Pushing My First Affiliate Offer..

[GriD]

Never go with the rec bid, at the end of the day they're a business and they want to get the most money out of you.

Thanks for the insights, yes I understand they make a living of a % of my spend and I am fine with that... I needed somewhere to start, short term thinking of them... Because if they eat up my testing budget and I don't see enough conversions (and I don't even mean break-even) then I will test other traffic sources... If they want us to stay they should help us instead or shaking our pockets... I'm lowering my bid on Propeller, let's see how that goes...

Most traffic is bad with push and pops because its spammy and is full of bots. If you're paying for lots of clicks on a website with no conversions blacklist immediately.

No specific one is showing up but I'm keeping an eye on that... EZmob rep also made me aware of this!
But still what is lots of clicks//no conversions... Hard to tell without experience But I don't mind the budget invested in learning... If I don't do it, I will never get anywhere! As you say it's my money... If I run out of it before 'graduation' then it's off to free traffic sources for me as I am out any income so this budget is not replenishable

a. Your push image and text is an oversell when compared to your landing page and the offer itself.
b. The unengaged were click bots you filtered at your landing page (most likely).
c. Some combination of a & b (possible).​

a. Not really overdoing it on my creatives as I don't want to be too aggressive, sending clicks to a sweepstake so never saying they have won anything... One of the creatives is a bit less clear as it reads "Did you win...?" - So whoever clicks might expect a 'yes' or 'no' answer instead of my LP stating they can win if they enter followed by the offer's 4 page questionnaire... Updating that one - removing the question...
b. Yup think so too... Propeller counts anybody staying on the landing page for more than 2 seconds as an engagement... I will need to learn how to read my server stats
c. That is obviously the right answer...

Thanks for your input Guys ! Highly appreciated !!!

 

[GriD]

So what's happening right now...
I wonder if I should post the data, I guess not as there is not much I can't write down in a few words...

- EZmob Almost 1400 total clicks sent to my LP but still only one conversion... Lowering the bid for tonight and will chat with EZmob tomorrow to see what they suggest...

- Propeller - I adjusted the bid as suggested (Thanks again @Aiden L) but I only got 40 more clicks and since then flow has seamed to drop dead... So raising it by one point and wait to see if the traffic will start flowing again... If not I also will have to chat with someone @ propeller... No additional conversion there either...

- Landing pages - My LP has an overall CTR of just over 10% which feels low Is it wrong for the audience or do I get bad traffic or is a fishy offer? Hope time will tell... In the mean time I worked on an additional LP... Had to fight a bad internet connection today... I need a reliable backup... I just added an additional LP, spreading traffic over the two LP's using BeMob - easy to do - happy with that feature...

- Next I guess would be either another LP or a complete other offer... Seeing I'm not getting much conversions here I'm tempted to throw in another offer Staying with a sweepstake or I will completely loose my marks

But first thing tomorrow is calling my internet provider (and lose an hour or two on the phone before reaching anybody knowledgeable after the usual 'did you reboot your computer?' and 'please reset your modem' questions)

 

[Graybeard]

144 clicks sent to the LP, only 67 identified as engaged clicks...

a. Your push image and text is an oversell when compared to your landing page and the offer itself.
b. The unengaged were click bots you filtered at your landing page (most likely).
c. Some combination of a & b (possible).​

 

[azgold]

But first thing tomorrow is calling my internet provider (and lose an hour or two on the phone before reaching anybody knowledgeable after the usual 'did you reboot your computer?' and 'please reset your modem' questions)

I absolutely dread those kinds of calls. It feels like such a waste of time sitting on hold for an hour then having to restart everything while on the phone with them, even though I've already done that. Hopefully, you get through it all fairly quickly.

Re your landers - existing and one yet to launch - are they different in appearance? I'm just curious. Initially, it can be the landing page, traffic or offer. I was thinking that if the landers are different designs and the new one does well, that will give you a big clue. The fact that people drop off at the landing page suggests that may be the problem. Not necessarily the design, image, button placement or colour - could be your text, maybe? That's why testing pages are so important.

