Home > > Newbie Follow-Alongs

---

Total Newbie 40 Day Follow Along (23)

---

Hello everyone, my name is Huy and I joined STM about a week ago. I'm a total newbie to the affiliate marketing world. I've decided to buckle down and take action following Amy's 40 Day Tutorial.

Thanks for following on my journey.

My Goal:
Take massive action to get through the learning curve. From reading just a ton of forum threads in this forums, I'm already getting quite the information overload so I really want to focus on one thing and learn the fundamentals as much as I can for now. I'm not worried about profits or anything yet. Just want to master the basics and fundamentals.

Budget:
$1,000 / month

Day 1 Objectives Completed:
-Approved by Mobidea
-Signed up to PropellerAds
-Signed for Voluum Tracker
-Registered and set up custom domain for Voluum

---

Day 2 Objectives Completed:
-Applied for about 10 offers. Got approved on all of them.


-Created my offer spreadsheet for the offers that I will be running. I plan on testing out just 5 campaigns and then gonna move on to the landing pages section because of Amy's suggestion on another post.




Day 3 Objectives Completed:
-Set up the Voluum tracker
-Created new offer & campaign in Voluum
-Tracking is still a bit confusing to me so I'm reading caurmen's guide on understanding it.





Day 4 Objectives Completed:
-Successfully set up my very first campaign. It felt good.
-Continued to set up all the other campaigns from my offer spreadsheet.


Here are the results from my 5 campaigns so far. I got my very FIRST CONVERSION from one of my campaigns. That felt pretty awesome to just check on Mobidea and hear that CHA CHING sound . ROI is looking pretty bad but I'm not worried as this is all a learning experience.




Just to DOUBLE-CHECK that Mobidea is triggering my conversion and the tracking is working properly:



Even though I got 2 conversions, none of my 5 campaigns that I launched passed the kill-whitelist calculator test to keep running. I pretty got the hang of launching campaigns. The whole process of transferring from Mobidea (affilite network) -> Voluum -> PropellerAds .

Question: I got accepted into a couple of other affiliate network such as ClickDealer and YepAds. Is there a link to a guide to doing the postback URL set up for other affiliate networks other than Mobidea?

---

Congrats on the first conversion! And kudos for starting a follow-along!

For clickdealer do a search on this forum for "clickdealer postback" and you'll find lots of threads. This one for example just got posted to today:

https://stmforum.com/forum/showthrea...on-clickdealer

Yepads, like clickdealer, also uses the CAKE software - so the postback url would be similar.

Feel free to ask further questions when you get stuck!


Important: Now that you have experience direct-linking to offers, please proceed to the latter lessons in the tutorial to learn how to work with landing pages. Please see this thread for details:

https://stmforum.com/forum/showthrea...Major-Overhaul


Have fun Huy - I'll be following-along!



Amy

---

Hello!

Congrats on your first conversions! Keep on going!

Sure you will meet a green ROI soon by staying consistent and working hard

---

Thank you Amy!

Thank you

---

After reading some of the explanation on Testing Bids, Applying to more traffic & ANs, I moved straight to landers section.

Successful Objectives:
-I successfully set up Hosting, CDN, DNS and got the test lander to work correctly.


-I signed up for Adplexity and downloaded 10 landers for an iPhone 8 offer+geo I found through YepAds and ClickDealer.

-This morning I got an email from Voluum saying how they were boosting their prices up to $249-299 . To a beginner at affiliate marketing, that is very discouraging to see. Some of the features they were putting out to justify the price jump didn't even apply to me so I felt like I would just be paying for a higher price for no reason if I decide to stick with Voluum.

Okay so what I did was that I switched trackers. I switched from Voluum to Binom. I did all the set up mostly myself with a some help from Binom support.
-I had to get a DigitalOcean account
-Had to install the tracker onto the server
-Had to redirect my namecheap domain from Voluum to Binom
-Set up all the ANs postback urls in accordance with the new tracker

After I set up everything, I ran a test DL campaign to one of my offers from Clickdealer to make sure everything was working smoothly and traffic from my traffic source, tracker, and AN were correct.

Here is data from Binom:


Here is data from PropellerAds:


Here is data from ClickDealer:


It seems like ClickDealer's data looks offcompared to Binom and PropellerAds. Any ideas why?

---

I'm really impressed you were able to figure out all the tech!

It IS a shame that Voluum decided to jack up their prices. I'm going to have to rewrite the tutorial for another tracker - probably Funnelflux.

