Home > Push Traffic >

---

The Road to Learning Push Traffic (42)

---

Hi, I intend to update this follow along everyday as I run my first Push campaign--

Short Intro:

I'm Najee. I started out learning entry level direct response copywriting, and then joined STM before committing to Facebook CPA in the middle of 2020 using Vortex and StickUpKid's FB tutorials.

Big mistake, lol. FB is definitely not beginner friendly at this point.

My ad accounts got nuked over the next few months. It did NOT help that I was running sweeps of all things.

So I decided to learn a different traffic source, which ended up being Push.

Learning Push concepts was SIGNIFICANTLY more complex than FB to me, but Vortex's 40 Day Tutorial + TwinAxe's Push guide helped my brother and I get through.

Funnel Setup:

1. Traffic Source: RichPush
2. Bids: $0.03 ($0.0312 Recommended)
3. Sub Freshness: Subscribed within the past 7 days

Creatives Examples / LP's



Right now I'm following TwinAxe's creative testing methodology:

I'm testing (5) creatives with no big image. They are all using the same icon and emoji.

They all have different Headlines and Messages for the first phase of split testing. They are all along the line of * Win a Free iPhone 12? / Click here to win an iPhone 12 *

We are running 3 landers: Spinwheel, Giftbox and Questionnaire. The 3rd offer actually comes with it's own questionnaire lander.

Progress So Far:

Here is my two-day progress so far via FunnelFlux (a bit of test traffic + running low budget once overnight):



Offers / Geos:

I'm just runnin SOI sweeps offers in Tier 2-3 countries right now. I'm running in Taiwan now, but I've also ran Thailand sweeps on FB.

Affiliate Networks:

Right now I have access to a lot of South East Asia offers via WeWe and BigBangAds.

I have also been accepted to multiple networks including Jumbleberry, CPA Affiliate Network, MaxBounty, AdToGame.

Next Steps:

• Re-read the Vortex / TwinAxe modules on cutting bad placements
• Take action + make notes in Campaign journal
• Commit to a testing budget (I just pulled the trigger on running traffic to step out of the fear of spending money)

Closing Thoughts:

I need to figure out why I have offer views but no offer clicks...

Is there any more information that I should be providing as I progress through the follow along?

Thanks in advance!

---

Nice start!

Basically, the more info you provide, the more accurate help and advice we can provide. But there's no need to report literally everything

We need to know about your setup for sure. What traffic network you are using, what do you bids look like, the freshness level of push subs you are targeting.
Then about the funnel, you might want to share some AD samples... LPs, are you using any at all?

Then, once you want some advice on optimization steps, we would need to see your traffic performance, possibly broken down by ad or LP at least.

The first thing you need to do is find some offers that do convert more or less steadily. Without conversions, there is no optimization possible. And to make any solid decisions, you need several conversions.

So I'd focus on this part, you need to test multiple offers and find some that can convert.

---

Thanks Matuloo! I will update those things now.

UPDATE 2/12/2021

- Problem: Realized that a full 1/3 of my traffic was eaten by a broken lander
* Solution: Fixed lander and ran $5 of traffic to test (didn't mean to spend that much)

- Problem: Conversions and bids were not being tracked in FunnelFlux
* Solution: Fixed the tracker postback and learned that bid postback tracking is important

Next Step: run traffic over night with a low budget to get some stats in. This is truly the start of my campaign as the previous day's stats were broken by a faulty lander.

UPDATE 2/13/2021

- Problem: misjudged traffic speed and spent $17.34 while I was sleeping.
* Solution: learned to toggle the traffic speed from ASAP --> Even

Next Step:

* I am buying up to $7 more traffic at $0.03 CPC. Per the 40-Day Guide, I'll measure the conversions after x10 ad spend per offer ($24 total in this case).

* Awaiting RichPush campaign approval

---

Always test your campaigns before putting them live. Click through the funnel to make sure it works the way it's supposed. Abd yes, tracking is of utter importance, so pay close attention to this as well.

When I was starting with paid traffic, I didn't run campaigns at night at all. Try to run during hours where you are able to actually check the stats often, it's a good idea to follow this rule when starting with AM.

---

Update 2/15 --

So, Taiwan is 14 hours ahead of me, which makes my daytime hours the middle of the night for them. Would the conversion potential likely remain the same?

All three offers have converted at least once per 10x offer payout in traffic cost. One has converted twice.

