Home > Mobile > Follow-along Campaigns

---

From Negative ROI -> Break Even -> $XXX Profit (53)

---

Hey All,

BACKGROUND:
I've been trying to break into Mobile for so long trying to form a system. I've spent thousands of dollars "testing", but really looking at it, I was just throwing shi*t at the wall hoping something would stick.

The closest I've ever been to breaking even was -20% and then I hit a wall. Thinking I knew how to cut placements, I started cutting and traffic slowly died.

Knowing my sticking points, I read up on the forum and I'm starting to realize I was doing a lot of things wrong - that -20% was probably a fluke lol

Hopefully, through this follow along I'll get help from minds more experienced than mine.

GOAL:
- Develop a system that will get me from negative ROI to a steady profit
- Win the STM award for best follow-along


CAMPAIGN:
Traffic Source: AdCash
Offer: Sweeps - Win an iPhone
Geo: Australia

PROGRESS SO FAR:
LANDERS:
- 4 in total - Ripped 2 and created 2 of my variations.
- Using a CDN to host the landers

CAMPAIGN SETUP:
- Budget for this test - $80
- Created 1 Main Testing Campaign in Voluum and in AdCash

Hopefully this will get enough data to determine different segments

I've started the campaign and will update you on the stats tomorrow!

QUESTIONS:
1. How's my campaign setup?

I've seen people break out their campaigns into WIFI/CARR or PHONE/TABLET. I was hoping using my main campaign at first, and then drill down in Voluum those type of stats.

---

Great start Glen! I'll be rooting for ya.

I agree with you that the lander CTR is horrible - but have you noticed that of the people that DID click through, a really high percentage of them went on to convert? This suggests that your offers are solid, and you just need better landers.

It's great that you were able to cut a lander and an offer as well.

Judging by the low CTR of your current landers, I wouldn't advise testing lander variations at this stage. Here's a job for you: Go rip at least 5 more landers that preferably look very different from your current ones. Wait until you can cut 1-2 more of your current landers, then start another around of lander testing by adding-in the new landers.

Adding an entry pop is a good thing to test yes - as are the backbutton redirect, vibrate, audio etc.

Also, do check your landers' loading speeds. Optimize your landers for fast loading. Caurmen's thread here can help you:

http://stmforum.com/forum/showthread...s-Loading-Time

Also, gtmetrix.com will analyze your page and provide good optimization suggestions.

With regards to screwing up test data - once you add more candidates to a test, any data previously gathered needs to be excluded from evaluation. So if you add landers to your campaign, when you're entering stats into the split-test calculator you can't include data gathered before for the old landers, because conversion rate will fluctuate from day to day.


You can do this for sure, but be aware that you'll be testing landers and offers for 2 campaigns instead of one, which may or may not be worth the trouble. Are you seeing different landers/offers perform better for each device? Or are you wanting to assign a different bid to each device? If I were you I'd wait until I have more data though before making this decision.


Why not send a support ticket to Adcash?


Summary: Focus on improving your landers!


Amy

---

Thanks Amy! I appreciate the feedback.

I'll start working on ripping new landers and include them on a new test.

Today's results are similar to yesterday.

I included the variations but there wasn't enough actionable data to do anything about it. But after reading Amy's advice, I'll do that first before I run another test.

On the bright side, although the landers were weak, I've narrowed the offers to 1!

THINGS TO DO:
- Spy and rip additional landers for testing
- Run them through an optimization checklist to ensure faster load times

---

GREAT work on data analysis! So here's what you're found, and suggestions on what you can check/do:


-Wifi traffic converts like sh*t - only 1 conversion after $50's worth of traffic, and in spite of testing multiple offers and landers. Basically would take a ton of work to make it profitable, so can write it off at least for now.

-Carrier traffic is in profit and with unreal ROI before you've even optimized landers! Have you drilled down to specific mobile carriers stats? At any rate, I would agree with you that you should set up a separate camp for this right away - to target just carrier traffic. You can include all OSs and not cut tablets for now, because like you've said, they may still turn out to be profitable.

In fact, try setting up multiple camps at staggered bids. With such a high positive ROI, you should definitely try higher bids. If this was my camp I would set up 5 camps at 5 bids to get a better idea of the bidding landscape. I would suggest at least 3 camps if you don't want to bother with 5. Remember to set a low budget on each ($20/camp should be enough for a rough idea).