Also, are you blocking/blacklisting some of the sites that are eating your budget without any action on your current landing page?

And did you do a bit of poking around to see if the offer seems reasonably good?

Have you considered trying either the same offer from a different network, or the same type of offer from the one you're using now? You can try them simultaneously to test.

Don't want to overwhelm you, just throwing out things that come to mind that others have done.

 

[GriD]

@azgold you're not overwhelming me...
I manage to do that by myself... Intentionally... It forces me to search a path...

Re your landers - existing and one yet to launch - are they different in appearance? I'm just curious. Initially, it can be the landing page, traffic or offer. I was thinking that if the landers are different designs and the new one does well, that will give you a big clue. The fact that people drop off at the landing page suggests that may be the problem. Not necessarily the design, image, button placement or colour - could be your text, maybe? That's why testing pages are so important.

LP1 CTR is still around 10%... I have no clue if that is good or bad as it depends on so many factors! The creative, the offer, the traffic, the page look and feel, the content,....
LP2 looks very different and with 210 visits it only got two clicks... Looks like this will go down again soon !

Also, are you blocking/blacklisting some of the sites that are eating your budget without any action on your current landing page?

Did not at the start, this is why this thread is perfect for me... Make mistakes and learn as I go... I can only encourage others to do the same, no matter how much you read, you will eventually forget it... Do it and see what works for you (or does not work, felling like i'm good at that right now )
So up to now only blocked out the ones that exceeded the offer payout...
I feel like I could do even better here as on BeMob I see the CTR per publisher feed.
It sounds like a good idea to cut down after a certain number of visits but from what number of visits do I evaluate? I guess 25 visits should give me one click at the bare minimum seeing that LP1 has a 10%CTR...
Or would I evaluate at 50 visits? This is a numbers game and I have zero reference... Building myself some basic reference here as we go (this will all be funny to read a year from now)

And did you do a bit of poking around to see if the offer seems reasonably good?

How the hell do I do that?
The offer seems interesting to me and the people around me, although we don't live in the UK we would certainly be interested if a similar offer came to life here in Belgium...
The AM told me their offer conversion rate (at that point in time)... It sounded not too bad (but what do I know)...
I am nowhere right now, just feels like I'm sending garbage traffic to that offer

Have you considered trying either the same offer from a different network, or the same type of offer from the one you're using now? You can try them simultaneously to test

Yes I am running this same offer on Propeller ads as well and it did convert a little better (but far from great) until I decided to play with the bid... Since then traffic has stopped, I have reset it to just over the original bid but will wait bit now, I suppose it takes some time to recover and start flowing again... Again part of learning...

As for the same type of offer, working on it right now, I built the same type of lander as LP2... Looks like it's not a good idea to lunch on that...

 

[Aiden L]

CTR of just over 10%

Thats quite normal i get 10-20%.

Also can i ask how much you have spent in total?

offer conversion rate

This is sort of inaccurate, people could be using fb or bing ads to the offer which have much better targeting options. The conversion rate could be geeting carried by one guy doing extremely well with a premium higher quality traffic source. Like azgold says make sure to test the offer on other networks, they may be shaving you.

 

[GriD]

@Graybeard, all push... So as well push on EZmob as Propeller...

Thats quite normal i get 10-20%.

Thanks @Aiden L, that gives me an indication that LP1 is good enough for now...

Also can i ask how much you have spent in total?

Easy...
EZmob Direct linking test 2500 clicks at $0.02 : $50
EZmob LP test 1,715 clicks at $0.028 : $48.07
Propeller LP test 185 clicks at $0.09 : $16.60 (they stopped sending traffic as soon as I adjusted the bid, I brought the bid back to where it was, if it doesn't restart I'll buzz them tomorrow)
And a whopping total of 7 conversions bringing me $11.20 in return
Total spend about $115 - Total invested about $100

Like azgold says make sure to test the offer on other networks, they may be shaving you.

