Home > Mobile > Follow-along Campaigns

Mobile sweeps on POP follow along! (44)

11-24-2015 09:27 PM #1 spartanen (Member)
Mobile sweeps on POP follow along!

Intro: Dutch SPARTAN who is convinced this is his future! Have a small recruitment company in IT (I dont know @$%! of IT i just do sales and recruitment) .
The money that i am generating (around 7k monthly (i am going down since i am focussing on AM) i have for AM. Offcourse i also live a little and i have a kid comming etc so i dont want to throw money away! Furthermore i think i saved like 10k.

Skype: dennie.verhagen

Total budget: depends on LIFE! But i think if you read above that it is enough....

Software/hardware:
Beyond VPN Hosting
Voluum

TESTROUND to get to know the process

Traffic source: PopAds
Vertical: Sweeps
Offer: Apple / Samsung voucher
Geos: MY

Budget: depends of a lot of things
Bid: $2.00 CPM

Angle1: are you the winner of a FREE... (translated in MY)
result:
Budget: 20 dollar
impressions: 8695
Clicks: 4610 (F5Media reports 3553 clicks??)
CTR: 53%
Conversions: 2
ROI: -95%
Revenue: -19,04
Date: in the weekend from Saturday on Sunday morning (Amsterdam time)
Frequency cap: 1/3hr

Ok... CTR looks good right? But its not converting very well with the offer... So what to do? Below is something i tried but tips are very welcome!!!

11-26-2015 03:43 PM #2 spartanen (Member)

Can anyone tell me:

What a good CTR is for a POP on Sweeps to descide if it's a good one are bad one?

11-27-2015 07:08 AM #3 vortex (Senior Moderator)

Quote Originally Posted by spartanen View Post
Can anyone tell me:

What a good CTR is for a POP on Sweeps to descide if it's a good one are bad one?
The CR (conversion rate) would be the more important metric here, because that's what's going to make/lose money.

How many landers are you testing? Did you include some ripped landers? It would be good to include some of the ones you see the most often during spying, even if only to establish a benchmark. If you test enough ripped landers and have a great offer, you may not even need to get too creative yourself.

"Are you the winner of a free..." isn't really that great of an angle. It's a great start though and the fact that you're considering angles gives you a big advantage over many other newbies. When it comes to sweeps, you need to give visitors a "reason why" you're giving them a chance to win something. You could be a company asking visitors to fill out a customer-feedback questionnaire in exchange for a chance to win, for example. Or you could draw people in with a game of some sort to get them to start engaging, and then say "wow you've beat our high score so here's your prize!" Any way you do it, try to provide a "reason why" that's believable, and make the visitor feel that they've done something to deserve the prize, and you'll see a difference in your CR.

More importantly, you should consider split-testing offers as well. Try to find similar sweeps offers for your geo. Finding a good offer + lander combo should be a first priority, and something that can only be done by testing.

Looking forward to your next update!


Amy

11-27-2015 09:28 AM #4 spartanen (Member)

Hi Vortex... offcourse the CR is the more important rate but i start from the beginning and i see it as a funnel... Get high CTR first then trying to get high CR....

I think thats abvious right... With high CTR theres a higher possibilitie to get a higher CR... Its like in sales... with a higher input you get a higher output... simple and clear...

SO for now i have tested a few POP Alerts and 1 is def the best: 66%ctr... Do you think i can get this even higher?

After that i am planning on a quistionairy in different styles to see whats the best here... i did direct and got a -50%roi...

11-27-2015 09:42 AM #5 thuglife (Member)

Clicks: 4610 (F5Media reports 3553 clicks??)
Use a CDN. (Amazon's or Rackspace)

Loading time is empirical when doing pops.

Beyond's VPS is located in North America if I'm not mistaken so loading time from MY can be huge = clickloss = low CR.

11-27-2015 01:53 PM #6 spartanen (Member)

Quote Originally Posted by thuglife View Post
Use a CDN. (Amazon's or Rackspace)

Loading time is empirical when doing pops.

Beyond's VPS is located in North America if I'm not mistaken so loading time from MY can be huge = clickloss = low CR.
Thanks! But its also possible that they like my POP alert and not my quistions or something right and went away before going to the offer page right?

11-28-2015 12:12 PM #7 spartanen (Member)

So the thing that i got so high CTR was because i only had a "entry pop" with the text "are you the winner etc" and then straight to the offer... So people dont see anything and get curious but when they get to the offer they bail! (I think)... ROI -50%...

