Home > Push Traffic >

---

A noob follow along (14)

---

EDIT:
oops. Tagged twinaxe, I thought it was like a notification to read my post seeing what you would do >.<.

So my background is in IT / Software Dev. Just had a thought that I should learn how to sell, even though I am an extreme introvert.
What better way than to learn through affiliate marketing on the internet!

Anyway this is like my second month into affiliate marketing. My first month was browsing through a lot of misleading guru information out there, before I found some gems (like STMForums!).

I had 0 idea about affiliate marketing until I went through the 40-day guide by @vortex and then the push guide by @twinaxe

Got pretty comfortable with the mechanics of Pop traffic, even though I've never had a profitable campaign (only did like 30 pop offers).

I've ran a number of sweepstake campaigns on Pop and Push in various Geos like Thailand, Malaysia, Brazil, Indonesia.

Felt like moving on to Push, and after reading a number of follow-along threads I thought I'd put another one up .
So here goes the follow along.

GEO: Sweden
Offer: Mobile Dating offer recommended from AdMobo Network.
- SOI email $3.5 payout

Tracker: FunnelFlux Prod
Traffic source: PropellerAds Push (Web + In-page)
Using CPC for "High" Quality traffic only.
My bidding approach is to start on the low end, then increase the bid over time.
- Just so I know what zones I can get at the "low end" before just blindly overpaying for everything.



Creatives: I've got 2 different angles
1) Explicitly showing an ad that it is a dating site
2) Mimic-ing a user that messaged the person.

Landers: 3 different angles.
1) Quick survey
2) Something like this

3) Interactive chat (as if they're talking to the person)

Angles and Landers were from spying on Adplexity Push and Anstrex (Trying out 2 ).
and also from Adplexity resource bundle when you purchase a subscription.



Background Note:
I've run this offer in Norway last week, with no success.

Also tried running with RichPush traffic.
RichPush left a bad taste in my mouth.
- They had an birthday offer crediting $200 on a $200 deposit, with 5 spots available to claim the coupon on their landing page.
- By the time I signed up and asked for the credit, I was told it was no longer available (this was still during the promotion)
- My AM was nice to give me some placements though, cant blame him.


Here are the results
Creatives

Classic Push


In-Page Push


Landers


End background note


Soooo I want to run this offer again to see if either the GEO I tried last time was not suitable for Norway, or if the offer is just not working for me.


Some questions to begin that I'm really need some advice on. Maybe it's obvious from another persons point of view

1) Is there enough data to determine the best lander/creative set (based on the screenshots provided)?
- I pretty much began with nothing, except some landers and creatives I found from spying. So I am trying to gain insights from my own data

2) Should I even bother trying this offer again for Sweden, even though I did not have a profitable campaign in Norway?
In my head, I'm trying to eliminate a few things.
- PropellerAds not being suitable for dating for EU geo's. (This is why I'm doing the second GEO with same offer) . I should try a different traffic source after this.
- The offer sucks. I should move on.
- My Landers/Creatives angles may not be good enough.
3) I'm at the stage where I'm confused on what to do.
- On one hand I know I should stick to one geo, one vertical and master one traffic network.
- On the other hand I want to explore and try different things.

---

I forgot to mention the structure of my campaigns.

I have Ad creatives that flow in two different directions.
Each direction has 2 campaigns (Web + In-page push)
After the first camp in Norway, I decided to keep the ad creatives consistent with the angle of the Landing page.

So for example:
Traffic (Web/In page) goes to ad creatives in the campaign under the "user mimic" angle.
That then flows into the landing pages that are consistent with this angle.

Likewise for the dating angle.
I have creatives that make the user expect to go to a "dating" style landing page.


Here's the "user mimic" angle flow (samples)
(I dont know what to call it. Just emulating a person messaging someone?)




and that goes to a landing page like this..



Here's the dating angle (samples)
(The ad creative is done, such that the person expects to see a Dating landing page)



and that goes to something like the second landing page in the previous post.


Day 1 results.






