Home > Mobile > Follow-along Campaigns

---

Newbie Start-Up: 100% ROI without Conversions! (26)

---

Hi everyone,

My journey on STM started a few months ago, when I clearly was having enough of my Office job!
So it all started with several hours here reading posts, guides, follow alongs but honestly, connecting the dots was pretty hard. Maybe because this is not the kind of job I could succeed by investing only two or three hours a day (especially when I had already spent most of my energy in the office and other part time jobs).
So the first decision had to be made, continue working the job I was already not enjoying any more or dive in the AM ocean? Well, I left my office job and started dedicating full time on learning how things work and what one needs to do to succeed in the long run in affiliate marketing. One thing I know for sure is that when I push my self to my limits I get the best I can

During this time, I took a short break from the forum posts and spent that time reading Cashvertising and Influence: The Psychology of Persuasion (haven't finished it yet). Cmdeal's post The Best Books on Copywriting is a goldmine! By the end of the day, our mission is to Convert (sell).

So after several weeks and time spent on STM reading, playing with design tools with real stuff (helping out local AD agency converting Image compositions in responsive HTML free of charge) and figuring out how things work, I finally decided to TAKE ACTION!

I may have done some mistakes (that I will mention below) on my two first campaigns but, the cost weights on my side so "Learning ROI is 100%"
By now I should have launched more campaigns, but I think it's better to understand tracking and bidding methods on live setup to have reliable data to work on.

Verticals, Tools and Memberships
Vertical: Sweepstakes / Leadgen
Tools: Voluum, Amazon S3, Adplexity Mobile, Linode, HMA VPN, Adobe CC
Traffic sources: Go2Mobi, Avazu mDSP, Propeller Ads
Affiliate Networks: Addiliate, ClickDealer, AffiliaXe and Adsimilis

Campaign A
GEO: UK
TS: Go2Mobi
Vertical: Sweepstakes
Creatives: 8 (320x50 Banners)
Landing page: n/a
Daily Budget: 50 USD

I decided not to go with a landing page on this campaign because offer's landing page seemed quite heavy for what the user had to convert.
As suggested on several posts I used Automatic Bid type (leaving manual bidding for later tests). No filters were applied on the campaign besides the country restriction and one exchanger that did not approve several of the banners.

The results the next day were negative (as expected).
Clicks: 161
CTR%: 0.51%
CPM: 2.097 USD
Cost: 66.84 USD
Convs: 2 (shown on Network side after pausing the campaign)
Banners: The one that I cared less about design and had a better angle had more clicks - More focus on angles on next campaigns.

Question: How is it possible that for not even 24h campaign run the cost was higher than the daily budget set on the campaign? Should I check with Go2Mobi support regarding this or it has to do with the time i Activated the campaign?

Campaign B
GEO: UK
TS: Propeller Ads
Vertical: Sweepstakes
Landing pages: 2
Offer: 1
Daily Budget: 20 USD
Campaign Budget: 25 USD

Because I had an over-budget cost on my first campaign I decided to go easier this time with this one, just to be safe and have the budget under control.

Next day results
Impressions: 14,585
Clicks: n/a
CPM: 4.36 USD
Cost: 63.59 USD
Convs: 0

The clicks were not tracked properly because I accidentally pasted the wrong link on the network side (pasted the preview link of the campaign generated by Voluum after having it pasted on my browser).

Regarding the budget control, still uncertain as from Campaign A.

Also I have to revise tracking following Zeno's Comprehensive Tracking Tutorial, as the two conversion that happened on Campaign A were being reported on Campaign's B Traffic Source in Voluum even though I have the feeling that the flow was pretty simple to have such results, maybe I'll need to check with Voluum.

I'm aware that the GEO I’m targeting is not an easy one for newbies, but I wanted to spend on translations later. First goal is to create angles that I know how they sound and how they perform.
Planning to launch 2+ other campaigns before Saturday, but first I would appreciate your assistance and critics.

---

You sure lured me in with that thread title! I can already tell you'd make a fantastic marketer!

Could you please elaborate on what you mean by that?

Actually, for sweeps offers, direct-linking without landers may not be the best idea - because pre-selling is usually necessary in order to convert the visitor.

There are offers that already come with landers built-in - for example, the visitor will be asked several "survey questions" before being shown the sign-up page. If you're wanting to direct-link, I would suggest asking your AMs for offers that come with landers (aff networks call them "pre-landers" so do use this term when making your request).


