Home > Paid Traffic Sources > POP / PPV / Redirect

---

Some things you just don't understand until you do it. (Hugely Long Post) (5)

---

Let's start with some numbers. They're not pretty.



Total from Propellerads, spend is obviously correct. The revenue is about $15 down from missing postback data.

And here's all data from Funnelflux, the cost is not accurate on these as I didn't update the numbers yet. Still figuring out a way to update those numbers easier.





Spent 646 and got about 200 back and another 10 on Propush (started a bit late on propush). These are numbers without fees as well. Add another 3% for spend and deduct another 3-5% on revnue for when I actually get the money.

Yeah. It doesn't look good.

THE STORY
Started on the 5th of September. So about 11 days now.

I started learning the whole IM thing late 2017, before then I had absolutely 0 knowledge on marketing. I dumped my analytical mind long time ago, so it's rusty, and it's still rusty. I was focused on music.

I started by doing some random stuff here and there, and started a dropshipping store and got it to 1k usd day revenue, for a short while. Things died down soon, still running but nowhere near the revenue ever since.

So, I had some facebook ads experience before coming into this. But my 1k days were achieved by selling 225-315 usd items. CR was always low. The margin was great, tho, itt was profitable for a while.

Anyway, fast forward to now.

WHAT I USE

Tracking - Funnelflux Pro
Hosting - Amazon S3 and Filezilla Pro.
Affiliates Network - I'm on quite a few. Clickdealer, Haka, Zeydoo, Mobipium, Maxbounty.
Traffic Sources - Propellerads, Zeropark.
Automation - Python.

THE START
I started with HAKA offers first, direct-linked. I've tested a 8 offers on 9 different GEOs (most according to Top GEOs from HAKA AM). I only found one type of offer to be converting at a decent rate. Also, only found one GEO to be doing anything >20%.

(My test process is still messy, and undocumented. I did not test 8*9 combinations, but I've done enough test.)

So once I found this offer/geo combo, even tho the ROI is at -70% I let it run for a bit and see if I could collect meaningful data.

---------------------------------------------------------------
BIG MISTAKE NO 1. (WARNING: BABBLING AHEAD)

To be fair, I wouldn't have known until I did. And hopefully someone will swoop in and say it wasn't a big mistake or maybe even I could save it somehow.

Anyway, the thinking was I could try BL/WL and multiple optimisation technique to bring it to the green. NOPE.

About half of my spending has gone into this and the lesson I learnt was. Unless there's significant data showing one way or the other, no amount of optimisations will bring it to the green.

WL - So I've tried Whitelisting from early on, from a few conversions, to at the end stage I had 1x conversions from one zone.

When you focus in on zones, the ROI drops immediately and the volume just becomes so small. So unless the zone performs at significantly better CR, it's not worth it WL.

BL - I've also cut a few zones that had huge amount of traffic and no conversions or majorly negative ROI. Yeah, if the funnel ain't converting much on the other zones, it isn't going to bring it to the green just cutting the big ones.

BIG MISTAKE NO 2

Picking up my programming chop is fun. It feels like learning something new, but it's not really learning something new. So it's easy-ish and it lets me feel I've done something new.

So, what did I do in the 10 days? I wrote a few python scripts for generating stats to a local server, so I don't have to login to check stats. I have a script for stopping campaigns according to predetermined details.

I also have a script for killing bad zones according to some rules I set. Basically the optimizer, for free. (These are not very useful at the moment, not until I have a winning campaign, as mentioned above.)

As opposed to, landing pages, finding offers, testing offers, these things I don't like much ---- LAZINESS.

I'm ok with python, but I don't like HTML/Javascript (I'm not good with those, not much knowledge.). So cleaning landing pages is a *insert swear word* chore and sometimes fruitless when they becomes too complicated. But I didn't want to spend $ on these as I *should be* able to do it with some more time. I'll probably have to give in sooner or later.
----------------------------------------------------------------------------------------

I did manage to code one very simple LP and also took @twinaxe's ultimate click2sms LP for a spin.

I tested these vs direct linking. Results:
My LP CTR 75% (high because of it just a popup->redirect) LP Views CR: 0.2% Visitors CR: 0.15%
Twinaxe's CTR 3.6% LP Views->CR: 4.5% Visitors CR: 0.16%
Direct-Linked CR: 0.18%

I don't think these stats are conclusive by any means and they are not meaningful differences.

------------------------------

I've tested some more offers since then, they all flopped. Got 2 sweepstakes landing page working, have not pitched them against each other yet, though.

This is the journey so far. And brings me to my questions, would really appreciate anyone's input. I mean, if you've read this far down! Might as well right?