The clickloss looks really big! Are you sure the time zone are all set to the same? Try to choose a bigger date range on clickdealer to make sure you "catch" all the data.

How many offers are you testing with the 10 landers?




Amy

---

Hi Amy,

I haven't tested any offers with landing pages yet. This campaign with the clickloss is a test Direct Linking campaign to see if my tracker is working properly. I checked the time zone and looked at the bigger date range but it still shows the same data.

I ran it for $10 to Brazil and PropellerAds spent all of the money within 2 hours.

---

That is strange.

Have you asked your AM at clickdealer to see if they could shed any light on the discrepancy?

Also - I would suggest to add another offer to the same campaign - one from another affiliate network -and compare clickloss.

If you get the same large clickloss, then perhaps you should look at your Binom installation in greater detail. I've never used binom so have no idea what you'd need to check. For Voluum for example I would check the custom domain settings.

That's all I can think of for now.




Amy

Sent from my SM-G930W8 using STM Forums mobile app

---

Hi Amy, so I ran another offer on another Affiliate Network (Mobidea) to compare the click loss. Here are the results with $10 budget

This is the result from Mobidea:

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



This is the result from Binom tracker:



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

This is the result from PropellerAds:


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



So it seems to me like the Affiliate Network's tracking is off by 10,000+ clicks.

I asked my AM from Clickdealer to see if it's any errors from her side and she said this:


Are there any chances that there is bot traffic going on?

---

So I've been figure out this clickloss issue that I've been having and the support team from Binom helped me switch the location of my server to Singapore to help with the DNS speed. I ran a test DL campaign again to see if it made a difference and it helped out a lot but there still seems to be a big clickloss issue from my affiliate network to the tracker.

During all this time, I looked for offers in 1 geo+1 vertical. I downloaded over 10 landers and edited the hell out of them. This seems to be the most time consuming and tedious task. Especially with all the different languages.

I finally grouped up my offers categories of each GEO and this is my first campaign in my first GEO that I'm launching. It's grouped in 1 geo+1vertical+1carrier.

It contains 3 offers with 5 landers. Which in turn gave me 15 paths for this campaign.

My budget will be $120 into this campaign. I'm gonna launch this and then will post results of the campaign. I'm excited as this is the first mass-offer campaign I've ever launched. WHOO HOO lets go!

---

Here are the results for my campaign.

After $120 spend, I got 2 conversions for $2.40.

Here's the result:

The 2 conversions both came from 1 particular lander and 1 particular offer aka both conversions came from 1 Path.

After $50 ad spend, I decided to cut off all the other 2 offers and just let the rest of the ad budget rotate to the 1 particular offer that got me the conversions to see it would stick to the other 4 landers I had, which it didn't.

When I was checking the zoneid's in PropellerAds, there were 3 placements that cost me $10-30 dollars without a single conversion. I feel like maybe I should have excluded these placements while the campaign was active. I didn't check all the results until the campaign ended.



I feel like the 1 lander that got me conversions is the most likely one that will help me convert. I have 2 other GEOs with the same vertical to test out and I will definitely include the lander from here to see if it will convert in another GEO also.

---

Just curious - Have you been able to solve the click loss issue since moving server location?

---

1) Clickdealer counts unique hits, and the tracker counts everything, so there may be a discrepancy in the stats.
3) Losses between Propellerads and any other tracker of 8% (screenshot) are normal for pop-traffic.

---

Less than 20% clickloss is still acceptable for pop traffic in general. This thread addresses some of the possible reasons:

https://stmforum.com/forum/showthrea...nearly-60-loss

And yes - bot traffic is one of the potential causes.

DNS speed is a big one as well. Basically there are several redirects each visitor has to go through: Traffic network > your tracker > your lander > offer (aff network + advertiser). And every step of the way you will likely "lose" visitors.

For the 1st step (at the traffic network): The link above discusses some of the causes there, including bot traffic, and how the tracker counts impressions differently from the traffic network.

For the next 2 steps (to tracker and to lander), speeding up DNS can help. For landers, optimizing them for loading speed will also help lots.

For the last step with the offer, that's pretty much beyond our control. The only thing you can do to optimize there is pre-fetch/pre-render:

https://stmforum.com/forum/showthrea...l=1#post254366


Woohoo to mass-testing!!

I've never used Binom, but did you manually set up 15 paths? In other trackers, you only need to specify the landers and offers and the tracker would automatically rotate evenly through them (or at any other rotation weight you specify). I don't imagine Binom to be different.