I'm going to see if either of the other two convert within the next 10x offer payout test.

If not, I'll cut the underperforms and move on to the landers cutting phase.

Also, I've been holding my effort back due to fear of losing money. I know good and well that fear isn't going to help, and I just wanted to let it be known that I'm tired of my crap

I need to be intentional about getting my 6+ hours of sleep a night. I've been slacking and working feverishly late at night. No more of that.

Updated Original Post:

Funnel Setup:

1. Traffic Source: RichPush
2. Bids: $0.03 ($0.0312 Recommended)
3. Sub Freshness: Subscribed within the past 7 days

Creatives Examples / LP's



Right now I'm following TwinAxe's creative testing methodology:

I'm testing (5) creatives with no big image. They are all using the same icon and emoji.

They all have different Headlines and Messages for the first phase of split testing. They are all along the line of * Win a Free iPhone 12? / Click here to win an iPhone 12 *

We are running 3 landers: Spinwheel, Giftbox and Questionnaire. The 3rd offer actually comes with it's own questionnaire lander.

---

Well yes, this is a problem. Most offers don't work well during the night, I mean not THAT good. Usually, the activity cools down after midnight. It is possible to stay profitable 24/7 for sure, but the night time hours are way lower in volume.

Not sure what your sleeping habits are, but if you stay up late, you could run campaigns until you go to bed, which would mean you could catch some of their daytime hours. And then relaunch early in the morning to catch the end of their day.

BTW: running campaigns during the day is perfectly fine we are all doing this, especially when targeting GEOs in different time zones. I'm only bringing this up to save you some frustration... once you gain some experience, you can start running non stop for sure.

We've all been through this, but it's something you need to get used to. Do not look at this as "losing money"... it's an investment, you're paying for data and education. There is absolutely no way around this.

The best result are usually achieved with the most fresh users, try to experiment with this setting too.

---

Thanks for this tip and the others!

Update 2/16 --

I've gathered enough data to select a winner from my set of offers. I'm cutting the worst performer, and pausing the 2nd place for the second round of funnel testing.

I'm starting the lander testing phase now with 3 landers. We'll see what happen.

---

Update 2/17 to 2/19 --

1)

Finally was able to work out my tracker issue for good.

About half of my incoming trafficwas getting sent to the wrong IP address. Turns out I somehow set up a duplicate A-Record in Cloudflare that pointed to the wrong IP address. So essentially half of my incoming traffic was eaten.

I ran a test with 11 impressions to make sure the fix was complete.

I probably shouldn't have ran traffic before getting that resolved. I don't regret it, though, because in the process I put "progress over perfection" into practice. Plus, that's one less mistake I'll make in the future.


2)

RichPush campaign just got approved. I'm retesting the landing pages again at 10x payout in ad spend max. I won't really count the first test run since half of my traffic didn't even see the landing pages.

3)

I applied to about 20 more affiliate networks to expand my selection of APAC and South American SOI Lead Gen offers.

I've learned that 80% of my success is going to come from launching and testing, so I need to make sure I expand my offer access within the next week.

I plan on getting both a grocery and fast food sweeps campaign running by March 10th.

---

February 20 Update --

1) My giftbox lander ended up beating the other two landers.

2) I'm moving onto the second phase of offers using the GiftBox lander. I'm reactivating the other two offers I cut earlier. I need access to new offers, so I'm going to wait until I get accepted to one of the other 20 Aff. Networks I applied to.

3) Whitelisted 5 placements that had at least once conversion since the the stats were marred by my A-Name DNS error. Two are in the green overall.

4) While Im wainting to get access to more offers for this GEO, I'm going to go ahead and get a second sweeps campaign up for testing

---

Don´t spread yourself too thin with too many networks.

Test few and when you see success on a network stick with it for some time.

When you send your traffic to too many different networks you risk to don´t make enough revenue per network to hit the payment threshold.

In worst case you then have all your money spread across different networks and can´t use it because you first need to hot the threashold or have to wait for payout.

---

I hear you, loud and clear.

February 23rd Update:

I wanted to get a ton of applications in just to increase my chance of ultimately getting access to one or two more plentiful offer sources. I only really had one.

After having taken yesterday to actually talk to AMs and study my target geos' pattern of popular offers, I now have 2-3 solid sources to select from for my two geos.

I'll want two more sweeps campaigns launched by February 7th.