Landing page results

Looks like LP2 and LP4 are doing well.
These LP's are flowing from the "Dating" angle Ad creatives. (The other two, LP1 and LP3, flow from the "User" angle ads.








Dating angle
In page zone ids

Can't do much for this one at this stage.

In page creatives

Looks like 10402531 is doing well







Web zone IDs


For zones taking up the majority of the traffic, I've decided to "pause" them to allow even traffic distribution from other zones.
So for now: 2609 and 1845 impressions in the Dating Zone ID results. Will un-pause later. Maybe.


Web creatives
Looks like 10402659 is doing well.





User angle
In page



Zone ID



Creatives


Web

Zone ID


Creatives



Results looking pretty bad, based on overall stats.

Question for you guys.
1. Would you guys still continue this campaign or let it run some more time??

Considering a similar failure from Norway GEO. I'm thinking of just killing the camps and moving on soon.

---

How could you get comfortable with a traffic type when you didn't master it yet?

Pops are great for beginners since they are very simple and rather cheap. Also they help a lot to get an understanding about statistics.

Push brings more complexity into the game since you have to deal with creatives and its daily dynamics (when you don't have big numbers).

I would start with way lower payout offers and focus only on finding an offer that converts like crazy.

In the beginning all you need is an excellent converting offer that runs at about -50 to -70% initial loss. Such an ROI does not give you panic for low payout offers.

Only with conversions you can learn something. Without conversions you are just a boat in a big fog having no idea which direction to go.

If you have no idea what offer to run read some of twinaxe's recommendations. These things still work and are excellent for training and even more.

Last but not least: Forget profits for a while. Just concentrate on offers that bring you quick and plenty conversions. Test 100 offers a week and you will find something.

---

I will drop a few thoughts here, since you are running dating and that's my playground.

1. SE and NO... two very competitive, though great GEOs for dating. These GEOs are known for very high user value and the offer payouts usually reflect that. Not the offers you chose though, $3.5 in SE or NO is way too low. So unless these have some super simple flow, it wont work. I've had offers in the "Nordics" that paid $10+ just to give you an idea.

2. PUSH traffic isn't a good match for dating, I've lost most of the offers that I tried to promote on PUSH traffic... the quality of these leads tends to be too low. It was very easy to convert though, so judging by the numbers you shared, there is a problem somewhere. Hard to say what the problem was, I see you have tried several ads, but just 2 LPs and 1 offer. The problem can be any of these. Based on my experience, when the results are really poor, it's usually a problem with the offer.

So my advice, for what it's worth, if you want to give adult dating a proper try:

- skip push traffic. If you want to keep it simple, try POPs... those have a bit better quality when it comes to the leads generated. The best would be to focus on adult sources and go with banner traffic.
- SE, NO etc... are not the best GEOs for starters, look at the newer EU countries instead. Poland, Czech Rep, Hungary... these can work, competition is lower, payouts are lower too so it's better for testing.
- OFFER IS THE KING in dating, singup for more networks, you need to have access to many offers. Focus on finding some promising ones first, the finetune the rest of the funnel. A great offer will work "decently" even with a weak funnel, but even the strongest funnel will fail, if the offers isnt good.

Get ready to test... a LOT

All the best.

---

No problem, here I am

I agree with what @larsometer says, sounds a bit strange that way

Sweden is no easy geo.

It´s very competitive, bids are rather high and volume low and with $3.50 payout it probably won´t be a joyride.

No, to determine what´s best you first need more conversions, otherwise you can´t know what didn´t perform.

But keep in mind that the offer is the most important element that has the biggest impact on your campaigns performance so when a campaign just isn´t converting also your creatives and LPs are at least "ok" and you run at competitive bids then you can mostly be sure it´s the offer that isn´t converting good enough.

A good offer will also convert somehow on mediocre traffic/LPs/creatives but a bad offer won´t convert on best traffic/LP/creative.

I probably wouldn´t do it.

Often that´s the case.

It doesn´t even mean that an offers really sucks, sometimes it just isn´t converting for us whereas it´s doing good for others.