-If you're feeling adventurous: Start setting up test camps on other traffic sources as well. If you want to play REALLY safe, just target carrier traffic for IOS. Also, only target carriers that are doing the best based on your adcash stats. Re-test all landers since you don't know which ones are best yet. You probably won't get a ton of traffic with this targeting, so stick to big sources that have a lot of AU traffic: Zeropark and Propeller come to mind.

-Lastly - I literally cringe when I see your lander CTR which is horrible across the board! But you're already using a CDN and have optimized them for speed. But your ROI is still so high - imagine if you could identify the problem with your lander CTR and fix that, how much higher the ROI can go!

If you could send me your lander in a PM - either a url, or just the lander code - I'll try to find some time to set up a small camp to see whether my CTR is similar. That way at least you know there's not a major problem with your setup.

You're making amazing progress - congrats and keep it up!


Amy

---

Will create new campaigns and stagger the bid. Right now I'm using the recommended bid of 3.7 - how should I stagger them?
Recommended - 3.7
Medium - 10
High - 15
Highest - 20

Also, i check pingdom for my speed test.

For 1 of my landers it's initial load is about 1sec. I test it again right after and it drops to about 200ms. Do I optimize it even further to drop the initial load?

Also, I used a prefetch and preconnect on my landers - I don't know if I'm using it right.

Preconnect
I create individual <link rel="preconnect"> entries for all the domains the user goes through after my lander - I used HTTP live headers to see the click flow.

Is this right?

Prefetch
I create individual <link rel="prefetch"> entries for all the larger files that's on the offer lander

Is this right?

Thanks again for you help!

---

Taking Amy's suggestions, I separated the CARRIER only campaign into 3 bids.

Stats from yesterday's campaign:
SPEND: $76.33
CONV: 7
REV: $21.70

CAMPAIGN OVERVIEW:


HIGH BID: $12
MED BID: $7
RECOMMENDED: $3.7

QUESTION: are the bids different enough, or should I make them higher from eachother?

NOTES:
- Landing CTR is still pretty horrendous - question from my previous post:

- Amy I'll PM one of my landers to you, hope that's ok.
- High bid and medium bid are getting roughly the same amount of traffic - do I keep running both and improve the funnel (since 3k isn't a lot of traffic)?
- What's a reasonable CTR for Pop traffic?

HIGH BID - LANDER STATS

Eventhough A1v5b and C25.A4v1 got conversions - I don't think I've reached statistical significance on the landers to start cutting, but do you think I could cut c25.A2V1 and C25.A1V1 since they didn't even get a click through?


MED BID - LANDER STATS

I don't think I've reached statistical significance to start cutting - will run for 1 more day.

RECOMMENDED BID
- I don't think I should continue this camp since it's not generating much traffic - correct me if I'm wrong



THINGS TO DO BEFORE NEXT TEST:
Work on my landers - really work on my landers

---

Hey guys,

Here are my stats from yesterday:

CAMPAIGN STATS



I'm re-evaluating my campaign and moving my traffic source to PopAds.

REASONS:

• I've used PopAds before and comfortable with it.
• Able to set minimum bids of $20 for each campaign whereas AdCash is $50. I would have to constantly watch the test campaigns until it hits $20 to pause.
• The traffic estimator will help provide me with a visual representation of the types of bids and traffic available.
• They have an assortment of placements (Popups, Popunders, Tabunders, etc.) for me to find a profitable campaign and scale it before going to newer sources
• I just feel it's more noobie-friendly
• I'll also be able to start fresh
  
  • Stagger different bids from the beginning of the campaign
  • Tweak my landers even further for speed
  • Restructure my campaigns, flows, and landers to make more sense - I felt unorganized

NEW POPADS CAMPAIGN

• I will still be using the same offer and including 2 new offers that were just provided to me by my AMs (said it's "BLOWIN UP")
• Will rotate 5 landers from the start
• Start with 3 Campaigns - Low Bid/Med Bid/High Bid
• Improve sales funnel - include YTZ and Monetize for remnant traffic

I have a lot of work to do for the weekend. Hopefully provide an update by Sunday with some fresh stats

---

There's no right or wrong way to stagger bids. But I'm very surprised at the low ROI. Are you targeting carrier traffic alone? If not you should!

I'd suggest to put some decimals behind your bids though to "further stagger" from your competition who may be bidding in whole numbers.

I like to bid for example x.222.


200ms is quite good.


http://stmforum.com/forum/showthread...l=1#post254366

Not being a technical person myself, you'd do much better to refer to caurmen's original post on the topic.