I am totally aware of this @Aiden L... But then I also know @SuryaJ from EZmob is aware of this thread.
I mentioned this before, shaving is a short term vision, it will bring the one that does it a little money right now but that is pocket change compared to what they might see in the future... I am in this for the long run and will obviously not return anywhere I feel I have been wrongly treated...

 

[GriD]

LP2 looks very different and with 210 visits it only got two clicks... Looks like this will go down again soon !

How interesting... Remember: Don't haste into decisions...

Today LP2 has 330 visits, the additional 100 visits made a big impact... CTR has raised from 1 to 4% and it got one conversion... (One more for Propeller...)
And looking at the stats over the past two days the CTR is even better as the LP1 CTR... (Yes I am sending more traffic to LP2 so I can evaluate it faster, LP1 already has a total of 1500 visits)

It looks both LP's need to run a little longer...

And if all works out, later today I am going to throw in another offer/Geo offer (still sweep/tier1) on both traffic sources... (if not that will be tomorrow )
All enough to keep my brains crunching...

 

[GriD]

Status today...
EZmob campaign hit the initial campaign budget... I upped the campaign budget so it can run a few more days...
One more conversion... Again from Propeller...

So for the time this campaign has been running I got 11500+ visits from EZmob traffic, 112 clicked on the LP and only one converted? The LP's are supposed to filter out bad traffic So I'm not sure what is happening...
Especially since 300 visits from Propeller generated 41 unique clicks resulting in 6 conversions...
This truly is a numbers game...
Off to reading stats again and filter out more costly non converting sources...
I added a second sweepstake offer to push traffic on both these traffic sources... Will report on those as well as soon as I start seeing visits...

 

[Graybeard]

What sort of ads?
Propeller Interstitial Ads 13/Nov/2020

root@server:/home/user/domain/logs/tests# grep -c 'pointer' tos_bot_hosts_2021-03-30
787
root@server:/home/user/domain/logs/tests# grep -cv 'pointer' tos_bot_hosts_2021-03-30
107
root@server:/home/user/domain/logs/tests# wc -l tos_bot_hosts_2021-03-30
894 tos_bot_hosts_2021-03-30
root@server:/home/user/domain/logs/tests# echo "107/894"|bc -l
.11968680089485458612
root@server:/home/user/domain/logs/tests# echo "787/894"|bc -l
.88031319910514541387
root@server:/home/user/domain/logs/tests# zcat ../access.log-2020-11-30.zip| grep "nwk=prop" |cut -d'-' -f1|wc -l
1520 clicks
12% bots
88% user IPs with hostnames (maybe real)

I say "maybe real" as it's not that difficult to run a clickbot on cable or fiber
pointer is a hostname that is identified and could send email

100.19.13.43,43.13.19.100.in-addr.arpa domain name pointer pool-100-19-13-43.phlapa.fios.verizon.net.
100.2.240.22,22.240.2.100.in-addr.arpa domain name pointer pool-100-2-240-22.nycmny.fios.verizon.net.
100.35.112.178,178.112.35.100.in-addr.arpa domain name pointer pool-100-35-112-178.nwrknj.fios.verizon.net.
100.36.54.2,2.54.36.100.in-addr.arpa domain name pointer pool-100-36-54-2.washdc.fios.verizon.net.
100.37.248.73,73.248.37.100.in-addr.arpa domain name pointer pool-100-37-248-73.nycmny.fios.verizon.net.
100.43.29.135,Host 135.29.43.100.in-addr.arpa. not found: 3(NXDOMAIN)
104.12.141.31,31.141.12.104.in-addr.arpa domain name pointer 104-12-141-31.lightspeed.irvnca.sbcglobal.net.
104.1.25.205,205.25.1.104.in-addr.arpa domain name pointer 104-1-25-205.lightspeed.cicril.sbcglobal.net.
104.137.120.28,28.120.137.104.in-addr.arpa domain name pointer 104-137-120-028.res.spectrum.com.
104.143.52.23,23.52.143.104.in-addr.arpa domain name pointer 23.host.columbus-ks.com.
104.159.177.196,196.177.159.104.in-addr.arpa domain name pointer 104-159-177-196.res.spectrum.com.
104.162.18.193,193.18.162.104.in-addr.arpa domain name pointer cpe-104-162-18-193.nyc.res.rr.com.
104.172.117.23,23.117.172.104.in-addr.arpa domain name pointer cpe-104-172-117-23.socal.res.rr.com.
104.172.236.94,94.236.172.104.in-addr.arpa domain name pointer cpe-104-172-236-94.socal.res.rr.com.
104.173.43.137,137.43.173.104.in-addr.arpa domain name pointer cpe-104-173-43-137.socal.res.rr.com.
104.176.23.244,244.23.176.104.in-addr.arpa domain name pointer 104-176-23-244.lightspeed.bcvloh.sbcglobal.net.
104.222.18.161,Host 161.18.222.104.in-addr.arpa. not found: 3(NXDOMAIN)