Test more offers, period

No need to peg to something that isn´t working.

Yes, it´s good to stick with one traffictype but I would test different offers and verticals and maybe few trafficsources to see what´s working best for me.

Once you found your pocket you can and should stick with it to improve your skills.

Seriously, in dating it´s all about the images.

There are not many verticals where the visuals are that important as in dating.

I personally also wouldn´t run the chat lander, just 2 or 3 variations of the "3 rules" LP and that´s all.

When I run dating offers (and that doesn´t happen that much) I use exactly 1 multilanguage LP.

3 questions, 1 CTA, that´s all and it works pretty good.

How do you come to the conclusion that LP2 and LP4 are doing well.

You have only 3 conversions in the campaign, LP2 and LP4 have 1 conversion each but LP3 has also 1 conversion so what make you thinking that LP2 and LP4 are doing well?

Even when LP1 has 0 conversions, that´s only 1 conversion less than the LPs that are doing well.

With so few conversions every new conversion can completely turn the tables, it´s too early to make assumptions.

Why do you think they are doing well?

The screenshot shows 0 conversions but this is only the stats from the trafficsource so probably you see more in your trackers stats

On the other hand there are conversions tracked in Propeller Ads in the following screenshots.

What´s total spend (numbers from Propeller) and what´s total revenue?

Apart from that, you already tested the offer in NO and it didn´t convert so don´t expect wonders when you run the same offer somewhere else

Seriously, better run lower payouts and less competitive geos first so that you can test many many many more offers because this is most important - testing offers.

That way you will probably find a profitable campaign faster and along the way you can learn much more when you run more.

Yup, probably not what you (or anyone else) want to hear but before you focus only on profits it´s important to learn the whole progress of running campaigns first.

When you test enough and stick to some basic rules that sooner or later a winning campaign is basically inevitable.

---

Hey all.
Thanks for the feedback!. Really appreciate the advice.

Thanks @matuloo for the notes! Will keep it in mind.


The offer was a SOI - email sign up.
I looked at the offer and now that you mentioned the part about "Simple flow" I have some doubts on the simplicity of sign up now.
The actual promoted offer URL is: https://snap-date.com/








@larsometer and @twinaxe mention I should look into lower payout offers and less competitive geos.
At the back of my mind, I fully am aware I need to pay for data. It's all about the data.


On the other hand, I have 2 questions which I have doubts about:


1) As far as bidding strategies go: I recall that I should bid slightly above the optimal price from the either the newbie or 40 days pop guide
- I found that PropellerAds just recommends the MAX bid price for "optimal" bid?
- I also found the "Optimal" bid in the traffic chart, is different when I'm making a campaign.
Which to choose?


2) If the traffic price for Pop is already high, then do low payout offers really matter?
- I must also consider the price of traffic of the geo in question (This is also why I moved to Push traffic, see my Indonesia Pop vs Push comparison below)?


Traffic chart.



Here are some sample screenshots for Indonesia, I took from PropellerAds traffic graph while making a campaign.


Push - $0.009 (CPC)



Pop - $2.728 CPM (SmartCPM) - Clearly much higher than the Traffic chart "Optimal CPM Rate" above



I ran a campaign for the same offer for both Pop and Push in Indonesia.
You can see I can get much more impressions with Push per $, compared with Pop.
Plus these are engaged clicks, compared to the intrusive pop-ads.





I've also run sweeps in these GEOs as well. And the CPM cost were pretty high.
Push costs relatively low compared to Pop like Indonesia, when considering the impressions I get per $ spent. Since I mainly pay for Clicks.


$2.7 CPM for Malaysia

$8 CPM for Thailand





So my thinking was if I can get more impressions with Push traffic, why not just go Push?


Right now Sweden CPM is $8. Push ~ $0.09, so similar numbers (I can get more impressions in Push vs Pop, just like the Indonesia comparison).


Now while writing this, I realised I have chosen Tier 2 geos.
But does the same logic apply for tier 3 geos? ; Why not choose push if its "cheaper"?