Haven't received your PM yet - will be waiting!


Leaving one to run is fine. You can do your testing there.


Depends on too many variables, e.g. geo, lander type, vertical, traffic quality, bots vs. humans. Too bad I haven't run sweeps in AU in so long, I've lost my data for that already. But I can tell you that I don't remember ever seeing a CTR < 5%. Some landers in certain verticals and geos have reached 70-80% with the backbutton and exit script and all that good stuff.


Right now you can't really cut landers based on lack of clicks, because they ALL have such low CTRs. For example, your lander that got 2 clicks and both resulted in leads - those landers that currently have 0 clicks may just as well catch up in a hurry if they get a couple of clicks. The difference isn't nearly great enough.


That wouldn't be a bad decision. Although that lander problem is yet to be solved. Your plans and set up sounds great. Will be waiting on your stats.

Oh! I just remembered - there's another member who's also running AU sweeps on popads - you can see his stats for sample lander CTRs:

http://stmforum.com/forum/showthread...l=1#post264538

So yeah - we REALLY need to solve your lander problem ASAP! That offer is converting extremely well! I suck at coding, but maybe I could at least run your lander and let you know if I'm getting the same CTR.



Amy

---

I don't know what could be wrong but i've also had a hard time with CTRs. I can literally run such a click bait lander and still be under 1% ctr it's weird... I don't know if it's my Voluum setup or what, but if that's the case why do I get any clicks at all? I just tried this page https://s04.adplexity.com/storage/im...410cd5091e.png (not exact) just to see how it performed and it only got like a .30% ctr lol in AU popads

also is < 5% with scripts or just pure clicks through cta

---

Have you verified how the lander looks from your smartphone? Try to show all the important elements above the fold.

Also - have you optimized the lander for loading speed? That's crucial as well. Speaking of which - are you using a CDN?

0.30% CTR is horrendous for sure. Very weird that a few of you are experiencing the same problem for the same geo.



Amy

---

Amy,

Thank you so much for your help and running that test!

I'm currently optimizing all my landers to make sure I can achieve the fastest possible response times.

Although you used a different TS, my experience with landers was still under 1%, regardless of the traffic source.

The new insight you provided about my CDN is shocking though. I'm using rackspace cloud files - which to my understanding was sufficient. Maybe I should try Amazon Cloudfront. This is very interesting though.

And to think with my previous campaigns I've run the past, I would come close to breaking even with crappy landers.. imagine the ctr boost... I'm drooling at the possibilities Lol

---

Yeah I've optimized for speed, lander looks fine on my phone and browserstack, but this isn't just AU I get pretty bad ctr in other geos too, using a CDN though I'm not sure if i set it up right, i mean it works but im not sure if it works like it should be.

I've been operating under the assumption that I just suck, or I'm too compliant but I don't have the money to test a campaign in the grey area and end up not being paid, my budget not so large

Do you think amazon doesn't send my files to Australia servers right away because low traffic? Maybe after running for longer load speeds get better since it's been distributed to more servers?

Hey glennb would you send me the same lander you sent Vortex? I'd like to run the same test I promise I won't user it after the initial $3 spend. (want to test my own setup.)


EDIT: Okay so I tested a landing page I tested a few days ago at $10CPM AUS on PopAds NON-ADULT traffic. This landing page got 0.67% ctr when I tested it the first time, with Amy's settings it's currently at 16% ctr... Wtf? Bots or what? Why does a lower bid have a higher ctr? I bet i don't get a single conversion from this too.

EDIT: I got a conversion now I'm in the green... Why was I bidding so high before??

---

I'm a bit technically-challenged, but as far as I understand, that's not how it works. If you're running in a single geo, I think the first time the lander is accessed it gets cached at the datacenter that's closest to the visitor, and then all the visitors that view the lander after that will see the cached version which makes loading time fast.

So it shouldn't be a matter of running longer.


Holy crap! I don't usually pay attention to CTR so have never noticed the difference in CTR at different bids. What I DO know is that when you bid lower you'll usually get more bot traffic. This is because people will naturally blacklist placements that don't convert, so competition for those placements will be low, which is the reason why you're able to get traffic from them at the low bid price. This is just a generalization of course.


This is exactly why I always encourage people to test several different bids. However, one conversion doesn't tell you much. Hope this keeps up for you!



Amy

---

Ok, I'm still getting horrendous CTR:



What I did:
1. Optimized all landers - some being around 150ms
2. change my targetting to what Amy suggested
3. Changed NS to Amazon Route 53
4. Moved CDN from Cloud Files to Rackspace CDN - i think that's essentially the same

This is really bugging me.