I've learned to never use autobid during initial testing stage.

I find that by bidding low in the beginning, I could save SIGNIFICANTLY. If you're patient in waiting for traffic, bidding low can save you - so that you can test more stuff for the same amount of money. The downside is traffic volume will be low.

Unlike for pop traffic, where if you bid too low you wind up with all the bot traffic nobody's bidding on, on display sources I find that even at very low bids I'd still get conversions.

So - start low, and gradually increase the bid until you're getting just enough traffic to test your stuff. You could always increase it further when your camp is green, and you need more traffic volume.


Go2mobi's system operates in UTC, so if your 24h period crosses midnite, then that will reset the budget.

You can use autorules to further limit your spending. I've listed several ways to control the budget at G2M, in this post (see point 4):

http://stmforum.com/forum/showthread...l=1#post294318


Several suggestions here:

-Try to target a less-competitive geo - you've picked one of the worst ones in terms of level of competition! Try tier 3 geos in Asia, LATAM, Africa. Lower competition + cheaper traffic = improved chances for a newbie.
-2 landing pages and 1 offer aren't nearly enough. Let me dig up a thread where I proposed a test strategy for sweeps+pop - I'll add the link to this post when I find it.

UPDATE: Found it! The methodology is at the end of this post:

http://stmforum.com/forum/showthread...l=1#post293497


Hm...not sure what you mean. Has this been resolved? If not please elaborate and we'll figure this out.


Nice attempt! Keep up the good effort!



Amy

---

Hi Amy, Thank You for taking time on this!

Regarding Campaign A, yes the offer came with pre-lander and that is why I came into doubts whether to have a lander in that offer or not. Anyway I still think at least a message to reinforce the banner angle should help the questions game. The next question that bugs me regarding this is the design standardization. By this I'm afraid the design should look familiar in order not to leave the spam effect on the prospect.

So, regarding Go2Mobi I am keeping an eye on the bids and develop the best practice on results. Does the same apply to other traffic sources? The same strange thing happened even with PropellerAds.

After setting up the campaign on Voluum, Previewed the campaign on a new tab and pasted the new tab's link on TS campaign.

I'll check the referred post and apply accordingly.

EDIT: Actually I'm working on a new lander, maybe thinking for something different from the crowd, the roulette kind of lander seems too mainstream despite the fact that still converts.

---

Goodness!

Good thing you were direct-linking - otherwise, clicking through the lander would lead the visitor to the Realm of NoWhere.


Yes - every traffic source has a different way of defining their "day". Some will have a standardized timezone applied to all customer accounts. Others will allow you to specify your timezone in account settings.

Another thing to watch out for when it comes to dealing with different time zones: When comparing stats on your tracker against those on the traffic source, make sure both are set to the same timezone.

I've edited the post twice to add 2 links - one for a sweeps+pop methodology and the other for how to control spending on G2M. For your reference.



Amy

---

Solid set up! Would be really interested in seeing results!

You would want to append a slash and a number to your original click url, for each offer's outgoing link on you lander - in this format:

http://track.Voluum.com/click/1
http://track.voluum.com/click/2
http://track.voluum.com/click/3
etc.

Then, when you add the lander to Voluum, put the number of offers on your lander in the "Number of offers" field.

Then, when you create a campaign, and select that lander, you'll be presented with the same number of offers to specify.


Eager to see results!



Amy

---

Update:

Campaign C
GEO: UK
TS: Go2Mobi
Vertical: Sweepstakes
Landing pages: 3
Networks: ClickDealer + Addiliate
Offers: 2 (iPhone 7 SOI)
Daily Budget: 50 USD
Campaign Budget: 37.29 USD

Following Amy's instructions to not use Auto-bidder which in Campaign B had a CPM=$2.1, I started bidding with $0.5 that surprisingly gave traffic (taking in consideration that I was bidding in a T1 GEO).
That being said I decided to follow-up the campaign budged consumption and performance through the day and played around bidding from $0.5 to $1.0. The results are as below.

Total Spent: $37.29
Avg CPM: $0.43
Visits: 500
Clicks: 5
Conversions: 0

I wonder if traffic quality is related to the bid, in which case I believe it is.

Pending approval for Campaign D (with 3 offers), and preparing for campaign launch in LATAM.

---

Actually - you can't really draw this conclusion.