One strong reason why I wanted to try out Push traffic was the cost of traffic in Pop relative to Push, as per screenshots above.


Am I having the wrong thinking here?







Hey @larsometer. Thanks for the feedback, much appreciated.


I probably badly worded. I meant to say I was comfortable with the process of doing Pop offers. I definetly have not mastered it thats for sure...
I focused on Sweep offers, with some mobile app installs (VPN, Anti-virus) offers as well.
So here's my workflow:
- Finding some offer
- Ripping/modifying landing pages
- Creating campaign
- Run for some time (leave for a day, or check when I can)
- Check the stats
- Look for placements to pause, if initial ROI looks good. Otherwise move on to next offer.
The workflow above is what I meant by "comfortable".


I get the point now. Need to test 100's of offers that convert. Thank you!







Yeah I badly worded that >.<
Definetly did not master Pop. I just meant the process of running Pop campaigns.
And then looking at pricing stats of Push Impressions (with CPC pricing) vs Pop Impressions (SmartCPM) I thought it was worth a shot. (see the reply for @larsometer)





Ok, glad to hear your opinion on this. I was starting to think the same to myself now. It was not a joyride!




Yes sir.












Yeah. If I focus on conversions then I cannot make any conclusions, since the sample size is too small.
Understand now its all about conversions.


I was too focused on CTRs, and each "stage" of the funnel seperately.


My thinking was if Ad creatives got clicks.
It means the ad was good enough to grab eye-balls and click. High Ad CTR = Ad must be good.


Then if Landers were getting CTR's then my lander would be convincing enough for people to go through to the offer. High lander CTR = Lander must be good.


Then from there, its up to the landing page of the offer itself.







Across 4 campaigns (I put $15 limit on each) - $60 total.
More or less 4x the payout in each.


I've found many people have different ways of doing the "initial" testing.
- Some do 2x. Some do 5x. Some do 10x of the payout for their campaign budget.





The one with 2 conversions for in page started off with -54% ROI.





Yes sir. Fully aware that its not about the profits at this stage. Am fully focused on learning and experimenting at this stage.

---

Ok, we need to set some things straight here.

1. Comparing PUSH Impressions to POP impressions is wrong thinking. It's something completely different.

In case of PUSH, it means a user has just seen your AD, what matter in this case is how many actually clicked it.
In case of POPS, every impression is actually an impression on your LP, since that's what opens in the POP.

Let's take a look at your screen now:

You have received 264805 PUSH impressions, but only 1320 people actually clicked the AD. With the POP campaign, you received 6421 impressions but all of those actually landed on your LP. So with the POP campaign you had almost 5 times as much eyeballs on your LP with 1/4 of the spend. Now, the actual clicks are way more valuable than a simple POP impression of course so we cannot compare it as simple as looking at these numbers, because these are two different traffic types with a different form of delivery.

A better comparison would be to look at the price of an LP click (so how many actually clicked through your LP), but the ultimate value to look at is the final ROI and in your case, these are quite similar.

The takeaway, ignore the price per impression when comparing push and pop traffic.

Keep in mind these are two different types. The initial cost per impressions is irrelevant. Push also produces lower quality leads, at least in case of dating, so that's another important thing to keep in mind. And don't forget about the final conversion rate, more expensive traffic often converts better than cheap.

2. Now the bidding.

Ignore the automatically recommended optimal rate, because it looks like the system basically sets it close to the max bid.

What I'm doing is looking at the graph and setting the bid so it sits at the end of the spike, just before the graph starts to flatten out. Let me use one of your screens for illustration.



I would set it where I made the black line, so around $1.4-$1.5. The graph clearly shows that any further increase in the bid wouldn't bring in such a dramatic increase in traffic. Know what I mean?

3. The offer: https://snap-date.com/

Yup, this looks like a full form submit offer, not a simple email SOI. That might be the reasons it doesn't convert that well.

4. let me quote

Wrong thinking again.