I have launched the second round of testing 5 offers, after having found that my giftbox lander performed the best.

I've set this budget to ASAP spend, so I'll be watching the traffic spend closely.

Awaiting results.

---

In first test stage it´s often better to spend the budget slower throuigh the whole day to get a better impression of a little bit longer timeframe.

When you set it to spend the budget ASAP you can blow through the whole budget in 10 minutes and only have stats for these 10 minutes.

In worst case these 10 minutes are very bad time for the offer and you would never know the true potential of the campaign because you don´t test the other hours.

---

I definitely needed to hear this. I'm very impatient.

February 23rd Update #2:

I'm in $12.84 adspend and 6 conversion, and I haven't quite reached 10x payout in adspend for any offers yet. Though I am close for my lowest paying offer.

I'm looking at cutting placements and I see one primary candidate. This guy is eating up the lionshare of the traffic with trash CTR.

Hopefully blacklisting this source will improve my bids.

---

There are always some placements/zones that send a lot of non-converting traffic. Usually they also have poor CTR performance, just as in your case. Cut these as they are just eating up your budget.

---

February 24th Update --

I cut the bad placement that was sending the lion's share of my traffic. I highlighted one more placement I'm cutting.

I'm going watch the overall performance of the others and give them until being in loss by $0.90 (3x my lowest payout) before considering cutting more.

I'm prepping two more campaigns while data comes in here.

---

February 25th Update --

Traffic is still coming in slowly, and I got tw0o conversions.

Increasing my bid from $0.0325 to $0.04 didn't help traffic speed or win rate at all, so I'm dropping the bid back to my original $0.03.

I have one source that's a little suspect. It's already losing more than my $0.90 threshold just slightly, but it does have one conversion.

Even one more conversion in a reasonable amount of time with bring the ROI up significantly, so I'm going to keep it open.

Also I'm not sure if cutting it would impact traffic too much. Most of my sources have only one visitor and I'm not really sure why that is. Maybe they're the more premium sources?

---

Yup, this is the problem when doing optimizations based on a single conversion, because as you said all it takes is one more to change the ROI dramatically. In cases like this, I keep the placements running unless they are losing heavily. In your case, it's coming close to 3x spend though, so if you don't get another conversion quickly, it's time to cut it.

---

February 26th Update --

Winrate has been hovering for almost three days around 27 - 32% since I dropped my bid to 0.04, 0.325 and then 0.3. Interestingly, all of there bids yielded essentially the same winrates.

Recommended bid is hovering daily between $0.047 - $0.052.

Most of my publisher sources are currently getting almost no traffic. Some (including the one I'm likely going to cut) haven't gotten traffic in 2 days.

Therefore, I'm raising my bid up to 0.045 to see if I can finally get some volume from other publisher sources.

---

Feb 27th Update --

Traffic finally started coming in at a decent pace with the slightly raised bid. Got five conversions overnight. I just need a little more traffic to start cutting offers.

However, I have cut the following placements (in loss by $0.90+). Yellow = freshly cut. Red = previously cut.

Most of the publisher sources have sent negligable traffic, so I am interested to see what will happen with the balance.

Turning traffic back on and awaiting more stats while I build a 2nd campaign.

Also, I unintentionally put the red / yellow dots in different places below

---

February 28th Update --

Finally got enough data in for all five offers to start cutting. Three are getting cut, and now the final two are going head-to-head.

I nearly cut these two sources before realizing that the revenue is coming in as Euros, instead of US Dollars. FunnelFlux support gave me some instructions on managing sooo.... I'll do that right now before I mess up.

I'm going to get that fixed, but these sources are CLOSE to being cut regardless.

---

March 1st Update #1 --

So the bid winrate fell from about 48% - 39% last night. Traffic came in at a snail's pace.

Now that I've cut quite a few bad sources, I'm comfortable turning RichPush's ASAP budget on, which just gives me access to more sources when I win a bid.

I'll be watching the spend closely.

---

Hey Najee.



Would be great when you could include the table headers in the image so that it´s easier to see what these columns are

Apart from this I just wait for some more stats that we then can check together

---

Ah, I'm not sure why I kept leaving out the headers

March 2nd Update:

Alright, so after reading Day 30-32 of the 40 day guide like 3 more times today, I've concluded my iPhone offer split testing.