Great start! But there are several, very valuable lessons you can learn from this spend:

1)YES to cutting placements from the start! You can set up your only rule-of-thumb, e.g. if a placement doesn't convert after spending 4x average offer payout, cut it. How you specify this rule-of-thumb would depend on how aggressively you're wanting to cut placements, which would in turn depend on how much total traffic you're getting and how quickly you're wanting to get this testing done.

For example, if you have lots of traffic and want to conserve costs, then set an aggressive cutting rule. You can always retest these placements later (and should - if you're cutting aggressively at the start), when you have identified your best offer+lander - those same placements may be profitable then. But cutting them out in the beginning can save you lots.

If you're running in a small geo though and already not getting a lot of traffic to start with, then you may want to be less aggressive when deciding on a rule-of-thumb for cutting placements.

2)Don't spend too much money testing each offer. It's not straight-forward to come up with a rule-of-thumb here, but I'd say that if you're running popular and optimized landers, if an offer doesn't convert even once after a total of 10x payout (so 2x payout per lander), I won't waste any more money on it.

It's also important to use proven offers in the beginning (when testing landers) as much as possible. This means: Offers recommended by AMs as having done well for at least a few other affiliates; offers you've seen in spy tools that have received lots of traffic from multiple affiliates; offers that have otherwise shown promise.

Once you've cut down to a winning lander using an offer that converts at least semi-well, you'll have a benchmark that will allow you to test any other offer you want. But in the beginning it's critical to use promising offers and landers and traffic - the 3 main variables of pop - as much as possible. Worst case is when you're not getting conversions and have no idea whether the offers are duds or the traffic is bad or the landers are bad.

There IS a possibility that all the offers you test initially would turn out to be duds. In that case, you can either swap out all the offers for a new batch, or if you feel you've already talked to all your AMs and there aren't many other recommended offers to test for that geo+vertical, you may want to test another geo/vertical altogether.

3)Use the most promising offer to cut landers. What I usually do is run several offers until one of the them reaches 2 conversions, then pause the others and just keep running that one offer to give me enough conversions to cut landers down to a winner. No need to wait for statistical significance when choosing this offer, because we're not trying to determine which offer is the best - we're only looking for an offer that converts well enough to help us cut landers. And an offer that reaches 2 conversions first is a pretty good indication (assuming it doesn't spend excessive amount of budget to make those 2 conversions - this is why the rule-of-thumb I mentioned earlier, the need to cut an offer that spends 10x average payout spend without converting, needs to be in place).



I hope you weren't discouraged by the results! It's difficult to get used to mass-testing, so you're already miles ahead of many newbies that are afraid of spending money.

But now that you've brought yourself to start mass-testing, we can tweak your approach to make sure you're testing for as cheaply as possible - and the 3 lessons above WILL help you to do that. Without knowing exactly how much you've spent on each placement, or how much was spent to acquire those 2 conversions, I can't say for sure how much money you could have saved by applying the 3 tips above. But my estimate/guess would be that you would have saved half of your budget or more.

Also know this: Once you HAVE run a campaign in a certain geo, and have invested money into cutting placements and identifying a good lander, next time you target the same geo on the same traffic network you won't have to start from scratch! Just regularly ask your AMs if they have more new and promising offers in that geo+vertical, and test them on the proven lander and good placements (with the bad ones already eliminated from a previous campaign). A common misconception is that mass-testing can drain your budget, but hopefully you can see how that doesn't need to be the case.

You're progressing nicely! So what's next? I remember you've prepared landers into several other languages? Looking forward to seeing you test those!



Amy

---

Sure, Binom rotates through all landers and offers in one path, either evenly or at whatever percentage you set it.
But of course you need to create different pathes.
You don´t want to mix Amazon Gift Card landers with AV offers for example.
So you create Path 1 for AV landers and offers, Path 2 for Amazon Gift Card landers and offers, Path 3 for direct linked mobile content offers and so on.

---

Yup that makes sense thanks twinaxe!

I was under the impression that the 3 offers + 5 landers were for the same type of offers.




Amy

---

Yes, then it would be little bit too complicated to set up own pathes for each comobination

---

Hi Amy,

Thanks for all the tips! Greatly appreciated.