I can make a misleading banner with clickbait style angle/copy and misleading elements and it will have a great CTR. But it won't convert for shit. Same for the LP, it can have a high CTR but won't be a good match with the offer lander and again, it won't convert good.

CTRs are a great indicator but the actual CVR and final ROI are what matters. Quite often, ADs with low CTRs convert way better than ads with high ones. And the same goes for the LPs.

Just to give you an example :

Let's say I have two ADs, one mentions the price, the other one says it's free. Obviously, the "free" one will have higher CTRs, but the one where the price was mentioned will convert much better, because people already expect that they will be charged.

Hopefully it's starting to make more sense now

---

Thanks @matuloo
Yes, a lot of things crystal clear now. Thank you

---

Also about the bids:

On the image that matuloo showed we can see very good where to set the bid, even when the recommended bid is higher.

The thing is that you often have the situation where the high bids come from already very good optimized WL campaigns where other affiliates can afford to bid very high on specific placements.

When you then just bid that high for your test campaign you ca easily overbid.

A good way to find a good bid is to check for the point where the volume doesn´t increase anymore when you bid higher.

That´s the point where the line in matuloos image is set.

Before the line you can see that the increases the higher you bid, after the line it´s only the bid that increases without much additional volume.

About creative CTR:

The trick here is to find a good combination of high CTR and high CR.

When a creative is very specific it mostly has a lower CTR but higher CR, when a creative is very broad it can get a much higher CTR but also much lower CR because it´s not that related to the offer anymore.

"Win an iPhone 13 when you fill your name and email in a form" vs "You have a new message. Read now"

Of course your creative should be related to the offer so that you achieve a healthy CR but when your CTR is too low your traffic will dry out pretty fast.

On CPC the creatives with highest CTR will receive most traffic because the creatives with highest CTR make most monoey for the trafficsource.

Try to find a good balance, a high enough CTR to keep traffic flowing but also a high enough CR to run your campaign profitable.

---

Just want to give an update - to let you know I have not disappeared.

I've been going back to Pops and mass testing offers from Traffic Company. I guess I lack experience and so I'm going to just run a ton of offers and see what sticks. Even though I'm "comfortable" with the process of Pop. Still gotta get experience points to level up!.

So I like how Traffic Company just lets you pick an offer and run it STRAIGHT AWAY (most networks need approval before running)

Previously I would just browse through an offer on the CPA network, or ask the AM to find me some offers. Wait for them to approve whenever I request. This sometimes ends up in a day or 2 of turn-around. I just want to test NOW.

I recalled reading @twinaxe 's programming thread.
https://stmforum.com/forum/showthrea...on-with-Python

Why did I bring up a programming thread?

Through this thread I was learning the "manual" process of what was being automated, so thank you for making that follow-along @twinaxe

So the main take-away from reading that thread was:
Getting IVR offers (and click2sms) for particular geos and carriers , and choosing what offers to run based on stats like eCPM. Throw them in a campaign and let it run!

Combined with reading this thread, I got quite excited:
https://stmforum.com/forum/showthrea...ing-to-profits

Random question:
- In TrafficCompany's eCPM stats, does this mean this is what they pay affiliates (like us) promoting the offers?. In other words, how much profit we can make on average?

So anyway, back to running my Pop campaigns.
PopAds was my go-to traffic source in the beginning while reading the newbie guides, but the turn-around time to approve my campaigns takes too long (sometimes I have to wait the next day, due to timezones I guess).
PopCash is much better in this regards, so this is why I ended up using them.
Both get me conversions, and they both have low volume compared to PropellerAds which makes it less attractive for some affiliates; so perfect for a testing playground!.

When I create campaigns on PopCash, I found it a bit annoying looking up the traffic price across different geos when I'm about to run a camp for a geo.
So I decided to read some API docs for PopCash.

I wanted a way to quickly look up CPM costs for a particular Geo on Android Mobile devices, across carriers. They only provide average CPM cost, but they used to provide top CPM bid. Not sure why they took that away.

Anyway I ended up quickly hacking something together that looks like this:


Yes. The UI is not that great, Function over Aesthetics .