Below were the two finalists, and the WeWe one currently has a 59% success probability over the Big Bang one. I'm not going to spend anymore money on them because I'm losing too much.

The only hope left for this campaign MIGHT be optimization.



iPhone 12 sweeps have been running by other affiliates for a while, so I'm not terribly surprised that I'm so unprofitable, especially as a beginner.

Below are the sources that weren't at 100% loss. Only three publishers were profitable, and a few at least are at least workable, I think.



While looking through all the stat breakdowns, I found that the lion's share of traffic and conversions came from Android OS phones... Took me by surprise because Taiwan's smart phone market share (through official channels) is supposed to be dominated by Apple.

However, Statista did say that number was EXCLUDING online phone sales.

Not sure what my stats means, but I know I'll just repurpose this iPhone campaign into a Samsung campaign next.

Regarding Ads and Landers:

Three out of five of my ads clearly outperformed the bottom two in conversions:



Lander CTR probably needs to be higher:



Conclusion:

My initial thought is that this campaign doesn't have a chance at breaking even, at least not with so many of my sources blacklisted.

I'm going to take a couple of days to read and apply the Optimization lessons in the 40 Day Guide.

My instinct is telling me I probably need to do a bit more research into the competition's landers on Adplexity, and potentially test more landers.

I probably need to replace those two ineferior ads, or just cut them. Maybe some combination of optimization will justify a second split test between my two offer finalists.

We will see after I do some reading!

---

Yes, just because an offer is better than another one it still doesn´t mean that the offer is a winner

How do you want to optimize?

You can only do real optimization when you have a converting funnel.

It´s pretty hard to saturate such prizes

You´re running push and classic push is not available for iOS.

Don´t just look at the number of conversions, also check how much clicks the creatives received.

In this case it´s ok but always check how different metrics are connected.

Just out of curiosity but what LP style do you run and what CTR do you expect?

Don´t worry too much about landers.

Fow sweeps there are spinwheel, giftbox and survey, mostly that´s it.

When you have these three landers focus on testing offers.

---

Thank you for all the advice!

I am actually not completely sure how I'd approach optimization yet. I am not done reading the optimization modules from the 40 day guide.

However, it sounds like I need to maybe test my landers' functionality since I have no other offers to test. I justgot an email from CloudFlare saying that I encountered a 35% origin web server errors in the last 30 days.

I don't know what that means, I need to fix that before I touch anything else in this funnel.

Once I fix this, I think I need to clone this campaign and retest, since my CTR has been signficantly impacted by those errors.

I am running one of each of the Questionnaire, Spinner and Giftbox style landers.

I didn't have a specific lander CTR target. My plan was to figure that out as a went based on my results.

---

Update March 7th

So I've learned that the "35% origin web server errors in the last 30 days" Cloudflare email I recieved was either:

1. A phising attempt (literally no analytics indicate such a high number of errors outside of this email)
2. An extremely blackhat Cloudflare marketing campaign (there were no links except to the CloudFlare load management upgrade

The only Cloudflare community post I could find regarding this exact email has input from various users who've concluded the same thing.

Their emails (from within the last 7 months) all start 54% errors, so I am no longer concerned about this thing. Maybe I'll get an answer to my post inquiring about it, but this server error thing is no longer on my radar.

I reviewed my campaign notes from Day 1:

I did a few things terribly wrong due to being impatient:

1. Continued to run traffic despite knowing my tracking was only working halfway (I had aduplicate a-record)
2. Didn't restart my campaign to get unmarred stats and performance
3. Rushed lander testing because of weak results and to save money [as a result of 1+2]
4. Rushed offer testing because of weak results and to save money [as a result of 1+2]

I'm just going to restart the campaign and do everything right from the beginning.

Now that I think of it, I never even got data to properly test zones because half of the initial traffic got lost.

I won't make that mistake again!

---

Come on, you have a text about the error and you already said what you did wrong...

The error message says "35% origin web server errors in the last 30 days" and you say yourself "I had aduplicate a-record".

With an A-record you point the domain to the server where the files are hosted = origin server.

When you have a duplicate A-record where one record points to a wrong server then such an error can happen.

Depending on how often files were tried to get called from the origin servers it´s normal that it´s not a clear 50% split.

No phishing, no extremely blackhat Cloudflare marketing campaign, just a techinical error because of a wrong setting you made there.

