Home > POP / PPV / Redirect > Follow-along Campaigns

---

Absolute Beginner Following Vortex's 40 Day Pop Tutorial (Binom) (13)

---

Here's a recap of my progress for the past few days.

Currently following two other members' follow-along who are using Binom:

• GregMorrison - https://stmforum.com/forum/showthrea...g-(with-Binom)
• sheet731 - https://stmforum.com/forum/showthrea...ing-with-Binom

20181029

Day 1 - Some Preparation Work

• Signed up for Mobidea and PropellerAds
• Registered a domain for tracker on Namecheap
• Set up Binom on DigitalOcean
  
  • There's a Binom promo code for STM members, check the forum
  • I also started during Binom's support team's off hours so there wasn't anyone to help me set up the tracker. I decided to try setting everything up on my own and it took a few hours into the night to get it up. For those with more patience, I would definitely recommend having the Binom team help. I had to install the tracker twice to get it right.

Day 2 - Picking Offers

• Signed up to 10 different offers for Tier 2+ countries on Mobidea
  
  • I chose offers by sorting by Revenue and looking at the CR. If a CR is 1/XXXX (four digit) or less, I signed up for. I assume this means a higher chance of conversion.
  • I avoided offers with complicated restrictions.
  
  
• Set up a spreadsheet to list all 10 offers
  
  • Also using the same spreadsheets to begin tracking all my expenses

20181030

Day 3 - Setting Up Tracking

• Binom's Postback URL is in Settings of the tracker
• The Global postback for Mobidea should look like this:
  Code:

  https://DOMAIN/click.php?cnv_id={{EXTERNAL_ID}}&payout={{MONEY}}

• Adding Aff. Network in Binom
  
  • If you're loading from template, the tracker is using an old token {{DATA1}} and should be replace with {{EXTERNAL_ID}}.
  
  
• Offer Final Link
  
  • Big ups to Greg, who I'm following, for providing Binom tokens for the tutorial. Vortex's guide uses this:
    Code:

    data4={campaign.id}&data5={trafficsource.id}&tag={clickid}

    • Binom uses different tokens and instead of {campaign.id}, use {campaign}. Instead of {trafficsource.id}, use {trafficsource}.
    
    
  • However, somehow I messed this up and this affected my Day 5 (will get into details later) and ended up with
    Code:

    https://www.rotatemyurls.com/offer/XXXXX|XXXXX?website={subID}&placement={sub_subID}&campaign={campaign}&trafficsource={trafficsource}&clickid={clickid}

  • I forgot to remove `website={subID}&placement={sub_subID}` and I deviated from Vortex's URL parameters. The correct link should be:
    Code:

    https://www.rotatemyurls.com/offer/XXXXX|XXXXX?data4={campaign}&data5={trafficsource}&tag={clickid}

  
  
• Creating Offer and Campaign in Binom is straight forward.
  
  • I chose 3 out of the 10 offers I applied for to start.
  • I created individual campaigns for each offer with a total of 3 campaigns.
    
    • One thing to note when you're creating a Campaign, the Path would be +Direct then +Offer and choose the offer for the campaign.

Day 4 - Setting Up a Campaign on PropellerAds - Part1 + Part2

• I added funds to PropellerAds. It seems to require additional authentication with your credit card company. I had to call them to approve the transfer.
  
  • If you check out sheet731's follow-along, I used the same promo code he used for an additional $50 credit.
  
  
• I followed all the steps to set up 3 campaigns. PropellerAds approved 2 campaign within 30 mins. I was exhausted at this point and went to bed.

---

Binom is recording more clicks than PropellerAds?

Try SmartCPM in propeller ads, you'll save money.




LeadCloak

---

I realized the difference in clicks between the platforms was because I was only looking at the "Today" numbers. Looking at the Last 7 Days or This Week shows a more accurate number.

After rerunning Campaign 1, here are their numbers:

Campaign 1 - Max Bid: $1, Spent: $12.87, No Conversions

• Binom: 11,637
• PropellerAds: 12,867
• Mobidea: 10,434

No conversions so I still haven't been able to confirm if the tracker is working properly. Is there another way to test this?

---

I don't think there's another way to test.

also not sure if anyone's told you, but the current suggestion is to only skim the lessons from day 5 up until the landers days. so learn about landers and coding asap.

if you post screenshots of your setup here, that might be helpful too as a sanity check

---

good luck man! thanks for your help on the JS script stuff today, looking forward to more progress

---

Great read, keep it up!

---

It has been a busy week but I wanted to keep adding on to my progress journal here. Definitely would love any feedback!