So now i am waiting on bannerlanders.com for 2 new angles with a "entry pop" and quistionairy... In the meantime i lauched 2 new angles but i shortly tested a lot more but these i think has the most potential..

Traffic source: PopAds
Vertical: Sweeps
Offer: Apple / Samsung voucher
Geos: MY

Budget: depends of a lot of things
Bid: $2.00 CPM

Angle10: http://www.spartansdigital.com/Landi...r%20Survey.htm
result:
Budget: 40 dollar or so...
impressions: 1469
Clicks: 15
CTR: 1.02%
Conversions: 2
ROI: -66.84%
Revenue: 0,96
Cost: 2.90
Profit: -1.94
Date: yesterday till now
Frequency cap: 1/3day

Angle11: http://www.spartansdigital.com/Landi...r%20Survey.htm
result:
Budget: 40 dollar or so...
impressions: 1459
Clicks: 13
CTR: 0.89%
Conversions: 2
ROI: - 66.46%
Revenue: 0,96
Cost: 2.86
Profit: - 1.90
Date: yesterday till now
Frequency cap: 1/3day

Quistions:

Total of this campaign till now:
In Voluum it says: 33.227 impressions x 0.002 (my bid) = 66.45
In popads it says: 41.382 impressions x 0.002 (my bid) = 82.76

But i actually payed around 66.45 so how is this possible??

Is this a normal CTR?
Is this a good CR?
Is it a smart move to buy a URL like "freestuff.com to put my LP's on? Now i am going to my own URL Spartansdigital.com.
Hi @thuglife or anyone can you answer the quistion i asked after you commend?
Hi @vortex or anyone can you answer the quistion i asked after you commend?
How do people update the cost in Voluum. I go to the campaign (right mouse button) and select the begin date i started the campaign till now and update all the cost i made together. Is this the good way?

11-28-2015 10:19 PM #8 vortex (Senior Moderator)

Quote Originally Posted by spartanen View Post
I think thats abvious right... With high CTR theres a higher possibilitie to get a higher CR... Its like in sales... with a higher input you get a higher output... simple and clear...

SO for now i have tested a few POP Alerts and 1 is def the best: 66%ctr... Do you think i can get this even higher?
Actually that's not always true.

I see landers that have high CTR but low CR. And then there are landers that have low CTR but high CR.

It all depends on WHAT is causing the high CTR in the first place. Some CTR-boosting tactics will actually lower CR and those are the ones you'd want to avoid.

For example, some landers have high CTR because they make unrealistic promises that attract people to click through, but then because the offer page doesn't fulfill the same promise, a lot of the visitors would bail at that point - causing high CTR but low CR. Another way to get high CTR is by making visitors so curious they just HAVE to click through to find out what it's about, but if your offer isn't appealing to most people, the conversion rate almost certainly won't be high.

At the other end of the spectrum are landers that aim to attract the RIGHT sorts of audience - visitors that are the most likely to sign up to the offer - so that even if there aren't a lot of people clicking through (i.e. lower CTR), a high percentage of the ones that DO click through will sign up (i.e. higher CR).

In short, aiming to increase CTR is a good thing - as long as you make sure to also achieve a higher CR in the process.


Amy

11-28-2015 10:40 PM #9 vortex (Senior Moderator)

Quote Originally Posted by spartanen View Post
So the thing that i got so high CTR was because i only had a "entry pop" with the text "are you the winner etc" and then straight to the offer... So people dont see anything and get curious but when they get to the offer they bail! (I think)... ROI -50%...
I LOL'ed at this! I read this post of yours AFTER I responded to your previous one - we basically pointed out the same thing.

However, an ROI of -50% is not bad at all assuming you haven't tested a lot of landers yet!

Your landers are good, but they're almost the same as far as I can tell. During initial lander testing, it would be good to include some ripped landers - ones you see most often, and make sure they're all very different. Cast a wide net to find out which lander styles work better, then make versions of the best styles to do more testing. By casting a wider net at first, you can achieve larger increases in ROI. If you just take one lander and try to improve on that, your ROI increases will be limited by how effective your lander style is.


In Voluum it says: 33.227 impressions x 0.002 (my bid) = 66.45
In popads it says: 41.382 impressions x 0.002 (my bid) = 82.76

But i actually payed around 66.45 so how is this possible??
Popads often doesn't actually charge you by what you're bidding. In fact, the rate you're charge for each placement is different, and will vary according to competition (look at placement stats in your popads account to see this).


Is this a normal CTR?
Talked about this already...