THE QUESTIONS
1) I had this conception that in my mind, that offers are really tied to the type of site. I.E. IF one of the zones is about watching video online and we present them to download a file, the conversion rate surely would be lower than a popunder for mediafire.com. Let's say.

If someone has categorized zone ids with type of offers and found success / failure, would you mind sharing?

Am I just thinking too much?

And 2) How much data do you need to put a zone into eternal blacklist of the ad network?

3) When an offer does -70% or -50% on no bl/wl after say 1 or 2 conversions, is there much chance of it succeeding later on?

I'm talking with techniques of optimisations. Obviously with 1-2 conversions at -70%, it might still suddenly take off on its own later on. Sampled CR is much lower than population CR here.

I have an offer, PO$2.8 I got 2 conversions from $15 spend. Kinda wide targeting on propel. Is it worth running it further?

4) Is there a systematic approach to throwing shit on the wall and see what sticks?

NEXT STEP
I'll continue throw shit on the wall and see what sticks, and hopefully design a system that I would follow. And also pitch some landing pages when I find the opportunity.

---

It´s good when you´re not afraid to spend money but you should start slower to first learn the basics.

With that pace you will quickly reach your limits and lose motivation.

When you don´t even know yet how to run campaigns properly it´s like gambling what you do - In the end the bank always wins.

Especially in the beginning there are so many things that can go wrong and most beginners make mistakes that could be avoided, that´s just normal.

A wrong checked setting here, a missed token there and your whole campaign is worth nothing.

All these mistakes can cost your money, avoid it if possible and learn how to walk before you try to run

Even when you find a good offer, would you know what to do with it?

I recommend you to read THIS THREAD and then THIS ONE

These are more than enough for now.

First learn how to do all the stuff manually, then start with automation.

How do you want to automate something properly when you don´t know how to run it properly manually yet?

Good that you don´t stop too early but in the end you can only collect meaningful data when you have a converting campaign.

Otherwise the only data you collect is that you know that your campaign is not converting but you don´t know if it´s the traffic or if it´s the offer or if it´s the landingpage.

Multiple techniques at once is never a good idea, especially not as a beginner.

It can lead to a huge mess and do more harm than good.

Don´t waste time at this stage yet, your focus should be

1.) Learn how to setup, create and run campaigns properly
2.) TEST, TEST, TEST

It doesn´t matter what you like or not.

Only thing that matters is: Do you want to succeed or not?

Yes, some things just suck and we have to eat dirt sometimes but it´s nothing that kills us.

You don´t have to become an expert in all these things but a basic understanding should be mandatory.

Also the earlier and more often you do it the easier it will become

How many offers and campaigns did you test altogether for the $646?

Forget assumptions, they are not helpful.

They can become your worst enemy.

Of course the site can be important for the choice of offers.

But here we´re talking about pop traffic.

It doesn´t really matter if you run sweeps, download offers, dating, it all can work on pops.

I never did and never will do.

There are good zones and bad zones, that´s all I need to know.

Yes

First requirement is to have converting campaigns.

Then there is not such thig as THAT amount of data.

Simple said:
- Is a zone non profitable in a converting campaign? Put it on a blacklist for that campaign
- Is a zone non profitable in all converting campaigns? Put it on a global blacklist

There is also no general rule.

For example when the targeting is tight or the geo is so small that there´s not much traffic available it will be hard to get a campaign into profit when it already starts with such low ROI.

When there is much volume available and you can cut aggressive and still have enough volume it gives you much more room for optimization and you can get lower ROI campaigns still into profit.

It also depends on the number of elements you test.

-70% with only 1 offer and 1 lander is completely different from -70% ROI with 5 offers and 4 landers.

There are many approaches that all can work.

Just read here on STM and you will see that many experienced affiliates all have their own way how to run campaigns.

Ask a question about something and from 10 different people you can get 10 different answers that all can work.

This is why it´s important (in my opinion) to learn the technical side and all things that are same for all of us ASAP.

Then just TEST TEST TEST to find what´s working best for you.

I could tell you to a T how I run campaigns but most probably you would fail when you do it exactly as I do because over the years I developed a workstyle that works best for me but could be a total fail for anyone else.

Every mind works different so just try to get the basics done and get a good understanding about the whole affiliate thing in general.

What are offers, what´s a trafficsource, what traffic can I run, what´s a tracker, how does the whole flow work and so on.

Once you get the hang of it test many offers and work your way towards your own style step by step.

Follow alongs are a great way to run campaigns on your own but at the same time you have some kind of safety net where you can ask for help or support for your very own campaigns so that you can continue with them.

And again, read the threads that I linked above.