What I'm going to do now is switch CDN to Cloudfront Completely and hosting to S3 - this should be a similar setup to Amy and steel's.

On that screenshot I spend $20 and made $12 so I'm not losing much.

---

This is actually VERY good news! If your popads settings are the same and STILL getting such low CTRs that means it's something that can be fixed!

I'm also using NSOne for nameservers. Have no idea how much of a difference this is making. Which nameservers are you using right now?

So we know your CTR is like 1/10th of what I got. This implies that if you can fix this (and you WILL), instead of making $12 for every $20 - you could be making $120 for every $20!!! Holy batman!

Let's get this fixed before doing ANYTHING else!



Amy

---

Ok, still not getting anywhere with this lander CTR thing.

Here are my 2 setups:

1. Normal Setup
Nameserver: Route 53
CDN: Rackspace CDN
Hosting Server: WiredTree (Linux Server)

2. Amazon Setup
Nameserver: Route 53
CDN: Cloudfront
Hosting: S3

I ran both setups head to head and they're both receiving brutal CTR's under 0.50%.

As always, there are always some takeaways from this test:
- The Amazon Setup produced a slightly higher CTR and produced more conversions than my normal setup - moving forward I'll just use the Amazon Setup
- I did stagger the 3 bids and my medium and low produced more conversions than my high bid - moving forward I'll just focus on my medium and low
- I included YTZ in the mix for remnant traffic and it's showing some conversions so I know that funnel is working - plan on including Monetizer

TO-DO
- Apply for a trial of DNSMadeEasy
- Do another round of testing

QUESTIONS
- Is there anything I'm missing?
- Could I try anything else?


IM DETERMINED TO FIX THIS

---

Thinking of trying out DynDNS and DNSMadeEasy since they offer a free trial.

---

I didnt read the entire thread as its pretty long, but I see you're having a problem with low CTR. I see you are using rackspace CDN, Im using that too and didnt notice any problems. DNSmadeeasy wont make a miracle either, when I started to use it, I didnt really notice any difference. Are you sure you're not buying bots? Some bots are not able to simulate the click through an LP, thats why the numbers MIGHT be so low.

---

Hey Matuloo,

I sent Amy one of my landers to test. She made a generic campaign and received a CTR of 4%. I created the same campaign on the same TS and I was receiving CTR's of under 1%.

Steel520 was also receiving horrible CTR but after he made changes to his campaign to reflect what Amy suggested, he too was getting 4% CTR.

At this point I figured it was something to do with my setup. I will take your consideration of bots though.

I'll look through the forum regarding bots and I'll include that into my testing.

Thank you!

---

Just to verify, did Amy run this totally unoptimized or she just split tested the LP in one of her running campaigns? If she had already paused some of those bot placements, the results would be different even tho the TS was the same. Im not saying it was the case, just making sure all options were considered Amy?

---

Haha matuloo nice to see you drop in! We're scratching our heads here and Glenn here isn't the only person having this problem, so if you have suggestions we'd sure love to hear them.

To answer your question: No I wasn't running an optimized campaign. We ran the same lander on the same traffic source with the same campaign settings. Otherwise we'd be comparing apples to oranges.

Other than handing over the FTP credentials of my server to Glenn so he could host his landers on my server I'm kind of out of ideas. I was just talking to a friend yesterday who's ALSO running AU sweeps on popads using ripped and modded landers - and he's getting 50-80% CTR. Something doesn't add up...



Amy

---

Indeed I've had this issue too, what's really weird is somehow the Click2conversion / ROI does "add up" to be reasonable...

---

Hey ysekse, thanks for dropping by.

I have a couple more options to test - which is good in a way since I'm learning a lot about the back end.

If those options don't pan out then I'm going to keep going on with my campaign and learn the process.

---

This is super weird. I'd still vote for a more detailed source analysis, try to target just one larger placement and bid high to see if it helps any. Somehow I feel the differences could have something to do with the traffic source algo.

The other possible options would be - create a new account with the traffic source. Make sure its not the hosting - get a virtual with the same hosting as Amy has. Nameservers usually dont cause so big problems, but lets register a new domain with some other registrar and use a service like dnsmadeeasy ... its still weird, not sure what else to suggest

---

Hey Matuloo,

You've actually nailed some of my testing parameters.