Is this a good CR?
The answer to this will depend on a lot of things - your vertical, your geo, your traffic source - even specific placements. This is why I focus on improving on CR as much as I can instead of setting a goal to reach. After a few rounds of lander testing you should have a very good idea on how far you can push the CR.

Is it a smart move to buy a URL like "freestuff.com to put my LP's on? Now i am going to my own URL Spartansdigital.com.
Yes! The domain name can affect conversion rates. This has been proven by Ruby Tunes.


How do people update the cost in Voluum. I go to the campaign (right mouse button) and select the begin date i started the campaign till now and update all the cost i made together. Is this the good way?
For popads, there's the [BID] token which you could add to voluum to track costs (if you do that, be sure to re-generate your campaign link and replace this in your campaign on popads). However, there will still be a difference between the actual costs shown on popads, and the costs shown in voluum. For other traffic sources I usually don't track costs in Voluum - I combine cost and revenue either in my head or on a spreadsheet. I know this isn't a very efficient way of doing things, but at least that way I know what the ROI is exactly.


Lastly - is there any particular reason why you're setting frequency to 1/3 hr? I like to set frequency to 1/24 or longer, until my campaign is in profit. Shortening frequency will almost always lower ROI. The advantage would be that you'd get more raw traffic and therefore potentially more conversions, so it would be like sacrificing traffic quality for traffic volume.


Amy

11-29-2015 02:54 AM #10 nebraska311 (Member)

Following this thread - I'm a few steps behind you but hope to be live soon

11-30-2015 01:19 PM #11 spartanen (Member)

Quote Originally Posted by vortex View Post
Your landers are good, but they're almost the same as far as I can tell.
These are ripped landers who i improved a bit to my offer.. furthermore i am waiting on bannerslanders.com for some landers with entry pop and exit pop.

Quote Originally Posted by vortex View Post
In Voluum it says: 33.227 impressions x 0.002 (my bid) = 66.45
In popads it says: 41.382 impressions x 0.002 (my bid) = 82.76

But i actually payed around 66.45 so how is this possible??

"Popads often doesn't actually charge you by what you're bidding. In fact, the rate you're charge for each placement is different, and will vary according to competition (look at placement stats in your popads account to see this)."
Yeah i noticed that popads doesnt actualy charge my on my bid so its also makes it really hard to exactly know my ROI right?
And how is it possible that i got 33.227 impressions in Voluum and 41.382 in popads?

Quote Originally Posted by vortex View Post
"Lastly - is there any particular reason why you're setting frequency to 1/3 hr? I like to set frequency to 1/24 or longer, until my campaign is in profit. Shortening frequency will almost always lower ROI. The advantage would be that you'd get more raw traffic and therefore potentially more conversions, so it would be like sacrificing traffic quality for traffic volume."
I changed this to 1/3days.. i read somewhere people have to see banners or something 7 times before they descide but that also depends on what you are selling so i changed it...

11-30-2015 11:16 PM #12 spartanen (Member)

It's not the worst thing there is but why do i see more conversions in my Affiliate Network report then in Voluum?

And why do i get less impressions if i rotate between landers?

I my ad network / traffic source and Voluum i name my campaign after code they give you from your affiliate network offer. Is this a smart move? Or if people wanna steel (i heard voluum steals for example) i make it a lot more easy?

12-02-2015 05:24 PM #13 vortex (Senior Moderator)

Yeah i noticed that popads doesnt actualy charge my on my bid so its also makes it really hard to exactly know my ROI right?
To get accurate ROI you'll need to export cost per placement stats from popads, export Voluum conversions by placement stats, then combine the two. Here's a tool that can help you with that:

http://stmforum.com/forum/showthread...kly-amp-Easily

It IS a lot of work for sure, so what I usually do is only take the few placements that have had the most traffic, and just manually get the costs from popads stats and conversions from Voluum stats, and calculate the ROI for those. This is also why I try not to cut too many placements while still testing offers and landers - to save on time and to avoid decreasing traffic. Only when I have a good offer and lander will I bother to export stats and combine them and plug everything into the stats spreadsheet to figure out which placements to cut.

If you have more budget than time, check placements less often; if you have more time than budget, check more often.


And how is it possible that i got 33.227 impressions in voluum and 41.382 in popads?
This is known as "clickloss". Usually with pop, 20-30% is acceptable.


It's not the worst thing there is but why do i see more conversions in my Affiliate Network report then in Voluum?
It's probably due to a delay in your affiliate network posting conversions back to voluum. Other than that I can't think of any other reason. If you set up postback incorrectly, you wouldn't be seeing ANY conversions show up in voluum at all.