20181104


Day 22: Downloading Landers from Adplexity

• Adplexity quickly shows how similar many landing pages are for multiple GEO.

Day 23-26: Fixing Up Landers

• Following up on similar landers, it makes it easier to work with because I start to see the common use of tokens and scripts being used on each lander. It definitely helps to know HTML/CSS and some Javascript.
  
  • Make sure Javascript code is in place for tokens to work properly.
    
    • For example, to show ISP, you would use this:
      
      Code:

      <script>document.write(getURLParameter('ISP'));</script>

    • But for it to work, you need to ensure this script is included:
      
      Code:

      <script>function getURLParameter(name){
         return decodeURI(
             (RegExp(name + '=' + '(.+?)(&|$)').exec(location.search) || [, null])[1] || ''
         );
      }
      </script>

    
    
  • Reuse landers but change the product image. If I’m running landers for iPhone X I can reuse it for iPhone 8 if I manually change the product image. Easy to find on the internet or apple’s official site
  
  
• Definitely test on mobile to see how it looks, especially if you’re running mobile campaigns
  
  • I almost made the mistake of launching one because everything looked fine on the web version. Luckily, I double checked on my phone and spotted an issue.
  
  
• Be detail oriented. Even if a lander is in another language, make sure to do a simple Google translate to see if some of the Javascript or token parameters didn’t get downloaded
  
  • An example is a sentence that translate to, “Hey, we have a special offer in .” Note the space and potential token that can be used to make that sentence work. If you don’t check, you might be missing a critical part that make the lander work!

20181105


Day 28-31: How to Test Landers to Find Winner

• Bot testing - I used https://stmforum.com/forum/showthrea...Live-Campaigns and it works great but as a beginner, it was very overwhelming and hard to decipher the data I’m looking at.
• @sheet731 directed me to Binom's own mobile bot tester (link to documentation: https://docs.binom.org/events.php). It makes the data easier to understand by following the same guidelines from the original bot tester: If bot traffic is > 80% with 60+ impressions, I would cut the zone.

20181106 - 09


I have ran a few campaigns across multiple GEOs and offers. Seeing conversions and have confirmed that my tracker is finally set up correctly. Very exciting!


Also, I was accepted into a new network, Advidi, and have started to run their offers as well as Mobidea’s.


Questions

• For split testing, if I haven’t gotten enough data to determine a winning lander, should I use CTR to determine which one to keep or should I wait for conversions first?
  
  • Follow up question, should I be pausing landers during my initial test run of campaign? Or should I wait until the budget is spent first?
  
  
• Should I continue a campaign if it has a low payout even though it has conversions? Using the Kill-Whitelist-Calculator, it tells me to kill the campaign. Calculator is referenced in Day 5-6 Part 2: https://stmforum.com/forum/showthrea...-ACTION-Part-2
• Should I be focused on multiple GEOs or solely focus on one and master it?
• Is there a guide to learning more about Adplexity and interpreting the metric for each lander drill down?

---

Facing a new issue where my domain has been flagged, twice in a week after running a converting campaign. Any ideas why?

---

First of all - many, many apologies about my late arrival. Been so occupied with the new tutorial I'm dropping the ball on things - will strive to be more on top of things and reply more frequently!

I'm very impressed by your systematic and organized way of recapping all the information! I'm sure other new members will find value in this follow-along as well - thank you so much for the detailed documenting! And your coding tips are invaluable as well.

Actually - and sorry for the confusion - but "data4" and "data5" were recommended by Mobidea IN THE PAST which was why I wrote them into the previous version of the tutorial:

https://stmforum.com/forum/showthrea...ng-Up-Tracking

The new tracking tokens as confirmed by Mobidea when I spoke to my rep there recently, were suggested in the updated tutorial:

https://stmforum.com/forum/showthrea...king-on-Voluum

data4 and data5 may (should) still work though.


There may not be enough traffic based on your targeting and bid.


Good advice!

When I was writing the old version of the tutorial, SmartCPM had not come out yet. This is recommended in the new version.

https://stmforum.com/forum/showthrea...ds-EXPLANATION


Also explained in the link above (written a few days after your post - good timing!)


Some networks will either allow you to trigger a test conversion, or do it for you - speak with your affiliate manager to see. But the easiest way would just be to wait for a conversion to occur.


sheet731 is absolutely correct! And I'm very glad you've noticed it. In fact, this is one of the most major changes to the tutorial, and the biggest reason I'm rewriting this updated version.