This is what I'm planning on testing:

1. Changing Trackers from Voluum to Funnel Flux
When I was testing my lander URL with pingdom and GTMetrix I was getting 150-500ms. This put me at ease in regards to Nameservers and I was content with using S3.

Then it clicked to try the actual tracking url Voluum provided. Testing the Voluum tracking URL, the redirect to my website added an additional 600ms-900ms!! I've contacted Voluum to help me with this since that is a significant amount of time.

I've also started a Funnel Flux install on a server. I'll also be testing that to see if that helps.

2. Test on a different GEO
I've heard of people saying AU is a little slower - I don't know by how much, but it's worth a shot to determine where my problem really is.

3. Fresh Setup
Start absolutely fresh. New Voluum account, new account on Traffic Source, new domain, new server. This would be the extreme but throughout this whole experience I've gained some knowledge and can get this up and running in a couple hours.

Wish me luck LOL

---

Yup, sounds like an extreme, but when shit simply wont click, sometimes its better to restart everything from scratch.

Im havign a hard time to believe GEO could be the problem, since Amy tested the same, but ok, this can be an easy test... try to run in a different GEO and check the results. Im sure Amy will be kind enough to run a small test for you on her host even in the new GEO If you still see the same difference, at least we will know GEO isnt the bottleneck.

Keep us posted!

---

Ok,

Where do I start lol

TESTING
Throughout the entire week last week I tested different variations of CDNs, DNS', Trackers, even going as far as creating new profiles on trackers and traffic sources.

- Tested to see which CDN (Rackspace vs Cloudfront) was faster and if it made a difference
- Signed up for multiple DNS services to see if it mattered
- Tested between Voluum and FunnelFlux
- Tested using a brand new Voluum Account and PopAds account

RESULTS?
All tests came back with similar landing page CTRs - lol

Actually, I take that back. FunnelFlux did provide a slightly higher CTR than Voluum, but Voluum had more conversions over FunnelFlux. Landers the same, offers the same, bidding the same, but Voluum consistently showed more conversions. There were some issues with my Voluum and FunnelFlux testing though. Since I wanted to keep all the variables the same, using the same offer proved to be difficult because (as far as I know) it only accepted 1 postback URL - so it either had to be the Voluum postback or FunnelFlux postback. Because of this, I couldn't run the tests head to head - instead, I ran the Voluum campaigns for 2 days, then I ran the FunnelFlux campaigns for 2 days (making sure I changed postback and making sure I started each campaign at the same time).

Funnelflux provided 5-10% CTRs, whereas Voluum was still under 1%. With that CTR, Voluum was still able to provide more conversions than FunnelFlux.

I don't want to make this too long regarding my testing, but if you want detailed information please PM me.

Back to my campaign

Moving forward, this is my setup:
CDN: Cloudfront
Hosting: S3
Tracker: Voluum

I ran 3 campaigns, low/med/high bid and I was able to get -50% ROI on my high bid with horrible CTR - so I'm not going to worry too much about it. Instead I'm going to focus on optimizing my system.

As I said yesterday I ran a Low/Med/High bid for my iPhone offer in AU.

Tested 5 landers for each.

LOW BID
Was able to get a couple conversions but not enough data to start cutting landers, so I'll run for 1 more day.

HIGH BID
2 of the 5 landers had all the conversions. I've cut the other 3 and included 3 more into my testing.

Drilling down into the HB campaign, I noticed the device types showing equal amounts of traffic - tablet and smartphone showed equal amounts of conversions.

Question - I'm thinking of splitting this campaign into Tablet and Smartphone. If I do split it, how would bidding work? Do I transfer over the same bid from my main campaign? Or would I create additional campaigns for staggered bidding on those as well?

---

April 2nd Update

Total Cost: $60.37
Total Revenue: $24.80

4 Campaigns:
iphone.ALL.LB
Spend: $15
Rev: $6.20
ROI: -58.56%

iphone.ALL.MB
Spend: $15
Rev: $0
ROI: -100.00%

iphone.TAB
Spend: $15
Rev: $9.30
ROI: -38.61%

iphone.MOB
Spend: $15
Rev: $6.20
ROI: -59.29%

TO DO

• Cancel iphone.ALL.MB campaign. Been running it for 2 days and not a single conversion
• Don't have enough data from the other campaigns to start cutting landers.