 

[GriD]

Thanks @Graybeard can't say enough how much I appreciate your help here!!!

These comments are a big wakeup call !!! (and knew I had to look into this, just had not given myself the time to do so)
It still is a numbers game but it looks like there is a lot of cheating going one...

A few moments ago... And one more conversion from propeller traffic...

Point 1 - Yes both EZmob and Propeller are running to the exact same targeting - Geo / devices and OS ...
Point 2 - from the IP's that clicked on my landing pages...
* EZmob 46 resolve into a hostname... 74 don't...
* Propeller 32 resolve into a hostname 16 don't...
These numbers don't add up to the total number of 'unique clicks' - but I see some Ip's have two 'unique clicks'
Point 3 - yes I can I exported the report to CSV to extract the IP's and run the reverse lookups...
Point 4 - will send you a PM but I guess what I found already explains a lot...

So I guess this whole campaign only sent about 80 real visits to the offer ... that's only 50% of clicks

As for the second sweep offer - in Germany...

300 visits from propeller - 56 clicks... 41 resolve into a hostname 15 don't...
241 visits from EZmob - 1 click... (it resolves into a hostname lol...)
Zero conversions
What a difference in LP CTR here ... But I've learned I now need to check for bots as well...

Going to go have to go through this whole lookup exercise for not just the clicks I got on my Landing Page but for all the visits (which are all clicks I paid for since I'm on CPC...)

 

[GriD]

Thanks a Million @Graybeard !!! (now I need to implement it , been busy searching easter eggs with the grandson today)

So what's happening with my campaigns... another day later...
UK sweepstake campaign - running out of budget again and I'm not decided yet if I want to put more money in that one...

More visits on my landing page, a few lore click throughs - no additional conversions today...

Germany Sweepstake campaign is looking better...

EZmob sent 780 unique clicks to my LP's today - 3 clicked through and none converted...
Propeller sent 426 unique clicks to my LP's - 81 clicked through and 3 converted...
(already removed one source on propeller that took 10$ and did not convert, I need to start using automated rules to prevent this...)

This offer seems to have some more potential than the UK one, again propeller traffic seems to convert better than EZmob...

 

[Aiden L]

This offer seems to have some more potential than the UK one, again propeller traffic seems to convert better than EZmob...

Tier 1 is always hard to make profitable without a required budget(I learned the hard way lol). Germany looks more promising and the cpc should be much cheaper as its tier 2. My advice to you with a limited budget is focus entirely on tier 2 geos, they are cheaper cpc and have much less bot traffic than tier 3 geos. After this campaign will you be pursuing with push traffic or testing something else? I'm currently doing google ads and its going ok, they offer a free £75 ad credit once you spend £25.

 

[Graybeard]

Today LP2 has 330 visits, the additional 100 visits made a big impact... CTR has raised from 1 to 4% and it got one conversion...