I my ad network / traffic source and voluum i name my campaign after code they give you from your affiliate network offer. Is this a smart move? Or if people wanna steel (i heard voluum steals for example) i make it a lot more easy?
If that's a concern for you, then just change your naming convention. Do anything that would give yourself peace of mind. Most people underestimate the power of expectation (positive/negative). If something bugs you, it would be good to either find reasons to put your mind at ease, or do something that would make you feel like you've helped the problem.


Amy

12-04-2015 10:39 AM #14 exclusif (Member)

Quote Originally Posted by thuglife View Post
Use a CDN. (Amazon's or Rackspace)

Loading time is empirical when doing pops.
I'm been thinking about this, and the importance of load time should depend on whether you're running pop-ups or pop-unders. On pop-ups load time is of course crucial, but on pop-unders it seems rather unimportant since the LP per definition is shown under what you're currently viewing, and thus you wont notice it right away anyway. Am I wrong here?

12-04-2015 04:29 PM #15 vortex (Senior Moderator)

Quote Originally Posted by exclusif View Post
I'm been thinking about this, and the importance of load time should depend on whether you're running pop-ups or pop-unders. On pop-ups load time is of course crucial, but on pop-unders it seems rather unimportant since the LP per definition is shown under what you're currently viewing, and thus you wont notice it right away anyway. Am I wrong here?
Actually, if you put something like an entry java alert pop up on your lander, you can force the visitor to "jump" to your lander right away without waiting for it to be discovered. Quite a few pop sources still allow entry alerts so if you're not using it on your landers currently, I urge you to include that in your testing very soon!

Now couple that with the use of a CDN and other techniques that will allow you to reduce page load times and you will likely see a difference in CTR and/or CR.


Amy

12-04-2015 06:18 PM #16 exclusif (Member)

Quote Originally Posted by vortex View Post
Actually, if you put something like an entry java alert pop up on your lander, you can force the visitor to "jump" to your lander right away without waiting for it to be discovered. Quite a few pop sources still allow entry alerts so if you're not using it on your landers currently, I urge you to include that in your testing very soon!

Now couple that with the use of a CDN and other techniques that will allow you to reduce page load times and you will likely see a difference in CTR and/or CR.


Amy
I use beyond hostings own cdn for images and heavier scripts. Currently have a 0.3s load time from New York. Will probably get a couple of dedicated servers at some point, when Im profitable

I tested an entry pop at one point and it actually reduced cr significantly. Maybe its worth another shot. It seems pop-unders are regarded as better overall, so is it smart to essentially turn them into pop-ups with an entry pop? Either way it seems better to wait with loading the script until the page is fully loaded to avoid the load time factor. I use exit pops btw which I really believe in.

12-06-2015 04:31 AM #17 vortex (Senior Moderator)

Quote Originally Posted by exclusif View Post
I use beyond hostings own cdn for images and heavier scripts. Currently have a 0.3s load time from New York. Will probably get a couple of dedicated servers at some point, when Im profitable

I tested an entry pop at one point and it actually reduced cr significantly. Maybe its worth another shot. It seems pop-unders are regarded as better overall, so is it smart to essentially turn them into pop-ups with an entry pop? Either way it seems better to wait with loading the script until the page is fully loaded to avoid the load time factor. I use exit pops btw which I really believe in.
Sounds great! Like you, I believe in exit pops as well. I like to tell the visitor what they need to do on the next page (i.e. offer page). Can do wonders to cr.

Amy

12-06-2015 09:29 AM #18 spartanen (Member)

Few quistions:

From where is a click registred? From the moment they press my entry pop alert or from the moment that they go to the offer page?

I was reading something about cpm media buying and ripping LP's and got stuck on the following:

"The best affiliates are constantly rotating in new ads. The original guy took all the best impressions, and now you’re left with shitty leftover impressions. You might be using his ads from 2 weeks ago and competing against his new superior ads."

Do they mean with leftover impressions that these are people who already saw the ads or if you bid to low you get less unique visitors? I understand that you want the best impressions if people offering the same offer on for example sunglasses... but with this "win a iphone" kinda offers isnt it all has to do with get a bit of a unique LP designed by yourself what nobody else has because the offers are already everywhere?

And what is a good impressions / conversion rate for mobile POP Sweepstakes?

12-08-2015 11:45 AM #19 spartanen (Member)

@vortex can you help me out with the above?

12-08-2015 12:19 PM #20 whtang (Member)

Im not Vortex but let me try to answer your questions:

A click will be registered when a visitor click on the button that redirects the visitor to the offer page.