April 3rd Update

Total Cost: $75.44
Total Revenue: $52.40

3 Campaigns:
iphone.ALL.LB
Spend: $25.13
Rev: $21.70
ROI: -13.65%

• Connection Type:Digging through the stats there's a lot of wifi traffic - I won't be optimizing between carrier/wifi because I don't have enough data to cut out carrier just yet. Also carrier made some conversions with the little traffic it received.
• Device Type:Tablet and Smartphone receive the same amount of traffic but tablet has all the conversions - will be optimizing for Tablet
• OS:IOS received 70% of the traffic as well as all the conversions - will be optimizing for IOS

To-do's for this campaign:

• Create 2 new campaigns. It will target IOS tablet traffic and lead it to the converting landers. 2 campaigns for different bids (1 will be the same bid, 1 will be higher)

iphone.TAB
Spend: $25.04
Rev: $18.60
ROI: -25.72%

To-do's for this campaign:

• Work on sales funnel.
• 1 Lander doing well, will include more landers to test

iphone.MOB
Spend: $25.04
Rev: $9.30
ROI: -63.20%

To-do's for this campaign:

• Work on sales funnel.
• 1 Lander doing well, will include more landers to test

---

Hi Amy may i know do you run on all websites every time in popads?

---

Wow! So FF gave you 5-10x the CTR for the same lander? I wouldn't call that "a slightly higher CTR"! That's a huge difference!

What surprises me even more is that Voluum got more conversions in spite of that difference in CTR. Thanks for giving me this idea - I'm going to run a similar test to yours and publish some results in a separate post.



One reminder: Testing different bids is great, but you're not meant to keep all 3 running to cut landers from all camps while your ROI is negative. This way you'd be losing 3x the test budget which is totally unnecessary! Would be better to pause the 2 camps that have lower ROI and just keep the highest-ROI camp running, and continuing to cut landers.


I would suggest looking at trends for offers and landers - i.e. are the same landers/offers performing well for both devices? Or are you seeing different trends, i.e. landers/offers doing well for smartphones are not doing well for tablets? If it's the latter case, it may be worth it to test and cut offers/landers for each device separately. If the trends are the same, then it may be better to leave both devices in the same camp.


Ah OK! Yes if tablets are converting a lot better and you're eager to reach green, separating it out into a new camp and testing bids. Or, if the trends are the same between phones and tablets (like I mentioned above), then it may make more sense to keep them in the same camp so you can gather data faster and cut faster.

Optimizing for IOS - sounds good as well! But again, if lander/offer trends for android are similar as for IOS, then keeping them into the same camp will let you gather data faster.

Later, when you have better landers, you may want to try this approach on testing offers for separate carriers vs. wifi:

http://stmforum.com/forum/showthread...esting!-Part-1

My guess is that right now you're optimizing for wifi traffic - because it usually comprises of the bulk of the total traffic.


How many landers have you tested so far? If you've tested quite a few already, then it may be better to focus on testing offers instead.



Amy

---

Yes by habit I always start with all websites. Don't know if it's the best way, but because traffic volume is almost always an issue, I like to target as broad as possible at the start, and narrow things down after that.


Amy

---

That's only because of the way data is calculated. FunnelFlux works like Thrive and Google Adwords and uses this definition for the CTR:

- source: https://support.google.com/adwords/answer/2615875?hl=en

Replace ad by lander view.

It means that if a user clicks twice on the same link, it will be counted as 2 clicks.

Voluum uses a different definition and counts clicks only once per user. If a user clicks twice on the same link, it will be counted as 1 click. Hence the lower CTR, even though it's not really getting less clicks.

The real definition of CTR is the way Google described it in the doc shared above. However, the way it's being shown in Voluum makes it easier to compare CTR from different landers.

We're currently adding an option in FunnelFlux to have the choice between both.

For the conversions, it doesn't really make sense that you're getting more with Voluum since your redirect speed was the same or slower. It tells me that all things were not equal. Maybe you didn't run the same days of the week, which affects conversions?

---

Hey Vitavee! Are you sure that Voluum tracks CTR like that? Here is a screenshot of a few placements on a new camp I am running:


PS: Thinking of switching over to FF soon!

Cheers,
Brad

---

Thanks Brad, I just checked on my side and you're right, it's counting the duplicates. It takes some time when clicking refresh as it's not really in realtime, so it's possible that didn't wait long enough to see the duplicate clicks coming in

The problem when we follow the real definition for CTR and count duplicate clicks is that it happens that some bots click several times per second on a link, and makes it very difficult to compare the CTR between different landers because of that. That's why we're currently adding an option to count clicks the way I thought Voluum was doing

---

Hey Glennb,
I've read through this thread and I'm a bit puzzled (technically) as to why your landers are receiving such low CTR's. Without seeing your landers, my only suggestion at the moment is to add more CTA's to your landers. Make all your images clickable, underline text (blue font) and make it clickable, etc, Feel free to PM me your lander and I'll have a look.
Goodluck,
Michael

---

Hey Amy

Thanks alot Now I understand that it is not wise to run all website for long term (just for testing)

But do people run on all websites and focus on blacklist instead?