Please see this updated lesson - noting the part in red especially:

https://stmforum.com/forum/showthrea...l=1#post357744


Payout should be "auto" - the postback that you put into Mobidea will include the payout when it posts back the conversion data.


You can pass conversions to PropellerAds too if you like, but I wouldn't recommend it. But if you wish to do it, Greg tells you how in this post (and you'll see my comments in the post below that one):

https://stmforum.com/forum/showthrea...l=1#post358726


It's actually better this way - because every time you make a change in the lander code, if you want to view the changes right away, you can just browse to the url version without the www (whereas the CDN version may still be showing the pre-edited lander).


Wow so you guys live in the same city? Always nice to meet other affiliates face-to-face!


CTR is usually not a good judge of a lander, or a placement.

I can use a ton of tricks to boost CTR - for example promise the moon, or even just automatically redirect visitors to the offer after x seconds - but it doesn't mean the visitor will convert when they get to the offer. For the examples I've listed, visitors would feel cheated when the content on the offer page doesn't check out, and they also haven't been pre-sold to sufficiently, so CR (conversion rate) would suffer.

So, judge a lander or placement based on CR, not CTR.

Having said that, there's an exception.

If the CTR is so low that it would require an unrealistically high CR for the campaign to break-even, then you KNOW the lander's a loser.

And when judging placements, you can view placement CTR for a SINGLE lander, and if most of the placement CTRs are in a certain range, but some of the placements have CTRs that are outside of that range, then you may want to cut those placements - as something fishy is likely going on (e.g. bot traffic).

e.g. Say most of the placements are showing CTRs of 30%+ for a certain lander, but then a few of the placements are showing -3%, then you should be able to safety cut those out without waiting for conversions.

And what I said about landers applies to placements as well: When CTR is so low that it would take an impossibly-high CR for it to break even, you can forget about that placement.


You can cut a lander as soon as it reaches statistical significance, regardless of budget:

https://stmforum.com/forum/showthrea...Banners-Part-1


That calculator is only good for judging direct-linked offers (as I've mentioned in that lesson).

It's great to use as a crutch when a newbie is analyzing stats for the first time, but it has 2 major flaws:

1)It's more accurate than efficient. An experienced affiliate would pause that campaign probably way before the calculator will tell you it's hopeless - this = more money wasted on collecting more data before you'll see a verdict. You may have seen one of my many rants about the accuracy vs. efficiency dilemma that we're always running into, when making campaign optimization decisions. For pop, efficiency trumps accuracy most of the time.

2)It's only remotely accurate when you don't change the current set of conditions. When you're just direct-linking to a single offer, without doing any sort of split-testing, that calculator may be half-accurate - it still won't be very accurate because you'd be cutting placements, which will improve traffic quality. And when you're split-testing offers and/or landers as well, that calculator becomes even less accurate (a LOT less probably). So - time to ditch that calculator.

Continuing with a campaign or not, does not depend on whether the payout is low. It's all about what the current ROI is, how far from ROI=30% you are (to me that's the minimum ROI I'd accept unless I'm doing volume), and how much more optimization you can still do.

For example, if you're running in a big geo with lots of traffic, and you're split-testing 3 offers and 5 landers, and getting ROI = -70%, I would say so far so good! This is because you've still got LOTS of room for optimization - once you have a winning lander and a winning offer, you'd be so much closer to breaking even. Then you can use that winning lander to test more offers which can potentially get you even closer. And then you can cut a lot of placements and still have enough traffic left to profit from, because it's a big geo.

I'm sure you get my point. Of course it would take running lots of campaigns to get a "feel" for when to continue a campaign or not, but you'll get there by keeping the above in mind. I wished I could provide some rules of thumb to go on, but every situation is different.


I wouldn't limit myself to a single geo - you need to test more extensively than that to increase your chances of finding profit. Plus, in order to be making consistent profits, you'll need at least several geos.

Sticking to a few geos though is a good idea, because then you wouldn't be starting from scratch all the time. You'd already have a blacklist of placements, one or more proven landers in at least one vertical, and maybe even knowledge of the culture (best converting times, what lander angles they respond to best, etc.) So when you get a new offer to test, you can just quickly throw it into the existing campaign with the proven lander, spend 5-10x payout on it, and be able to tell very quickly whether the offer has potential. (i.e. Because traffic quality and lander are both proven, if the offer doesn't start performing well right away, you'd know it's a dud.)