There is an excellent 3 part guide, written by Caurman, with the title "Product your NUTS". Read those guides and you will understand what its mean.

"What is a good impressions / conversion rate for mobile POP sweepstakes?"
Let the data speak. try to get statistical significance

12-08-2015 04:58 PM #21 Fabio (Member)

Are you still only testing with one lander?

I have been wanting to try popads, what has been your experience working with the platform?

12-09-2015 10:41 PM #22 spartanen (Member)

Quote Originally Posted by noxxyy View Post
Are you still only testing with one lander?

I have been wanting to try popads, what has been your experience working with the platform?
i did a lot of testing... 21 angles or so...
Popads works fine for me...

12-09-2015 10:43 PM #23 spartanen (Member)

Quote Originally Posted by whtang View Post
"What is a good impressions / conversion rate for mobile POP sweepstakes?"
Let the data speak. try to get statistical significance
Thanks for the before answer... and for this... i just lowered my bid and i am in green... so everybody can go to green it looks like... just lower your bid... but i wanna know if i am with my LP on the right track thats why i wanna know what a normal / good impression / conversion rate is for mobile POP sweeps?

12-09-2015 10:44 PM #24 spartanen (Member)

BTW... yesterday i got not so much impressions comparing to today but yesterday i got the same amount conversions as today... let me show you...

yesterday: 7353 impressions / 24 conversions
today: 6622 impressions / 12 conversions and the day has not ended yet

is there something that i can do to upgrade impressions / conversions? or should i just shut down a day like this? maybe throttling ?

12-10-2015 12:24 PM #25 spartanen (Member)

can anyone anser me please?

12-10-2015 02:08 PM #26 refresh (Member)

you're acquiring data spartanen. Let the machines do the work.
Have you achieved statistical significance yet? If not, do not shut it down. Don't focus on keeping it green right now, if you do this you'll kill your volume, and you'll be limiting your potential upside.

12-10-2015 08:14 PM #27 spartanen (Member)

Thanks @refresh... And i will do that but still i dont understand why nobody can say what a normal impression / conversion rate is for mobile sweeps on pop???

How do i know if i achieved statistical significance? because i think so but i dont know what i means exactly...

12-11-2015 03:53 AM #28 caurmen (Administrator)

Check out the Getting Started Guide - there is an article in that which goes through statistical significance and how to figure out if you're there yet.

12-11-2015 06:22 AM #29 sushiparlour (Member)

My 50 cent: if you're comparing different days of the week there is going to be a difference so let the stats speak for themselves before worrying (patience is also not my strong suit but am learning)

12-11-2015 11:48 AM #30 spartanen (Member)

@sushiparlour... hours etc are i think super important but you need a lot of data for that... not there yet.. but for example yesterday i got no conversion all day and in 2 hours i got 10 conversions... today i woke up a hour ago and i got 31 conversions with 8000 impressions... so totaly different with yesterday...

12-22-2015 11:01 AM #31 caurmen (Administrator)

Hey all! Sorry I didn't see this thread sooner!

OK: prerender.

First thing to realise is that there are two different approaches that you can use to prefetch/prerender. The one I showed at AWA is more powerful but works on less browsers - it's also simpler - but it does cause problems with click counts as you've seen.

That approach is prerender. It's been covered pretty well in this thread already: just use the line

Code:
<link rel="prerender" href="http://xxxx.voluumtrk.com/click">
in your header. This should be your tracker link, not your offer page - the prerender code follows all the redirects in the background and renders the offer page so that it appears literally instantly.

That does mean it will sometimes register a clickthrough.

Prerender only works on Chrome, Chrome Mobile, and IE11 at present although more browsers are adding it, I believe. It's very powerful but limited.

For a more general solution that works with almost all modern browsers and doesn't mess up your stats, but doesn't give quite the page speed boost that prerender does, you can use

Code:
<link rel="preconnect" href="http://xxxx.voluumtrk.com/">
and

Code:
<link rel="prefetch" href="http://xxxx.offerpage.com/sodding_great_png_background.png">
Both of those should be added in your <head> just like the prerender code.

Preconnect connects the browser to the domain name you specify. It handles domain lookup (around 300ms per domain usually) and negotiating connection (100ms ish), both of which are part of the reason that pages load slowly. You should Preconnect to ALL domains that your visitor will go through on the way to the offer - your tracker, your aff network, and your offer url.

(If you think about it, on domain lookup alone that's about a second! No wonder offers load slowly.)