I have often said you need a sample of at least 576 perhaps you can see why ...

 

[GriD]

After this campaign will you be pursuing with push traffic or testing something else?

That will depend a lot on the outcome... I still have budget with both Propeller and EZmob so I might easily pursue with Push... (remembering it's important not to chase the shiny object )
But they propose other options so I might test something else as well (I also have a Banner ad running with EZmob on the German offer but did not mention this as off topic and it would just confuse the hell out of any newbie trying to follow this thread lol...)

I'm currently doing google ads and its going ok, they offer a free £75 ad credit once you spend £25.

I know about the additional ad credit, this is tempting but as I said before, it is a great bonus if this advertising is part of your plan... Understanding the Google traffic source is a great skill in this market...
@Aiden L do you mean search or display ads? And what kind of offer do you run?
Correct me if I'm wrong, If I want to pursue sweepstakes Display might work for some offers, Search not really...
An option I have in mind with search is click2call...

 

[Dirkskep]

I might be off but what about counting bots by inserting an 'invisible button' and counting the clicks on that button? (as only a bot will see it when reading the code)
But I understand this does not help as you already paid for the click so the traffic source already 'wins' by sending you that bot click...
It will take time but by identifying the IP's you can blacklist the originating publisher feeds if the bot ratio is too high... Your traffic quality should improve over time... (This is a never ending story...)

 

[Graybeard]

1. The GEO, device, os is the same with both traffic sources?
2. look at JUST the IPs that clicked and check their hostnames.
3. can you sort a list in your stats or tracker of this only?
4. let me know --I will run a check for you.

If #1 above is a constant/control it is NOT simply just a numbers game.

US GEO I can profile the economic value of the 'traffic' by micro GEO segment (postal code areas) for UK I can't yet easily.

Usually, it's bots but I got a false positive in Iraq today (looking at an old stats sample).

 

[GriD]

It will take time but by identifying the IP's you can blacklist the originating publisher feeds if the bot ratio is too high... Your traffic quality should improve over time... (This is a never ending story...)

This is VERY valuable information, I wish there was an easy way of doing this (there probably is)...
I really need to sit down and create a procedure on how I will handle this...
@CPAHOUSEOfficial, if your course becomes reality, please talk about this subject in your course, maybe you can dedicate a chapter to 'fighting the bots!' Knowing how to handle this from the start would have greatly improved the quality of my follow along!

 

[Honeybadger]

This is VERY valuable information, I wish there was an easy way of doing this (there probably is)...
I really need to sit down and create a procedure on how I will handle this...
@CPAHOUSEOfficial, if your course becomes reality, please talk about this subject in your course, maybe you can dedicate a chapter to 'fighting the bots!' Knowing how to handle this from the start would have greatly improved the quality of my follow along!

Good that you are sharing this
It will help others avoid similar problems in the future

 

[Graybeard]

well, this can add time to the page load

$hostname = gethostbyaddr($ip);

     if (preg_match('/[a-z]+/' , $hostname)){ $var=1; }
        else{$var=0;}

if($var==0){echo "<p>you're banned!<br>";}
    elseif($var==1){echo "<p>WELCOME!<br>";}
        else{ echo "go fuck yourself";}

you should open a file and print the data: date,adNetwork,source,ip,nohost data
do something with that list

even 41/56 =0.7321 IMHO, for push ads .2679 bots is bad.
My only defense is to lower the bid and see if the CR remains constant --devalue the traffic
10% is my acceptable 'bot-rate' at the suggested bid price

 

[Aiden L]

@Aiden L do you mean search or display ads? And what kind of offer do you run?

Running search ads and its a lead gen offer.

 

[GriD]

I have an extensive list of datacenter ip ranges that has take a lot of my time to build, it's proprietary and not for sale --so don't ask. 2 million ips or so (at last count).

I know you do.. Wasn't planning on asking I'm very pleased by the fact you share your knowledge, up to us to work it!