Unless the combination of the offer AND the landers AND the campaign targeting have been proven to convert well, it wouldn't be fair to blame the lack of conversions on the bid (although it's possible that it is the culprit).

It's harder to make display camps work compared to pop, because you have an additional variable to get right: The banner.

A potentially better way to approach this, would be to test sweeps camps on pop traffic first, and only scale the best offer(s) and lander(s) to display traffic, and test just banners and bids.

That way you can remove 2 variables from your testing.

It would probably also be cheaper to test offers and landers on pop traffic first, rather than using display traffic - because as you've seen, spending close to $40 on G2M only got you 500 views to your landers, and 5 views to the offer. On pop, the average price for UK mobile non-adult popup traffic on ZP is around $1.5CPM - so you would have gotten the same 500 lander views for a small fraction of the price. Granted, display traffic has higher quality than pop in general, but with the low banner and lander CTRs you're getting, I bet pop traffic will allow you to test offers+landers for cheaper.

What are the offers' payouts? I know I'm always advising to focus on CR and not CTR, but unless your offer payouts are really high, the current CTRs you're getting are not going to work. Banner CTR is around 0.6% and lander CTR averages 1%. You'd need an impossibly-high CR and payout to even break-even.

It's really up to you what you want to do next. If you do want to test on pop traffic first, be sure to use compliant landers without aggressive elements or misleading ad text - otherwise you wouldn't be able to use the same landers for display traffic later.

Let me know what you think!



Amy

---

In fact, blaming the traffic quality for sure must be one of the bigest temptations for a newbie. I hope I already patched the bug regarding this

Yes, you're right again in the above post Amy, regarding testing through pop traffic and not display when it comes to sweeps. A detail I missed when using spy tools was taking note of the traffic sources the landers were being served.

Also when it comes Banner->LP->Offer page, one of the things that should influence the CTR & CR, is the design standardization (to be the same design line from start to end), and what I'm trying to do is to keep at least the color+font set Banner->LP. Anyway, I guess I'll find this out on my way to green results

As suggested, moved the offer from Display to POP traffic through PropellerAds (this time with a better control on mudged from the first campaing). Also tested two other offers in a second campaign.

The results for a day run from each campaigns are as follows:

Campaign C1
Offers: 3
LPages: 3
Visits: 2,955
Clicks: 2 (CTR:0.07%)
Convs: 0
Spent: $14.31

Both clicks came from the offer that was not too mainstream.

Campaign D
Offers: 2
LPages: 1
Visits: 2,628
Clicks: 7 (CTR:0.27%)
Convs: 0
Spent: $13.63

Even though the data are not enough, on the coming days of test I think it would be possible to come to conclusions that the copy on campaign D was better than the one on C. Definitely the “What’s in it for me?” was missing, despite my focus to make the prospect picture the situation of winning the offer's product. That being said, WIIFM coming to the second landing page on campaign D.

Also maybe conversions and click may come with a certain delay (I remember of reading about this), due to the pop-under nature.

The temptation to approach AV Products is pretty high, but for the moment the focus is on the basics and procedures.

Thx, Lsd

---

Great to hear you've decided to use pop for initial testing!

Yes this will help, but is more a fine-tuning task. The first step would be to test completely different landers - ones that look completely different, with different themes and/or functions. You're not testing nearly enough landers. Please see a rough strategy I've outlined at the end of this post:

http://stmforum.com/forum/showthread...l=1#post293497

Also - the lander CTRs are horrible! I know it's CR that's the most important metric, not CTR. But 0.07% and 0.27% are so low that the math would NEVER work out - you'd practically need 1 in 2 visitors that land on the offer page to convert, just to break even!

I would recommend to rip a ton of landers, and try to optimize page load times - please see the "On Landing Pages" section in this thread:

http://stmforum.com/forum/showthread...-AM-FAQ-Thread!

Again, I would suggest picking a less competitive, non-English geo. For a geo like the UK, you'll need to test a lot of offers and landers, and cut a lot of placements to get in the green (unless you're really lucky and find an offer that converts very well). Although, you're willing to put in the money to find the best placements + best offer + best lander, the pay off can be big. I'll leave that to you.

Either way - looking forward to seeing more encouraging stats soon! You'll get there.



Amy

---

I spent these last two days browsing for offers and Adplexity to create a better picture of the GEOs I'm moving for the next campaigns. Definitely not satisfied with my first results (even regarding total lunched campaigns).