Prefetch lets you fetch large assets that will be needed on upcoming pages - images, js, css - and cache it so that it appears to load instantly. This takes a bit of work to get to do a good job - you'll need to go to your offer page, run a waterfall test on something like Pingdom to figure out what assets the page uses that load slowly, and then individually set a prefetch line for each one of them.

It's a pain but it will speed up a lot of offer pages a great deal.

Hope that helps! I'll probably write a full tutorial on this since it has been so popular some time in the new year. Let me know if you have more questions!

12-22-2015 11:07 AM #32 kenny_digital (Member)

Quote Originally Posted by caurmen View Post
Hey all! Sorry I didn't see this thread sooner!

OK: prerender.

First thing to realise is that there are two different approaches that you can use to prefetch/prerender. The one I showed at AWA is more powerful but works on less browsers - it's also simpler - but it does cause problems with click counts as you've seen.

That approach is prerender. It's been covered pretty well in this thread already: just use the line

Code:
<link rel="prerender" href="http://xxxx.voluumtrk.com/click">
in your header. This should be your tracker link, not your offer page - the prerender code follows all the redirects in the background and renders the offer page so that it appears literally instantly.

That does mean it will sometimes register a clickthrough.

Prerender only works on Chrome, Chrome Mobile, and IE11 at present although more browsers are adding it, I believe. It's very powerful but limited.

For a more general solution that works with almost all modern browsers and doesn't mess up your stats, but doesn't give quite the page speed boost that prerender does, you can use

Code:
<link rel="preconnect" href="http://xxxx.voluumtrk.com/">
and

Code:
<link rel="prefetch" href="http://xxxx.offerpage.com/sodding_great_png_background.png">
Both of those should be added in your <head> just like the prerender code.

Preconnect connects the browser to the domain name you specify. It handles domain lookup (around 300ms per domain usually) and negotiating connection (100ms ish), both of which are part of the reason that pages load slowly. You should Preconnect to ALL domains that your visitor will go through on the way to the offer - your tracker, your aff network, and your offer url.

(If you think about it, on domain lookup alone that's about a second! No wonder offers load slowly.)

Prefetch lets you fetch large assets that will be needed on upcoming pages - images, js, css - and cache it so that it appears to load instantly. This takes a bit of work to get to do a good job - you'll need to go to your offer page, run a waterfall test on something like Pingdom to figure out what assets the page uses that load slowly, and then individually set a prefetch line for each one of them.

It's a pain but it will speed up a lot of offer pages a great deal.

Hope that helps! I'll probably write a full tutorial on this since it has been so popular some time in the new year. Let me know if you have more questions!
Thanks Caurmen for your response on this!

I think the pre-render is defo the coolest but with the CTR sku'ed I dont know which lander is doing the best! But I guess since finding out about the Probablility calculator (shout out to vortex for that one!) I guess CTR doesnt matter as much anymore?

Thanks K

12-22-2015 11:41 AM #33 spartanen (Member)

Thanks so much @caurmen! So we can also put in all 3 codes?

so that if someone uses the browsers that prerender can handle it uses the prerender code and if not it uses the other ones etc?

Also i noticed "fucked clicks" i dont care about the skewed profile anymore... some LP's have lot of clicks and almost non conversions and some has almost non clicks but when someone clicks 25% makes a conversion also! I was stuborn as fuck and @vortex was offcourse right! CONVERSION is the only thing i will watch! Testing time has arrived after my initial MY offer (learned a lot)... from now on i am going to test EVERYTHING...

3 or 4 kind of sweeps (and comparing sweeps from different kinda AF networks)... 3 or 4 kind of LP's... all kind of angles... 3 or 4 kind of traffic sources.. 3 or 4 kind of countries!! TIPS are welcome for this @vortex @caurmen! Is this a smart move?

Offers are also welcome!

12-22-2015 11:51 AM #34 spartanen (Member)

Can anyone tell me more about this?

What no one tells you when they say you really gotta know the traffic source before you jump around means, is that every traffic source responds differently to change, for example if you change the bid on 1 source, it might kill your campaign because it restarts you in the chain of bidders; whereas on the other traffic source, if you remove a targeting option it might kill your campaign, or if you decrease and increase budgets - so yea before you go crazy scaling horizontal, put in $500-1000 and focus on getting to know that 1 traffic source.

12-22-2015 05:29 PM #35 spartanen (Member)

And... like you know i am doing a offer in MY at the moment...