I've put together a list of threads here:

https://stmforum.com/forum/showthrea...flowguide-here


This has become a common occurrence, both for tracker domains and lander domains. There are lots of threads with good recommendations on how to handle this (read through the entire threads and not just the first posts - lots of good advice):

https://stmforum.com/forum/showthrea...ent-domain-yet

https://stmforum.com/forum/showthrea...flagged-domain

https://stmforum.com/forum/showthrea...ed-anyone-else

https://stmforum.com/forum/showthrea...etting-flagged

https://stmforum.com/forum/showthrea...wsing-phishing

https://stmforum.com/forum/showthrea...omain-flagging

https://stmforum.com/forum/showthrea...gged-by-google

https://stmforum.com/forum/showthrea...gged-by-google

https://stmforum.com/forum/showthrea...flagged-domain

https://stmforum.com/forum/showthrea...ks-for-any-tip

https://stmforum.com/forum/showthrea...our-lp-domains

https://stmforum.com/forum/showthrea...lagged-domains

Hope that helps!

Now that I'm all up-to-date on your follow-along - looking forward to seeing future updates!




Amy

---

Thanks for the detailed reply Amy! I really appreciate your time!

Notes
- One thing I've noticed from a converting campaign is that a good offer page tend to convert better. If you're on the iPhone and testing the offer links, use the Safari browser. I'm not sure why but when I use Chrome (the browser I normally use), testing a offer link would send me to a different offer.
- If you're using Binom's bot test script, make sure to replace the sample domain with your own! I got into the habit of copying and pasting it from Binom's documentation that I forgot to change the domain. Silly mistake, so I save the script with my tracker domain somewhere handy so I can reference it the next time I use it.

Questions
- I have been testing more offers. I usually test a minimum of 3 landers for each campaign. After the initial $10, if I don't get 2 conversions, I tend to move on. Is this the best approach? Should I expect conversions during my initial test?
- When an offer allows leads from all carriers, how should I set up my campaign? I'm currently just testing Mobile only (or targeted carrier based on offer) and Other (only Phones) on PropellerAds. Should I also drill down further and create a camp for each carrier?

---

It may be (it's BETTER be!) because the offer doesn't accept chrome traffic?

If this is NOT the case, you should notify your AM - they may be redirecting visitors incorrectly.

In the meantime, consider taking chrome out of your campaign targeting (which would require that you target IOS devices in a separate campaign so you can blacklist chrome), or at least keep an eye on conversion rates for this OS+browser combination.


Glad you figured that out!


I'm assuming your payout in this example to be $1?

If this is a smaller geo where you can't do massive cutting of placements etc., and you're confident your landers are good performers (have received lots of traffic over a few days according to Adplexity; load fast display well function correctly), then it may be a good approach.

Any rule of thumb may end up cutting an offer that may have ended up profitable - it would be difficult to find he perfect balance between not wasting money on testing and not cutting out potentially promising offers. But I'd say your rule is as good as any other.

For the bigger geos though, where a lot of placement cutting may be required to see green, you can go from very negative ROI to profit by cutting heavily. In that case, your rule of thumb may work less well.


For initial testing, if you want to save time, you can just target all carriers to see how each does.

Ultimately though, it may be better to target them separately, especially if you're getting lots of traffic from each carrier.

This is because prices are likely different between traffic from different carriers. So by targeting each separately, you'd avoid over/underbidding.



Amy

---

I think I have my lander page workflow down and can fix a bunch of landers within an hour and have it live.


I have been running campaigns here and there for different GEOs (NL, SG, ZA, FR). No green campaigns yet.

• Is there a list that outlines the minimum offer payout to be worth pursuing for each GEO?

I feel like $10 may not be enough for data gathering. Maybe my starting bid is too high? I’ve been using $2 with SmartCPM. I need to run $25 worth of traffic to have enough data to start cutting zones (bots and placements).

• In Day 5 about cutting placements, the non-aggressive rule is we should cut placement if Profit > 2x payout in loss with no conversions. But what if the initial test doesn’t give enough data to determine this? Should I reduce this to 1x or 1/2x?

At the moment, AM at my current level feels like it’s all about luck. If other AMs are running the same landers for the same GEOs, aren’t we just competing with each other to see who has the better offer?


Am I missing a piece of the puzzle?

---

That is a feat in itself!


Why not show some stats here so we can analyze them together?

Cracking into a geo can take quite some testing. How many landers have you tested? How many offers? Is the vertical you're testing even one of the prominent ones for that geo at the moment?

You've picked some hard geos to start - NL, ZA, and FR aren't easy. SG is easier by comparison.