So now my workflow looks like this:
- Look up some offers on TrafficCompany to run.
- Look up the the traffic source CPM stats for the geo
- Set up the offer in tracker & setup the campaign url
- Run it on PopCash traffic source.


Anyway. The point of this post was to announce I've just got my head down, mass testing offers; while doing the day-job.
I'll be honest, this is extremely fun. Even though I've yet to see profits, and down in the red.
The anticipation of discovery and learning is fun!

This follow-along thread might become a hybrid of Pop and Push traffic.

---

So I've been testing a bunch of offers.
Some Haka, Some Zeydoo survey offers. All direct linking BTW.


I also recently switched to using Binom Tracker, honestly I loved the user interface through the demo they provided.
I just could not get "in the zone" using FunnelFlux UI to make campaigns.

Anyway I have a massive concern that I want to verify if I screwed up with my setup of Binom.

I was looking at my tracker stats and could not help but be surprised at the lack of clicks.
This is me checking 8 hours after the campaign finished (due to no budget). So I do not think it is a delay in reporting between PopAds and Binom stats?.

This is a sample report of creating a campaign URL in my tracker and just direct linking to a Zeydoo offer.

You can see PopAds is showing 17393 impressions vs the ~3000 clicks in my binom report.
Thats 14000 clicks lost!..WTF?




This is the GTmetrix..honestly not sure if this tells me much. My networking knowledge is basic



Other info:
I'm using PorkBun for my domain. It's also where I set the A name and CNAME records.
DigitalOcean server hosting Binom is in Bangalore (seemed like it was closest to the tier-3 geo's I'd be testing out)
(No landing pages are used, pure direct linking from tracker -> Offer)
I was targetting a handful of geos on PopAds.

Anyone able to chime in and provide input on this click loss?

EDIT:
I've asked Binom support, but it wont be another 2 hours till I get a response.

I did another test with PopCash traffic on the same offer.
PopCash impressions are pretty consistent with the number of clicks I'm seeing on the tracker, while looking at the "live" reports

I'm starting to suspect its bot traffic. I guess they "insta-close" the Pop-up before it closes, but PopAds counts it as an impression.

Best to verify with support....Anyway back to the grind

---

Let me qoute this:

The ECPC is calculated based on the click on the call button and the ECPM is where the CTR on their own pre-landers is included in the calculation.

Yes, the go-to sources should have fast approval times, especially when you run time sensitive stuff like IVR.

But be aware that Popcash only has fixed CPM and no Smart CPM bids.

Apart from that Popcash has some pretty good traffic.

What about this?



This is definitely a huge loss, make sure that you LP is working properly.

We can see that traffic arrives at the tracker so the redirect seems to be working fine so I would rather look at the LP first.

About your setup, do you use a CDN and DNS?

Please check also THIS THREAD.

---

Support got back to me and it turned out the click loss problem is because the server I used was too weak (1 CPU 2G Ram)
I just went with the recommended ~200k clicks per day, when I first signed up.

Turns out the recommendation was based on average traffic distributed throughout the day per minute. Buying pop traffic results in spikes of traffic... I didn't think that far when setting up the server. And also I didn't read beyond the graph, about server description. I'm a simple guy
I will just upgrade the server. Problem solved.


I use cloudflare CDN and DNS for my S3 buckets hosting my landing pages, back when I used FunnelFlux.
For Binom server, It's just a single droplet. DNS is same as where I got the domain name from (Porkbun.com). No CDN for tracker.
- Do I need to setup the CDN for tracker server, if I already use Amazon for my Landing pages?

BTW The results I showed were from direct linking from Traffic source straight to the offer ( I didnt use my LP at all )

These are my thoughts before reading your thread..I will read your thread now.. thanks!

---

Important lesson: Don´t be cheap with your infrastructure or it can easily become your bottleneck.

In Germany we have a saying: "When you try to save money at the wrong end you have to pay twice"

For your tracking domain you don´t need a CDN but use DNS for it, can help for faster redirects.

---

Home > Push Traffic >