile Carrier -> Offers
7)OS -> ISPID -> Offers

Let's start with these. I don't need to see ALL the expand data - if in each case above you could show me the top few entries with the most traffic, I may be able to get a general picture.



Amy

---

April 19th - Just a little RANT

Long time no updates.

I really wanted this month to be green (or close) and then just drop the screens here, but as close as I get to any green camp, I get slapped in the face by AM or advertiser.

I feel that I have the ability to take camp to green, but then.. The offer gets pulled, landers get banned, and every other thing, that should nothappen, happens. And it is SO discouraging.

This month has not been too good:


Today was promising until it turns out all of the landers I got from Adplexity are banned from the GEO.


Even using RED color or countries flag is not allowed. Those rules are just *%$@.

Just double standards, I created thread about ethics of PIN submit offers and we discussed how wrong they are if you look at them from certain perspective. And then, lander that has testimonials where people just say how much they love their iPhone gets banned. But for some reason "I didnt thought this offer is real" is allowed..


Maybe I should try different offer types?

---

Yeah I hear you - offers getting pulled, creatives getting rejected by the advertiser/traffic network etc. are a part of the game. Once you have a good campaign strategy down you'll be able to juggle multiple camps on multiple sources so you'll always have something profitable running in spite of the fluctuations and things dying on you.

Why don't you just ask your AMs for offers that ALLOW aggressive? That way you'll save yourself a lot of grief. Plus you won't need to worry about having to cloak etc.

Actually - if you're running sweeps - conversion rates for aggressive vs. compliant landers can be small depending on how you word things. In some cases, "you could win" and "you have won" can perform similar.

When it comes to stuff like antivirus though - I find that the best approach is to just take offers that allow aggressive, and use landers that look really scary - the system-message type landers work very well.

Hang in there my friend - AM is always hard when you're just starting out!



Amy

---

Home > Mobile > Follow-along Campaigns