So as suggested by Amy in previous reply, actually I've applied for like 10 offers in LATAM GEOs (Spanish+Portuguese), pending aproval.
Also checking what's running on Arabic and Asia-Pacific (Based on KPCB report there's a huge raise of smartphone usage in this region). The process is pretty slow due to language barriers but for the basic "thanks again google translate". How to Test More Geos Faster saved me a lot of time

Now apart from testing offers in different GEOs, in terms of best practices - is it worth testing back button scripts in first campaigns?
If yes, in order to keep track of the redirected traffic I was thinking to create a dedicated campaign on Voluum where to send the redirected traffic, which then can target [SameOffer-DifferentLander] or [Different offer Same GEO]. According to Andy's post on BackButton Redirection we can keep track of this traffic too using our tracker and get info from which landing page that traffic came.

Regarding traffic sources I singed up to ZP (still pending approval); PopCash, PopAds also there's PropellerAds where I already ran some campaigns. What TS would you suggest to start with for LATAM, if any of these performs better in testing? If PopAds, is it worthy to start parallel campaigns with Blacklisted targets as suggested on Mr.Green thread, or should I consider this one for later?

Lastly, sounds/vibrations and Chrome/FireFox blacklistings. Inspired by Erik's Case Study I'm planning to add sounds to landers (those how don't have yet) and if i can technically achieve that, even vibrations. Most of the landers I'm actually using are ripped from Adplexity and altered by me.
I guess adding such features on landers will help a lot to get blacklisted pretty soon from these browsers, please correct me if I've created a wrong perception here.

I hope this post isn't too messy, but some little hints would be helpful to avoid any misunderstanding

Thx, Lsd

---

Its up to you what strategy you will use, but you can use the back button script right away. Why would you not use it if everyone else does. It would just put you way behind and give you a disadvantage (in my opinion).
You can try Andy's solution (never tested) or there is also a good back button script on Affkit.com, they have solution to track the tokens from the back pages, so your conversions will not show up as from "Unknown" zones.
Regarding the logic behind it - yes you can redirect to custom campaign and then setup rules in it. Basically you can even have 1 universal campaign and do rules by geo if you want to. Or you can resell or that traffic to services like monetizer. There are lot of options what to do with this traffic and you will have to experiment and find the one that works best for you.

You may consider Adcash, maybe Plugrush, but the ones you got should be pretty enough for start.

What do you mean by getting blacklisted? You talk about getting your domain flagged? If so, that is pretty random and you never know when it happens. For the you just need to have some spare domains and ready to switch once it happens. Nobody really knows what exactly triggers those.

---

Thanks Erik!
Yes by blacklisting i meant domain flagging. The strange thing i noticed when testing some ripped pages offline on chrome, was that they reported as harmful even though they were offline and no domain was involved in opening them. After disabling sounds (withoutother alterations) the pages looked ok for the browser.
But anyway, I've already got some domains (if needed will add others) and will start out using wildcards to ease the process.

---

Following your valuable guides, below is the setup of the 4 following campaigns.

Targeted countries: 3
CDN: Amazon CloudFront
Payout range: $0.8 - $2.4

The traffic for each campaign rotates on allocated banners and offers in order to discard the low performance ones in the next days.
Next task is to replicate these four campaigns in another traffic source, so i can split test even on traffic sources. Also, prepare campaigns for Country 4 and 5, plus pending approval offers.

Campaign E
GEO: LATAM
TS: PropellerAds
Landers: 3
Offers: 2
Daily Budget: $10

Campaign F
GEO: LATAM
TS: PropellerAds
Landers: 2
Offers: 1
Daily Budget: $10

Campaign G
GEO: AFRICA1
TS: PropellerAds
Landers: 4
Offers: 2
Daily Budget: $10

Campaign H
GEO: AFRICA2
TS: PropellerAds
Landers: 4
Offers: 1
Daily Budget: $10

It took me a little time to setup the landers and configurations due to my limited programming skills. But anyway, so far the landers in these 4 campaigns are all fitted with audio and vibration alerts. Found an article regarding IOS video autoplay feature, unfortunatelly user intervention to start playing was required when the video was hidden. If anyone has found a solution to play sound/vibration alerts for IOS devices, would be helpful

Based on statistics these three countries have different browser dominance so, for some things like sounds and vibration doors are still open I guess.

In addition, set the back button script to forward traffic to Monetizer, even though I wanted to try out Andy's method (but needed more time for this, i'll try later).