Somebody is bidding 0.02 cpm... (and he is 5th in the row...) 20 dollar for 1000 impressions... Sorry but how is it possible that for examle me with a payout of 0.48 can compete with that? That means that i at least have to convert 41,66666667 times in 1000 impressions for breakeven. So every 24 impressions a conversion... Even with a payout bump etc this seems impossible!?

Otherway around it seems also impossible if you have a payout of 20 dollar that that converts every 1000 impressions right?

12-24-2015 01:55 AM #36 Gabreeze (Member)

That might be one of his strategy
Caurmen has great post about it
http://stmforum.com/forum/showthread...he-Competition

12-24-2015 02:54 PM #37 vortex (Senior Moderator)

Quote Originally Posted by kenny_digital View Post
I think the pre-render is defo the coolest but with the CTR sku'ed I dont know which lander is doing the best! But I guess since finding out about the Probablility calculator (shout out to vortex for that one!) I guess CTR doesnt matter as much anymore?

Thanks K
Haha that calculator wasn't introduced to this forum by me - it was caurmen - years before I even knew what STM was lol!

But yes, when deciding what to cut, CTR does not matter in any way. CR is the metric that determines ROI. You want to cut something because it's not making money, no matter how much interest it may generate (i.e. no matter how high CTR may be).


Amy

12-24-2015 03:22 PM #38 vortex (Senior Moderator)

Quote Originally Posted by spartanen View Post

Also i noticed "fucked clicks" i dont care about the skewed profile anymore... some LP's have lot of clicks and almost non conversions and some has almost non clicks but when someone clicks 25% makes a conversion also! I was stuborn as fuck and @vortex was offcourse right! CONVERSION is the only thing i will watch! Testing time has arrived after my initial MY offer (learned a lot)... from now on i am going to test EVERYTHING...

3 or 4 kind of sweeps (and comparing sweeps from different kinda AF networks)... 3 or 4 kind of LP's... all kind of angles... 3 or 4 kind of traffic sources.. 3 or 4 kind of countries!! TIPS are welcome for this @vortex @caurmen! Is this a smart move?

Offers are also welcome!

LOL No need to be hard on yourself! Actually you weren't wrong. CTR is important too - just not when you're deciding what to cut.

CR is the important metric when deciding what to cut. However, CTR is important when you're designing landers - you'd want to figure out how to make CTR higher without negatively affecting CR. For example, if for a specific lander, 1 in every 10 people that click through to the offer converts. Now if you can modify that lander in such a way that you double its CTR, but STILL have 1 in every 10 people that click through to the offer convert, then you will have doubled your ROI.

Regarding your test plan - if you have a big test budget I'd say go for it - test as extensively as you can to uncover the gold and then zoom in and focus on those areas. However, if you want to save on test budget I would recommend doing more spying and asking around to narrow down the number of geos. Also, you wouldn't want to be doing full testing on 4 traffic sources all at once - that would be a bit of an overkill. A more wallet-friendly way may be to rip a few popular landers (or use your current best lander if you have one), take an offer that you KNOW lots of people are doing well with (your AM should be able to tell you this), set up a low-budget camp on each of the 4 traffic sources as an initial test. Then you pick the traffic source that seems to be giving you the most bang for your buck and do all your offer testing and lander testing there, and then scale to the rest of the sources after you find a promising offer + lander combo.

To me, the most important campaign factors are the offer, the angle, and the traffic. Spy to make sure the type of offer you're running is doing well for the traffic type you're planning to buy - if there are other people already promoting the same offers on that traffic type then you know someone's probably making money. Beyond that, it will mostly be about testing angles (i.e. banners and landers) and offers until you find a good combo. Other than that, stuff like playing with bids, cutting placements / devices / OSs / browsers / etc. will help push a camp further into green still.

Reminds me of how some machinery have bigger dials to make big changes, and then smaller dials for fine-tuning. In my mind, the traffic type, offer and angle are the big dials. Other tweaks are the smaller dials.

Best of luck with your testing efforts! It's a pleasure to be following along your follow-along!


Amy

12-24-2015 03:44 PM #39 vortex (Senior Moderator)

Quote Originally Posted by spartanen View Post
Can anyone tell me more about this?

What no one tells you when they say you really gotta know the traffic source before you jump around means, is that every traffic source responds differently to change, for example if you change the bid on 1 source, it might kill your campaign because it restarts you in the chain of bidders; whereas on the other traffic source, if you remove a targeting option it might kill your campaign, or if you decrease and increase budgets - so yea before you go crazy scaling horizontal, put in $500-1000 and focus on getting to know that 1 traffic source.
What attila means is that every traffic source has little quirks that you could only find out through experience. On some traffic sources when you change one setting the whole campaign just won't perform anymore.