There is a saying:

"When you try to save at the wrong end you often have to pay twice"

---

Yesterday I didn't even connect those two instances... Maybe if I didn't rush through the AWS + CloudFlare setup module, I would have known what the error notification meant *grimace*


Understood. Fail fast in the beginning, as they say in the business world.

---

March 8th Update [Beginning Attempt Number 2]:

Restarting offer + lander testing since I rushed through the testing the first go around.

Running one of each standard sweeps lander: gift box, questionnaire, and spinner.

I'm running 5 offers with no lander.

I'll experiment with one offer that has it's own lander.

I cleared my blacklist to properly test all my sources.

---

March 9th Update:

So much less stress running traffic at a steady speed, rather than trying to keep up with ASAP speed traffic from a timezone 14 hours ahead of me

I guess some lessons I'm too stubborn to just listen to. Well my wallet and small budget don't appreciate that.

Trying to brute force my way through bad stats with lots of placement cutting SMH.

Back to work.

The offer that came with a lander didn't convert within 10x payout adspend. Cutting it. Five offers still running.

---

March 10th Update:

Finished first round of offer testing. This Big Bang Ads offer got to two conversions first within a rate of 10x payout adspend.

Pausing all other offers to test landers.

---

Good to see you´re making progress

---

Thank you.

March 12th Update:

I was actually about to type that I needed to test more landers due to only (1) conversion the past couple of days.

I somehow missed the fact that this offer converted twice within slightly below acceptable adspend.




Both conversions came from the same lander. I'm going to see if the offer converts once more before hitting $21.90 in spend. Just interested to see if the gift box will convert.

The questionnaire is bombing in CTR, so I'll just turn it off. I don't have a target CTR just yet, as I'm just focused on one step at a time.

But I do know 7.69% is far lower than the other two CTRs lol.



Also dropping bid to $0.03. I'd been thinking about it as the recommended was hovering at just over $0.06 for a few days.

However, the recommended bid is now just under $0.04. So I'm curious to see how $0.03 will perform when the recommended bid jumps back to $0.06+.

---

It´s pretty normal that questionnaires have lower CTR than spinwheels for example.

But also often the questionnaires have a higher CVR that can compensate the lower CTR

---

Good to know! I'll let it run after all ��

---

March 12th Update #2:

So, no landers converted with another 10x payout in adspend. So I'm running with the spinwheel to test my 5 offers.



One thing I actually forgot to do was cut placements during the initial testing phases....... So I could have spent far less on testing.

Lesson learned.

Anyway, I'm cutting the following placements since $0.60 in loss puts them 2x my lowest payout. Seeing the one on top hurts... I'm also cutting the two at the bottom since they have 0% CTR and are close to $0.60 anyway.

---

Cut the top placement because it spends much more than the others.

If you want to also the second placement because for the moment it shows much worse CTR than the others.

The remaining placements I would keep running because there are no more big spenders.

Also test the blocked placements again when you have a good funnel.

---

March 15th Update:

Not the point I wanted to update at, but my RichPush account rejected all of my cards two days ago right when I ran out of traffic spend. The issue is being troubleshot for now.

One offer has converted twice within about 10x payout adspend.

I paused it to conserve traffic spend, and to compare the other offers at their respective 10x payout adspend.

I'm just waiting to be able to spend more money so that the other offers can actually get to 10x.

---

Update March 17th:

So my RichPush account issue is halfway fixed. Had to process my transaction through the rep's company payment portal.

I have been running traffic for about 12 hours.

Traffic spend was unusually slow today. Hopefully I'll have everything ready tomorrow to set up a second campaign while this traffic is coming slowly.

---

Good to hear that the issue is more or less solved, keep us updated

---

How much did you spend total on these offers and campaigns?

---

Well, they didn´t really won.

They were better than the Wewe offers but they are definitely no winners.

Honestly man, move on and test something else.

You also should keep in your mind that a successful campaign is not counted from the day it runs on profit.
It´s counted from the very beginning so even when you would get one of the offers from -60% ROI to +20% ROI it would mean that you still need to compensate about $100 losses.

With a $0.30 payout offer at 20% ROI it would mean that you make $0.06 profit per conversion.

You would need more than 1.6k conversions just to break even for the whole campaign and only then you would run on profit.

As we can see from the stats you get only few conversions per day per offer so at this speed you would need a year or so to break even.

So again, better move on, test more and then focus on something that shows more potential.

---

Home > Push Traffic >