Thx, Lsd

---

It is true that by testing fewer landers and offers at a time, that it would take less budget. But the flip side is that when you don't see conversions, you're left wondering whether it's your offer or lander that sucks, or both.

This is why I suggest using 2-3 AM-recommended offers, that other affiliates have gotten conversions with, and using them to test all the landers you can find on adplexity that look different. (Note: The offers should have similar payouts. When you use the peakconversion calculator to check lander stats for statistical significance, it assumes all conversions to be of equal value. So the more different the payouts are, the less accurate the results.)

That way, you'd be using proven offers, and by including lots of landers, you can be quite certain that at least one or more will be a good converter.

It's completely up to you how you want to approach testing. The way you're doing it can potentially work out too - although it'd be a bit like stabbing in the dark. But if you set up enough tests, you could hit on a winner.

I would suggest to leave all the camps to run a bit more though - not a lot of spend yet to see whether there's promise yet.

Propeller is pretty strict about landers that are misleading. And logos are a no-no. For a list of restrictions please see:

https://support.propellerads.com/Kno...paign-rejected

Try removing all logos for a start and turn the camp back on. If they reject again, take another look to see what else may be wrong, fix it and resubmit. And of course if you can't figure out what's wrong, you could always file a support ticket - be sure to include the campaign ID in question.


Hopefully all is well now. Did you get a response?




Amy

---

Actually LATAM was not the suggested one by the AM, the Africa country offers were the good converting ones. Anyway, I have an updated list of offers ready to be tested. As mentioned in previous post, I wanted to test the same offers with different traffic sources in order to figure out which TS was converting better.

I see there are policies regarding payments and paypal accounts, where one has to have the same email address of the account with that of paypal. In terms of security for the TS it looks good, but the downside is that you'll have to loose a day or two with the support team just to sort out the connection with the accounts and verify that you're not trying to make anything illegal. As long as we're all starting out with personal free email accounts doesn't seem a good idea to keep all the accounts tied down to only one email address.

Regarding Propeller's rejection, I think i found the problem, which should have been caused by an svg file that contained copyright protected logos in it. My bad, I didn't upload the correct version.

Yes it took a day and a half, but it is solved now.

---

I've been somehow quite for a few days on my follow along but I think it was worth it and here is why.

Well excessive pressure for sure will not lead to quality work and that pressure is related with long working hours, overflowing new information and last but not least higher fear of mistaking on low budget.

Not very proud to admit the following mistakes, but Yes I'm still learning and felt something was wrong in my approach.

Mistakes.
- Tracking: Had a mess on some offers with the post-back variables where the ClickID was not correctly configured on AN side.
- Language: Testing same/alike landing pages on two different GEOs, I noticed that in Africa's campaigns landers content was English and the exit pop Spanish (this one blew away 15k visits on one campaign)
- Angles & Landers: On 5 Landers, 5 Different Angles and different content.
- Offer Type: Using same language on exit pops for either SOI Singups and PIN submits
- Overthinking offer selection on sweeps (maybe under evaluating good ones)
- Ripped Landers: Social Proof pictures, comments, etc. Like me other Affiliates are approaching these same GEOs so I guess prospects are the same and for sure have already seen those same landers read same comments and yes spotted the same face.

Revised approach.
- Tracking: Fixed mess with variables
- Language: Double checking Start-End All materials before upload and flow before sending any traffic.
- Angles/Landers: For each campaign at least 2/3 Angles x 3 Landers (design) = 6-9 landers and able to analyze and optimize.
- Ripped Landers: Rebuilding the amateur photo library.
- Campaign Run: Budget (AvgPayout x NoLP x NoOffers) x 2 TrafficSources.

That being said, decided to test 3 of the offers from which I hit my first conversion and see results.

GEO: UK
Offers: 3
Avg Payout: $1.17
TS: ZeroPark
Daily Budget: $16*
Visits: 14,934
Clicks: 45
CV: 3
Cost: $16.06
Rev: $2.85
*Being aware that the campaign is on a T1 country, and 1st campaign on ZP decided to play carefully

Budget was burned within 2-3 hours. No visible budget distribution option on campaign creation/management. Also another problem I noticed is that, unlike other traffic sources, I couldn't set the campaign to pause and manually start it after approval.

With a list of offers ready to run other campaigns, among which I also have Carrier Specific offers, I was thinking to simplify the targeting using rules on Voluum rather than founding specific traffic sources for inventory from those carriers. Please advise if my below approach is correct.