Ok so I did it the long ass way. All my offers were the same . I could have put all the landers+offers into one path. What I did was I created a path for each offer+lander combination. Thanks for the tip. It'll save me a lot of time haha






So here's what happened. After seeing 2 conversions for that particular offer+lander combo, I decided to put more money into it to see what will happen. I gave it another $40 dollars ad spend.



Only after about 1 hour, I got 2 more conversions. The ROI is -61% for this offer+lander combo and in my mind this might show some potential. So I went to sleep and let it run the rest of the adspend.







After I woke up , this is the final result of the additional $40 ad spend. I got only 1 more conversion and the ROI is at -89%.



Total adspend for this campaign: $160
Revenue: $6.00
Profit: $ -147.95

I'm not sure after seeing the ROI that if I should take the time and money to try to dig in at this offer+lander combo. Any suggestions?



Next Focus:

Right now, I want to focus on increasing and optimizing the speed of my landers. I resized the images and did all the tips in the tutorial but the loading times for my landers still seem to be in the 2-3 seconds . One of them even went near 4 seconds. I read and saw how important it is that your lander speed be when running POPs and I don't want to lose out on CTR by ignoring the issue.

Does anybody have any suggestions? I'm running Amazon S3+Cloudfront. I plan on having a developer put all my images into a sprite file and see if that will decrease my loading time.

---

Sorry to butt in on this thread! As newbies should we be trying to find offers that are the same (such as Iphone X) to test for landers? Or is it okay to find different offers like you said...Amazon gift card, supermarket voucher, Iphones, etc. and create separate paths? Does it make a difference at all for narrowing down landers?

---

You setup different pathes for different offers/landers.
Here is an example:
Path 1= 3 iPhone X landers and 4 iPhone X offers
Path 2= 4 Dating landers and 4 dating offers
Path 3= 2 AV landers and 3 AV offers for carrier XYZ

Then you set a percentage for each path, this is to set how much traffic each path receives.
When you set each path to 100% then they all will receive same amount of traffic.
You can also set the percentage different so that each path receives different amount of traffic.
For example you could give the path with most lander and offer combinations most volume and the path with less lander/offer combinations less volume.
Then the traffic will be routed to the paths and Binom will then send traffic to each matching lander/offer combination so that you can track what's working best.
That's why you group the different offers and landers in paths.
You don't want to mix sweep landers with AV or dating offers for example because it just doesn't make sense.
You group them together that they match.
iPhone landers with iPhone offers, Amazon Gift Card landers with Amazon Gift Card offers, dating landers with dating offers and so on.
It's absolutely ok to find different offers, and in fact you even should test different things to find your winning combination.
Just make sure that you don't mix the different offers and landers.
Offers and landers should always fit together.
Also you shouldn't mix different campaigns together.
For example you shouldn't mix WiFi and carrier traffic.
Or you shouldn't run different Geos in one campaign.
So first thing in your tracker is to create separate campaigns based on targeting like Geo, vertical, traffic type, OS, and device or whatever.
Inside these campaigns you separate the landers and offers that match together in different pathes.

---

Thanks, twinaxe! That makes more sense. It sounds like a great way to mass-test a bunch of offers within a vertical. I will try it this way!

---

I would definitely have stopped that sooner - unless there were a few big placements that have been zapping your budget without converting, that you could eliminate to DRASTICALLY increase the ROI.

Can you show some placement stats please? Here's what we'd be looking for: Whether there are enough green placements making enough profits to justify keeping the offer+lander running.

If you have enough good placements that are jointly making enough daily profits to justify your effort in managing this campaign, then all you need to do is keep cutting the unprofitable placements.

But if you DO NOT have enough good placements giving you enough daily profits to start with, then no amount of cutting will net you profits.

Hope that makes sense! It's like blasting away rocks to find diamonds - there needs to be diamonds in the first place to justify all the effort of eliminating the rocks.

(The same thing applies to other traffic segments, not just placements. e.g. OSs, carriers, browsers, whatever else - basically you need to find enough green, then cut the red.)

Did that exercise generate enough conversions to help you cut down to one lander? But yeah - before you start testing landers again, make sure the load time is below 2s - should aim for below 1s actually, but 2s is the absolute maximum.


Try converting the images into base64 code.

If you're still having issues: Rip more landers, and test them for speed BEFORE you even fix them up. If they load really slow, pick a faster one to work on.

Failing that: Post the lander code in pastebin or similar, and post a link here and let's take a look! I suck at coding, but maybe someone else can help.



Amy

---

Home > > Newbie Follow-Alongs