Other things you can test to really get to know a traffic source are:

1)When you vary your bid, how much difference are you seeing in conversion rates? (This is not just traffic source specific, but also geo-specific.) Are you getting traffic from better placements when you bid higher/lower?

2)When you set up several camps with the same targeting, does it give you extra traffic? Or are you just getting the same traffic overall with traffic volume split equally between the 2 camps?

3)Are you actually charged the amount you bid? Or higher/lower? Is what you're actually charged relative to what other competitors are bidding?

4)When you're the top bidder how much traffic do you get? What about when you're 2nd? 3rd? 4th? (On some traffic sources if you're not the top 1-3 bidders you get little to no traffic at all.)

5)Does the source have a lot of bad placements? Does it have a lot of good placements too? (For example, some sources have a lot of bad placements, but if you spend money to identify and block them you can get to the good placements.)

6)What are ALL the different ad / targeting / settings options available? Have you tested all of them? For example different frequencies, different banner sizes, popunders vs. popups vs. tabups vs. tabunders, pops vs. redirects....etc. etc. etc.


The list could go on, but you get the idea.


Amy

12-24-2015 03:57 PM #40 vortex (Senior Moderator)

Quote Originally Posted by spartanen View Post
And... like you know i am doing a offer in MY at the moment...

Somebody is bidding 0.02 cpm... (and he is 5th in the row...) 20 dollar for 1000 impressions... Sorry but how is it possible that for examle me with a payout of 0.48 can compete with that? That means that i at least have to convert 41,66666667 times in 1000 impressions for breakeven. So every 24 impressions a conversion... Even with a payout bump etc this seems impossible!?

Otherway around it seems also impossible if you have a payout of 20 dollar that that converts every 1000 impressions right?
Several thoughts here...

-Just because the person's bidding that high, doesn't necessarily means he's actually being charged that much. It depends on the traffic source.

-He may be promoting an offer that has a higher margin (e.g. higher payout due to doing volume, or running direct with the advertiser, or maybe he IS the advertiser that owns the product). He may have already identified the best placements and targeting which allows him to bid higher. Spying would be helpful in this case - this is why whenever I see people running tech support offers in a certain geo on a certain traffic source, I stay away, because I know it would be near impossible to bid as high as them unless I pushed the same type of offer.

-Small payout offers can convert very well if you test and find the right ones! An offer is priced the way it is for a reason. I've seen offers that pay peanuts that convert like gangbusters.

At the end of the day, only testing will tell you whether something will work. A safer bet would be to spy on your traffic source - if you see other people pushing your offer (or the same type of offer with similar payout) it would be a good sign. Also, if you have an experienced AM, they should be able to tell you which traffic types work well for your offer. In that case just do a small test to see what happens. For an offer with such a small payout, it won't take a lot of money to see whether it has any potential.


Amy

12-24-2015 08:03 PM #41 caurmen (Administrator)

@spartanen - yes, you should be able to use all three codes. Put the pre render one first.

01-06-2016 03:17 AM #42 simon_89 (Member)

Quote Originally Posted by caurmen View Post
Code:
<link rel="preconnect" href="http://xxxx.voluumtrk.com/">
Preconnect connects the browser to the domain name you specify. It handles domain lookup (around 300ms per domain usually) and negotiating connection (100ms ish), both of which are part of the reason that pages load slowly. You should Preconnect to ALL domains that your visitor will go through on the way to the offer - your tracker, your aff network, and your offer url.
For Preconnect do you just use your campaign link that is created in Voluum(http://fffff.voluumtrk.com/fffffff-f...476?websiteid=[WEBSITEID]&keyword=[KEYWORD]&categoryid=[CATEGORYID]&country=[COUNTRY]&ISPID=[ISPID]&IMPRESSIONID=[IMPRESSIONID]&OSNAME=[OSNAME]&BROWSERNAME=[BROWSERNAME]&DEVICENAME=[DEVICENAME]&BID=[BID]"
or would you still use the Voluum postback link: "http://xxxx.voluumtrk.com/click" like the prerender does?

01-10-2016 11:39 AM #43 spartanen (Member)

Peeps i am changing to the hardest vertical there is... yeah i know it sounds stupid but i am gonna do REV SHARE with a SUPER AFFILIATE on FB/Nutra... So good luck everybody! And wish me luck to!

01-11-2016 11:59 AM #44 coconut (Member)

Good luck bud! let me know how it goes!

Home > Mobile > Follow-along Campaigns