---

If my view rates go up to a very high amount, will it be better to switch to FF? Am a voluum user atm mate

---

I can only reply yes to this question

Seriously, FF does more, and if your server is in the same region as your users, it's faster. It's even faster in some cases when your visitors are not in the same region as the faster redirect speed makes up for the latency caused by the distance.

---

There are definitely different approaches, and it depends on the particular situation. Some examples:

-If there's a lot of traffic available, it may make sense to build a whitelist, both for testing and for longterm.

-If there's not much traffic available, it may make more sense to just target all sites and blacklist, and only after your funnel is optimized.

-If traffic quality is good overall, may be better to blacklist rather than whitelist.

-If there are lots of bad placements (e.g. bots) it would make sense to either start cutting rigorously from the start or build a whitelist ASAP and use that for testing and beyond.

I tend to not like spending a lot of time messing with placements. If a campaign isn't profitable after a couple of rounds of lander testing, and a few rounds of offer testing, I'd just move on. Ideally I want to hit green before messing with placements, and then by cutting placements keep my campaign profitable longer.


I'm planning to publish a tracker comparison very soon. In short, yes - if you're driving a lot of traffic volume, FF will be cheaper.



Amy

---

Hi Mate!

Thanks alot. My Voluum ends 26April16. I would like to give FF a try as my traffic left 700k events only and its just less than a week

---

Thanks Amy for the sharing

The one that STM has is the blacklist for popads right?




Thanks Amy. I need to decide fast. In another 20 days I need to pay for Voluum again

---

CAMPAIGN OVERVIEW
SPEND: $105.67
REV: $37.20
ROI: -64.80%




0401.IP6S.PU.HB.MOB - CANCEL
• Going to cancel
• Getting data for the past 2 days shows 2 conversions and both those conversions came from an ISP with very low traffic. I checked PopAds to see the amount traffic it generated with my current bid and a higher bid, but it showed 2000-3000


0404.IP6S.PU.HB.TAB.WIFI.ISP.LOW - BARELY MAKING PROFIT
This campaign I've targetted to the converting ISPs


• Not enough data to cut TPG yet
• Increase spend to get more data
• Maybe throw in 2-3 variations of converting landers?

0404.IP6S.PU.HB.TAB.WIFI.ISP.REG - CANCEL
• Going to cancel

0404.IP6S.PU.HB.TAB.WIFI.ISP.HIGH
This campaign I've targetted to the converting ISPs


• Not enough data to cut the other ISP
• Increase spend to get more data
• Maybe throw in 2-3 variations of converting landers?

0403.IP6S.PU.LB.TAB.OS-SAMEBID - CANCEL
• Going to cancel
• Getting data for the past 2 days shows 2 conversions and both those conversions came from an ISP with very low traffic. I checked PopAds to see the amount traffic it generated with my current bid and a higher bid, but it showed 2000-3000

0403.IP6S.PU.LB.TAB.OS-HIGHBID

• Don't know what to do here
• Do I create a new campaign for Telstra and pause this campaign or keep running until I get more data?


ADCASH - 0404.IP6S.PU.HB.TAB.WIFI.ISP.RECOMMEND
Wanted to try moving to another traffic source with this campaign. When scaling to another traffic source, do I try a lot of landers and offers, or just the ones that converted in PopAds?

---

Hey Michael, I too am puzzled by it but I needed to move along.

Great points about making everything clickable. I'll create a new version of the lander thats converting right now and apply your recommendations. I'll include that in today's test.

If the results are the same, I'll forward you the lander to take a peek. Thanks for the help!

---

Id segment the landers by os, browser, and browser version to see if there are any abnormalities, as well as check them in a tool like browserstack. Id also look for malicious javascripts if u ripped the lander.

A pretty good proportion of nontechnical affs run landers with redirect scripts on them.

---

I've checked the landers and removed all the necessary scripts on them. I've also replaced them with my own back button redirect, timer redirect, etc., scripts - still the same result.

---

Put RUM on my lander that was converting - here are the results.