Campaign link -> Wifi/3G -> If Not Carrier A or Carrier B -> LP X -> Non CarrierSpecific Offer
..............................................If Carrier A -> Carrier A -> LP X -> Carrier A Offer
..............................................If Carrier B-> Carrier B -> LP X -> Carrier B Offer

Ps. I hope, it is not becoming a boring follow along, but besides keeping track and analyzing my progress, I hope this will overcome same mistakes to other newbie members

Thx, Lsd

---

Thanks for sharing all your "mistakes"! They're not boring at all - like you said they'll probably help other new affiliates to avoid some of them. So thank you!

These should work as intended. Actually, instead of creating the "If Not Carrier A or Carrier B" rule, you could have just put that as the default path. Any traffic that is NOT carrier A or B would end up at the default path.


Several points:

-Sorry if I've asked this before, but how many landers have you tested / are you testing? If you're only running one lander right now, it may be difficult to increase ROI drastically enough to hit green, given you're at -70-80+% ROI right now.

-Those stats you've shown - are they for your best offer or all offers total? If the latter, could you please show stats for each offer?

-Have you also drilled down to mobile carriers (not just ISPs)?

-Basically I'm trying to see how promising your camp is and whether or not you'll be able to get to profits quickly enough for you to justify running it. UK is a large and expensive geo. If you want to get profitable, be prepared to spend quite a bit of money in finding a really good offer, and then spending money to cut out all the worst placements. As a long-term strategy though, after spending the initial money to cut out most of the bad placements, you'll be left with good traffic you can use to test/run future offers on this geo and traffic source, which will give you a big edge. But if you're just starting out, you probably won't want to spend this kind of money. Of course it's ultimately up to you.

-For offers that come with their own pre-landers, I 100% agree that they should be tested using a separate set of landers. Landers will need to be very simple if the pre-landers are already complicated. For example if the pre-lander is already asking 4 survey questions, then your lander can't be doing the same or the visitor would bail. Probably something real simple like the spinning wheel, or even just a page where the visitor is asked to pick a favorite color for the iphone 7 (for example; you may not be promoting iphone 7 offers), or even just a js popup with simple "congratulations" text, will be worth testing. And don't forget to test direct-linking! For some of the offers that come with their own pre-landers, direct-linking alone can result in profits.


I would really like to see more-detailed drill-downs of your data, just to see if the camp is worth running further. Let's start with these if you wouldn't mind:

Offer > Lander > OS

Offer > Lander > Mobile Carrier


We shall see what happens!



Amy

---

Thanks for taking the trouble to post all those screenshots!

However...I was hoping to see multi-level drill-downs, so that I could see which offer+lander+OS combinations are in profit or close to being in profit, for example.

Also - what do you mean when you say "most excluded websites"?

Copying entire campaigns from Adplexity can be a good way to start if you need to save on budget. Of course not every one will work, because there are things you can't see (such as payout, or the actual landers used if they're cloaking), but if you copy enough of them some will.

Please do show me screenshots of that UK camp where you drill down to progressive levels - I'm seeing data that MAY turn out to be promising if you run more traffic to the camp. But we shall see.



Amy

---

Thanks so much for posting the new screenshots! Very clear!

OK - I see 6 or 7 combos that are either in profit or are 1 conversion away from being in profit. And given wifi is converting so well and that wifi has a ton of traffic volume, if this was my camp I would definitely run it some more to see what would happen.

You could definitely do this! In fact, you can use JUST TalkTalk alone to test offers and landers for cheap. Once you find a good offer and lander, you could then open up your targeting and just cut anything that isn't profitable then.

Sounds like a plan?



Amy

(And the low CTRs for UK always make me cringe. Once you have a winning lander, it would be worth it to optimize it for loading speed if you haven't done so already. May bring that CTR up a bit.)

---

In fact - Challenged by CTR and CR yesterday ended up with 7 new landers! Spent quite a lot of research time on discount/offers forums and sites where only women have the patience to explore...

I'll give a shot with Talk Talk and see what happens case this has made me work hard, has to pay back definitely

Regarding speeds I'm quite sure some improvement is needed mostly on the code side, as for images I've remade everything from scratch vectorial/non-vectorial and passed through Kraken to make sure there is no delay from images.

Thanks!

---

On my last post above, the overall spent for this campaign on ZeroPark was around $43 with only 6 random conversions (ROI -85.23% ).