As you said before :

do something with that list

I understand this is of great value, and bringing it out in the open would devaluate it's value and effectiveness! As no bot operator wants to be on any list...

So for now I'm not adding more $ on the UK Sweepstake offers...
EZmob 50$ budget spent direct to offer and 70$ budget spent to LP's - This brought me 3 conversions in total...
Propeller 50$ budget (still running but almost spent completely) brought 8 conversions...

Second offer still running - slowing down since I'm blocking out publishers sending bad traffic (either bot traffic and/or overspending without conversions) not even reaching the set daily spend..
EZmob campaign 40$ spent of the set 50$ - 0 conversions...
Propeller campaign 40$ spent of the set 50$ - 3 conversions...

@GriD
If you receive great customer support I'm sure they'd appreciate a profile review
Is incredibly helpful, if you have time

Not forgetting this one, first impressions are not always ideal... One of the parties made a statement but has not followed through yet... Planning reviews for affiliate network and both traffic sources with my findings...

 

[GriD]

Update before the weekend. Let's finalize and move on.
I stopped all the Push campaigns, additional finetuning (bid adjustments for working sources and blocking non-working) did not bring any additional conversions.
Looks l cut down the biggest traffic sources as traffic is really slowing down, yesterday only 33 LP visits... Zero LP CTR..

AM directed me to the UK sweepstake offer - I did not get this to work at all, tested direct linking and on 2 ad networks through 2 LP's
One of the ad networks pointed me to the second offer - I did get some conversions here but nothing spectacular .

Staring from scratch, what did I learn here...
- Use several creatives and monitor their performance
- Traffic source - lock out non working sites (overspend without results) and eventually increase bid on the publishers that do perform (LP CTR / Conversions)
- Bot identification - Lock out Site id's and publishers sending bot clicks

I had to work everything out manually and was running behind the facts as I was learning on the go what was going on and how to handle it... But that was the whole point! Learning
I need to dig a little deeper in the tracker stats as they might help me here...

If you plan on running a push campaign you will need to work out a procedure of actively monitoring that bot traffic as to act way faster than I did locking them out.

Before writing my traffic source reviews I will send them my findings and if (and how) they respond to this.

Thinking of where I should go next, the results I got were so bad that I'm not sure if I want to pursue Push traffic now or come back to it later... I'll be happy to read your ideas here...

 

[Sheenue]

This is VERY valuable information, I wish there was an easy way of doing this (there probably is)...
I really need to sit down and create a procedure on how I will handle this...
@CPAHOUSEOfficial, if your course becomes reality, please talk about this subject in your course, maybe you can dedicate a chapter to 'fighting the bots!' Knowing how to handle this from the start would have greatly improved the quality of my follow along!

Hey there. You can find pretty good post about bots on binom blog (even if you're not using binom).

My advice for your journey: try some really cheap traffic with some cheap payout offers on cheap countries just to get experience in zones, os, carriers etc optimization.

Also use smartlinks if you're not sure what offer to use (for example if you chose sweeps, take trafficcompany smartlink, pin sumbits - goldengoose for example, adult dating - imonetizeit or smaller network.

I don't wish you luck, I wish you patience!

 

[GriD]

Hey there. You can find pretty good post about bots on binom blog (even if you're not using binom).

Thanks for the link... Unfortunately I'm no good at Javascript/ajax etc... So it's hard for me to understand how his scripts help hem... Not sure if I need to learn scripting.. I guess I will be better off trowing this at a programmer... Can't keep on doing everything myself... Even though I love to do it so I understand what is going on...

My advice for your journey: try some really cheap traffic with some cheap payout offers on cheap countries just to get experience in zones, os, carriers etc optimization.
Also use smartlinks if you're not sure what offer to use (for example if you chose sweeps, take trafficcompany smartlink, pin submits - goldengoose for example, adult dating - imonetizeit or smaller network.
I don't wish you luck, I wish you patience!

Thank you @Sheenue , I know it is not about luck!!!