The basic approach would be to test landers and offers until you can make a large-enough portion of the total traffic profitable, then focus on cutting the unprofitable. This applies to all geos - but the "harder" a geo is (i.e. the more competitive it is and the more volume it has), the more testing of landers and offers will be involved before you can focus on cutting, and the more you'll need to invest into cutting because there's so much volume.

My advice: Either pick tier 3/4 geos that have good volume, OR if you want to stick with the geos you've listed, pick 1-2 and focus on testing LOTS of landers and offers (emphasis on OFFERS) until you find something promising, THEN focus on cutting.


The payout is just one part of the equation. The other important variable is conversion rate.

e.g. If you're running 2 offers, OFFER1 pays $5 but only converts 1 visitor in every 1000, OFFER2 only pays $1 but converts 1 visitor in every 100, which would you rather run?

You get the idea.

Having said THAT: Offers that have high payouts stand to have better profit margins. Note I said "stand to have" and not "tend to have". What I mean is: Given 2 offers, one with high payout and the other with low payout, if the high-payout offer converts well, it can make a lot more profits than the low-payout offer. The conversion rate though will always be important whatever the payout is, and the reality is this: There is almost always a reason for a payout to be what it is. Meaning, advertisers aren't stupid (most of them aren't anyway) - they would pay you more for a conversion only because it's worth more to them, but this also means you'd need to either make the visitor jump through more hoops (i.e. a more-complicated conversion flow), or spend more money to bid for the higher-quality traffic, or both. Typically, high-payout offers don't convert as well as low-payout offers. In the end, you just won't know until you test an offer - because until you can find out what the conversion rate is, you can't draw any conclusions on the profitability.


Is the $10 the budget you're using to test a single offer?

Whether or not a bid is high enough, would depend on how competitive the traffic you're targeting is. Which traffic source are you running at? Does it have stats/graph on suggested bids? While those may not be 100% accurate, you can always use them as a starting point. All you need is traffic with not-too-bad quality in the beginning, that you can use to adequately test offers and landers to provide the conversions you need to split-test.

And I don't know where the $25 comes from - you can start cutting placements right away if you like. After you've found a good offer+lander combination, you can always retest some of the placements you've cut before.



Sometimes I regret having "developed" that flow chart at all - as was stated in that lesson, it was meant as a crutch for newbies running direct-linked campaigns for 1/2-click offers.

You're right - you can absolutely make your cutting rules more stringent if you're dealing with a competitive tier 1 geo with lots of traffic volume. As was mentioned above, once you've find a promising-looking offer+lander combo, you can always retest some of the placements. But cutting aggressively in the beginning can potentially help you save a lot of money while you're split-testing offers and landers.


You're partially right. Most newbies have access to the same offers and using ripped landers. But that's only a start.

Don't forget that every experienced affiliate that's doing big numbers, was once a newbie affiliate, with the same limitations as you.

Testing variations of your best landers or even new lander angles can give you an edge. Having a solid testing and optimization strategy that will allow you to test the most amount of "stuff" efficiently on the lowest budget and least amount of time possible will give you an edge as well. Having access to lots of traffic from lots of networks that you've invested time and money into knowing inside out plus built good placement blacklists in multiple geos can be a BIG edge too.

Once you start running volume, doors will start to open: AMs (affiliate managers) will give you cap to the best offers and even exclusive offers, more/higher pay bumps, etc.; traffic source reps will give you access to the best traffic that isn't available to everyone, feed you intel on how best to monetize their traffic, etc. The best affiliates would even run at a slight loss JUST to generate the volume needed to build these relationships.

And when cashflow is no longer as much of an issue, you can work with direct advertisers (skipping the affiliate networks) for higher profits. (Trouble is their payout cycles are typically longer - thus the cashflow requirement.)

Pop is good in that it's easy and cheap to get started with, but that can also be a curse, because it's harder to get an edge.

This is why I'm always advising more-experienced affiliates to expand away from pop. However, I still wrote the newbie tutorial around pop, because it's relatively easy and cheap to get started. And you're practically guaranteed to get conversions with pop even on a low spend, which is what newbies need to see to keep that fire alive - and this isn't always possible when starting out on other traffic types.

You're certainly encouraged to explore other types of traffic though! If you have the tenacity to stick to something until you make it work, then you can jump into any traffic type - just be prepared for a steeper learning curve.

You've got the landers part down which is awesome! Now you just need to get your testing and optimization process down. Every case will be different but the general testing approach I've outlined above can act as an outline to follow. Don't be afraid to post stats, analyze stats, and ask questions - it's the details that will make all the difference.



Amy

---

Home > POP / PPV / Redirect > Follow-along Campaigns