Since I still have an open deal with UK - and as suggested also by Amy, I thought it was the case ti give it another shot!

On 14 Dec I gave this same campaign another $19 spent resulting in 18,945 Visits / 91 clicks / 3 conversions and $2.4 Revenue.

Satisfied? Not because:
1. Traffic Source was consuming offer's budget within 2-3 hours. Maybe cause of the bots or maybe country specific traffic flow.
2. The offer was converting but not for my traffic setup
3. Could not setup the working combination due to limited knowledge and/or tools capability to filter the desired traffic.

Based on tracking records I had a few combinations of ISP/Carrier traffic, device types and OSs for which the filtering on ZeroPark was not enough to channel the traffic (or most probablly I have to get more familiar with ZP interface) ...and this is where I stuck for a moment!

Maybe I pushed it too far with this and perhaps there is a shorter way to deal with these stuff but it seemed a better aproach to me.
With $62 worth of tracking data I was not ready to give up and here where SQL comes into play.

I'm not a wizard in SQL but at least I find it better than excel to return the desired set of data based on certain conditions.

Imported CSV Reports from Voluum.
- All campaign traffic from day 1 till 14 Dec grouped by Targets, ISP/Carrier, Mobile Carrier
- All campaign traffic from day 1 till 14 Dec grouped by Sources, ISP/Carrier, Mobile Carrier
- All campaign traffic from day 1 till 14 Dec grouped by Sources, Targets, ISP/Carrier

By knowing the performing ISP/Carriers that were performing on this campaign, I could generate the following reports.
1. List of all targets and sources that were sending traffic to the performing ISP/Carriers
2. List of all targets and sources that were not sending enough traffic.
3. Available traffic ( number of visits visits) each target and source had for the selected carriers
4. List of target/sources to temporary blacklist.

This way I could pause at once all target/sources not working for my campaign as well as the major non-converting consumers. Below are today's results:

Visits: 15,534
Clicks: 63
Conversions: 4
Spent: 15.96
Revenue: 4.80
ROI: -69.92%

It is not a huge progress, but knowing that I don't have a killer landing page and also lots of non-converting targets to be excluded is ok.

Thx Lsd

---

I'm quite confused - weren't you going to just target TalkTalk to test landers and offers to find a best offer+lander combo first?

I don't understand what you meant when you said:

Why did you need to program a MYSQL script? Are you using Voluum? I'm afraid I'm completely lost. Could you please elaborate what it is you're trying to do? I know you've decided to ditch the camp for now, but perhaps we can learn from stats you've already spent money to collect.

Regarding the AV camp - some thoughts:

-For future reference: I wouldn't spend money on translations before I see that a camp has promise. My thinking is that if an offer doesn't convert on English landers, then it probably won't convert on Malay landers either.

-With AV on pop, you have to run very aggressive using scare tactics that look very realistic in order to get offers to make offers convert. "Your machine may get damaged by viruses" is not going to do the job. On the other hand though I see you're using the really popular lander with the Google logo. This one WILL do the job, but some traffic sources will ban the camp or even your account if they catch you using it, and most advertisers / offer owners will either stop you from promoting the offer and even refuse to pay for leads if they catch you. So you'll need to balance effectiveness with risk.

What most people will do is either 1)run aggressive against the rules and cloak, or 2)just run offers that ALLOW aggressive promotion. I cannot encourage you to cloak, but you could definitely ask your AM for offers that allow running aggressive, and choose traffic sources that won't ban you for running landers that have google logos.

(A minor point: There's spelling error on your landers - "damaget" should be "damaged".)

-I would definitely suggest to test more landers that look completely different. Right now you're just testing 3 different landers (not counting the language variations). The CTR is extremely low - I know I always say not to focus on CTR but with such low numbers the math would never work out to be profitable.

So to summarize, if you want to pursue AV, you'll need to rip a lot more landers. And if you decide not to cloak, at least ask your AM for offers that do allow aggressive promotion - or you won't stand a chance.




Amy

---

The reason I decided to stop with UK campaign right now was because I've made quite a lot of mistakes there.

Yes, but when I went to change the campaign setup in ZP only for TalkTalk (since it was converting on WiFi) I considered the "Carrier" selection limited to only Mobile Carriers. For this reason I went with SQL, in order to filter all targets/sources which had reported to have most of the TalkTalk traffic WiFi traffic I was looking for and test the offer again with only those targets - basically I tried to filter the traffic behind the traffic source in order to get 2/3 target lists based on ISP/Carriers.