I can imagine that it can help you

In any way, we are here for you so if there´s anything else just ask and we try to help you as good as possible.

---

Thanks very much for the long reply. I do agree with you - "walk before you run." My problem isn't technical though, if by technical you mean setting up links, postback, tracking and such. Those, I can do fine without error after 132 campaigns. And, these are easy to verify and I do double check after i've submitted.

Yes. At the end of the day, doing what it takes is what's needed. I did clean up 2 of the landers from Adplexity. What I wrote was just to highlight it's a mistake to let laziness take over.


OK. I guess it's just that simple. In the grand scheme of things, I guess the connection between site and offer carries not much weight.

Yeah. Technical doens't worry me, I have my script killing everything that is >2usd spend 0 conversions for now. It maybe too aggressive and this is what I have most question about.

If you don't mind, let me dig in a little bit about what I did with testing and rules about cutting campaigns deciding offer doesn't work.

With that one offer and one geo. These are the results.


So with this geo, I've let one offer run for very long (134 conversions), and got some data. Then I looked at the zones with >1000 view (or 500, or 100, I've tested a few). Created blacklist, whitelist. Also took some of the CTR datas (0% ctr, or 100% ctr) with high views. Created separate blacklist. Tested them all, (on different campaigns), not mixed together. Already created campaigns for single zone, the zone with most view and positive roi. ROI tanked. Volume, not enough.

Also tested 2 landers (some variations) and no landers, they didn't make a meaningful difference, in fact no lander converted the best.

One was twinaxe's download lander, original and some word changes for this particular geos. My own simple popup lander, (JS alert -> redirect). 0.165%, 0.152% and 0.179%. Respectively. These numbers makes a difference when I'm trying to squeeze that last bit of profit, but don't make or break the campaign.

I have pretty much tried everything and beat this into submission, or I think I have.

No luck there. The original campaign had the best ROI. All subsequent campaigns just took the ROI down from -40 to -65.

So, this brings me to my questions. The evidence is there, it's just not working for me. Someone else definitely has them working, but it's not me. Is there something I could have been doing? Is there some much more elaborate landing page->offer combo that has insane conversion rate?

I've spied on Adplexity and couldn't find anything. In fact, @twinaxe's download lander comes up quite frequent there.

------------------------------------------------------------------

The rest of the budgets and campaigns goes into testing another 20 offers and about 15 different geos, not all combinations of course.

They mostly tanked. <-80% ROI. I tested mostly low payout offer and stopped them at 2-3 usd spend.

Except Indonesia, where it just ate my whole campaign budget before the stats are reported in Propellerads. And one more campaign where the python bot failed to stop because of an error.

I use my python bot to do two things, report stats in html (own server) so I can see it anywhere without logging. And stop-loss on campaigns with spend and no conversion, or higher spend with way negative ROI. So I don't waste a single dollar.

But it could be too aggressive.

My question is:

1) Is there anything more I could have done to that offer/geo combo? (I do think for now, without further knowledge on more elaborate funnels/landing page, I should just go looking for another offer).

2) Is 2 dollars spend for low payout (0.35) offer too aggressive a threshold?

3) I have today tested a $2.8 payout offer, I spent $15 and got 3 conversions. Now this is the point where I'm not sure how I should proceed. If it goes down that path of the original campaign, I'd be in quite a bit of loss.

I can make that mistake one time, but I shouldn't make it twice. Should I go look for something with positive ROI at the onset?

Thanks!

---

Improving your click-through rate requires improving ad copy that lasers in on your target audience. If you’re looking for the best platform to reach your customers, then you should consider Facebook for lead gen. It’s a huge, ready-made market and we can help you

---

@mariam-gcg Please take care what you´re posting.

Such posts are not related to the topic and don´t help the threadstarter.

More of such posts and it can be considered as promo spam so please refrain from it to avoid trouble.
@noobmarketer

What´s the payout of the offer?

You lost $260 on one offer, why did you keep it running for so long?

When you mean the offer I mentioned above: You should have stopped it way earlier.

When you spend $2 without any conversions it´s fine to stop it.

Seriously, it´s hard to tell without more details.

You´re at about -50% ROI with this offer, could work or not, depends on the setup.

Not necessarily, you should look for something with potential.

Of course it´s better when you have positive ROI from the very beginning but it´s also important for example to check if you could scale the campaign when you exclude non converting placements.

When the geo or your targeting brings enough traffic it can also be worth it to start with rather low ROI because you still have enough volume when you start optimizing.

It also makes a difference if you start with negative ROI when you only have 1 offer and 1 LP in your campaign or if you have 3 offers and 3 LPs.

---

Home > Paid Traffic Sources > POP / PPV / Redirect