GEO - AU
Median Load Time - 0.9 sec

Here are the states for the lander:

---

Yes - but I would recommend only using that blacklist for geos that have a lot of traffic. I suspect that many placements are neither the best nor the worst, so some people are making money and some people aren't. But the bigger a placement is, the more often it will get blacklisted. So I don't even use that blacklist anymore - I focus on testing landers and offers first, so that I can make more placements profitable.


Glenn - how many offers and landers have you tested so far? And are you seeing any promising offer+lander combo? Just trying to see whether it may be better to test something else.

Regarding that low lander CTR, I remember a couple of other people who are ALSO running sweeps in AU have reported extremely low CTRs as well. Really makes me wonder whether it's a geo+vertical phenomenon. If you do want to test something else, I would recommend antivirus pin submits - the right offers can convert very well for tier 3 geos, and lander CTR should be much higher (not that CTR actually matter unless it points to a technical problem somewhere).



Amy

---

Hello! What service did you use to check the speed?

---

Hey Amy,

I've tested 5 offers - narrowed to 1 offer now.

I've tested 10 landers - narrowed to 2 > then I had 3 different variations of those winning 2 - so about 16 in total.

Here are today's results:


Actually green on one of the campaigns - $10!! So that's a big win!

Campaign 1 can be optimized a little bit more. I noticed a trend with certain ISPS > OS > Landers that do well. I could create a new campaign with those variables, but then I'd cut down my traffic more, as well as test for bids again.

Instead, I'm going to try to form rules in Voluum where if those trends/conditions are met, goes to the specific lander. This way I'll know if that funnel works, without introducing a new variable of bids.

Also Amy, you might notice I'm pretty stubborn about this campaign lol. It's not that I have to make money from this campaign, but more of the fact that I want to know what possible optimizations I can do so I learn the process and the logic so I can transfer that to other campaigns.

That being said, going through this campaign I feel like I can branch off and try something new. I've already reached out to my Affiliate Managers to see what offer I could test with AV.

Mr. Braun, I'm using pingdom's RUM service. Caurmen wrote a detailed post regarding "page loads" and this was one of the ways of identifying how fast the page loads in the actual geo. Here's the link > http://stmforum.com/forum/showthread...highlight=real

---

Green!

Nice!

---

That's great to see!!

Cause really - with 5 offers + 16 landers if you're still not seeing any green combo then it would probably make sense to try something else.

Have you tried scaling this lander+offer combo to other traffic sources yet? That's what I would be doing now if I were you!

And dayum look at that conversion rate! One of every 4-5 people that click through to the offer is converting. Still gives me the urge to physically rip out that lander CTR from those stats and replace it with something sensible haha. But like you've pointed out, as long as the ROI makes sense, then it's worth your while.



Yup - I like to do this too when I don't want to "waste" the data I've already collected in the current campaign.


I really respect your attitude! And you're right of course in saying that the learning is more valuable than profits while you're learning the ropes. I just wanted to make sure you're not going to throw in the towel from the lack of results lol!

Regarding testing other stuff - now that I know your camp is green, I would suggest focusing on scaling instead of testing new stuff - but I'm sure you know best what you're wanting to do.

Congrats on the green BTW! Very exciting after seeing red for so long ain't it?



Amy

---

Little update from the past couple days.



High = Campaign 1 from last update
Low = Campaign 2 from last update that showed profit

I decided to cut Campaign 2 (Low) mainly because it wasn't getting enough traffic and it wasn't profitable after that fluke day.

Instead I focused on Campaign 1 (High) because it was still a similar segment as Low, just with a higher bid.

During the past 3 days on Campaign 1 (High), I tested 3-4 variations of the winning lander. This gave me the chance to not only see which lander would work best, but it gave me a controlled environment of ISP, OS, OS Versions, Browser, Browser Versions, WebsiteID.

After today, I was able to blacklist Browser Types and OS Versions.

I also realized that I need to work on my funnel still. I'm fortunate PopAds provides targeting to that level, but I may not have that luxury on different TS.

TO-DO
- Do more testing on winning lander
- Scale to different traffic sources and setup 3 campaigns (bid low/med/high)

---

Decided to try out PopCash. Through the data I noticed a lot of mobile traffic and their targeting is limited. For this particular campaign, I'm targeting Tablet traffic so I won't be continuing with PopCash at the moment.

Currently trying out other traffic sources as well as moving to a different GEO on PopAds.

---

Home > Mobile > Follow-along Campaigns