Regarding the AV camp when asked AMs, they said they were doing good but at the same time I couln't run aggressive, which is why I went with "May get damaged".
Taking in consideration of the fact that English is widely spoken in MY, I decided to test with translated content with the intention to target those users that convert better on local language.

About landers, on Adplexity I noticed that the selected themes were like the most clicked for MY and as it was listed too, I was courious to test the Google themed lander which seems to be quite present. And you're right, it didn't take too long before noticing the fact that that lander could get me into trouble.

Regarding cloaking honestly I'm not even thinking about at the moment.

I'm aware of the procceses separately but when it comes to combine them I fail and right now I have to re-organize everything and decide wheter to continue with sweeps or focus on another vertical but not where cloaking or alike techniques are required to make it work (it's too early for that)

---

Wow!

First of all I missed the part about the camp being run on ZP - didn't occur to me that you couldn't target by ISP there - apologies for this oversight on my part!

Usually people wouldn't bother to target indirectly via a method like yours. But you could be onto something here - this method could be your key to targeting by parameters that the traffic source doesn't allow advertisers to target directly. Useful or not, that was very creative! Were you doing that with stats from a tracker you're hosted yourself?

A more straight-forward way would be to just launch the same campaign on a traffic source that allows targeting by ISP.

But I completely understand and respect your decision to leave UK alone for now. It isn't the best geo to target for a newbie, and will take money to find a really good offer and then more money to optimize a camp properly. You can take another shot at this geo again once you have more experienced and have more confidence that your investment can be recuperated.


I understand. But I can tell you with almost 100% certainty that the other affiliates running the same offer are running aggressive and cloaking. Unless you're promoting big brand name antivirus products that people have heard good things about, running aggressive is the only way to go as far as pop goes. And if you're not wanting to cloak, then asking to run only offers that allow aggressive is the only option I can think of.

And don't start thinking that offers that are strict and don't allow aggressive, will convert better than those that do allow aggressive. This is just not the case. You can make money just by running offers that allow aggressive.


Sweeps is still hot, and is a vertical where you can still make a lot of money without cloaking.

You just have to be willing to test more offers and landers than the next person, and be able to test and optimize efficiently (i.e. for as cheap as possible).

Just as for AV offers, some sweeps offers will allow aggressive promotion as well. You can prepare 2 sets of landers, one for offers that allow aggressive and one for those that don't.

If you're not wanting to run sweeps, go do some exploring on adplexity to see what other verticals are doing volume on pop. Basically anything with broad appeal (i.e. stuff that appeals to the general audience) can be promoted on pop.



Amy

---

No need to apologize at all - Thank you for taking time on my follow along and helping me out!

Yes, normally even I wouldn't do that if I'm not that disctracted lately, ending up even with terrible gramatical mistakes. As for the stats, I only used data from drill-down reports on Voluum and proccessed it with MS SQL Express locally on my PC.
I don't know how effective that way of cutting targets down would be cause I couldn't test it enough, plus other variables like bidding position, time and most importantly how the traffic distribution itself works must be considered as well. For the moment I'll let it sit while I focus more on getting things work.

BTW... There's one thing that I'm having a hard time to deal with - Over-judging the offers!
By that I mean, when browsing for offers I get very sceptical wether to test X offer or not beign affraid it'll end up beign just a spam or deceptive one. I've realised lately that this sh*t is pulling me back on aproaching new offers.

I know I'll beat this one but just in case anyone has any proven solution to this it'll help a lot!

---

Is your concern:

#1)Wasting money on offers that won't convert? Or

#2)Promoting offers that will not provide value to people?

If it's #2 - we may need to pick a more whitehat vertical or even traffic type.

If it's #1 - Testing landers first to find a decent one will save you money when you get around to mass-testing offers. You can either do mass-testing and let the data tell you which offers are good, or you can comb through intel tools like adplexity to find offers that are receiving healthy traffic volumes and preferably exhibiting an uptrend, and just test those. To me though, if you've done all that work upfront on ripping and cleaning up and testing landers to find the best, you may as well get more value out of all that work by testing more offers using that lander. But again, if you choose to rip entire campaigns from adplexity, then you can choose to skip lander-testing altogether, and just run the same landers as in the adplexity camps.



Amy

---

Home > Mobile > Follow-along Campaigns