Home > Mobile > Follow-along Campaigns

---

Follow Along Mobile pops with pin submits (19)

---

Had a follow along a while back, but was using decisive and they banned use of the name of the offer I was hitting even with. Switched to tapit and was paying much more for traffic with lower success rates so I decided to try something different.

Mobile pops.

Campaign 1:
Geo: Poland
Offers: 2
Payout: 3.7 & 5.5
LPs: 5
spend: 30.00
conversions: 0
Rev: 0
ROI: -100%

Damn, that is some low CTR for pops! Not sure why, is this bot traffic and I need to kill a bunch of plcements to improve it?

Don't mind my lack of costs, I forgot to enter what I was paying per pop view in Voluum

Placements


Based on talking to my mastermind, and the camp that will follow this, I know these ctrs are BS for pops

So I killed any placement that had over 50 impressions and a ctr of less than 2%, or the top 19 placements by spend, or 50% of my total traffic and spend

Updates soon...

Campaign 2:
Geo: ZA
Offers: 1
Payout:1.9
LPs: 3
spend: 15.00 (only one of two days do far)
conversions: 0
Rev: 0
ROI: -100%


Much better ctr so far, but still under the ideal 20%


And placements here are doing much better. Looks to be some promising, high CTR


Actions taken: blacklisted top 4 placements by spend that had ctrs under 10%.
Bumped up bid to be able to spend my daily budget based on traffic planner.

Update soon....

What do people think of my process? Am I blacklisting too early, too aggressively, not aggressively enough?

---

aarg, how do I make "follow along" in the title be red like everyone else?

---

Campaign 1, day 2 :
Geo: Poland
Offers: 2
Payout: 3.7 & 5.5
LPs: 5
spend: 10
conversions: 0
Rev: 0
ROI: -100%

CTRs slightly higher, but still lots of shitty traffic it seems


if this is bot heavy traffic, direct linking will not really tell me which are bot heavy placements or not until I get some decent conversions happening, and still none yet. Not sure if I should leave direct linking or take it out until I find placements with a decent ctr that is not shit traffic. Any thoughts?

Lots of shit placements it seems still. Killed top 20 by spend again.


Also blocked wifi traffic, carrier only now.

Campaign 2, d2:
Offers: 1
Payout:1.7
LPs: 3
spend: 15.00
conversions: 4
Rev: 6.8
ROI: -55%

One of each lander and direct linking converted lol

Similarly, placements exhibited no strong tendency to convert, and no placements were money sucks, so nothing done beyond funding another day to gather more data before taking further action

---

Don't you feel it's kind of early to kill those placements? Some of them got less than 20 Visits.

The payout of your offers is $3.70 and $5.50. You can't expect them to convert after spending just $0.45.

---

I think your problem is lander, not placement. Test more landers to get higher CTR before cut anything.

Just my $.02

---

When you post the thread, you can choose "Follow Along" as prefix, before writing the title.
Try to edit your first post and see if you still can do that.

---

Dropped the Poland campaign to focus on the other. I bumped up spend on poland, removed my shitty blacklisting, and let it run with no conversions. Maybe my landers suck, maybe the traffic does....back to the drawing board.

Campaign 2:

Offers: 1
Payout:1.7
LPs: 3
spend: 15.21
conversions: 3
Rev: 5.1
ROI: -66



Finally seeing the beginning of some statistical difference in LP performance!
But wanted to see more conversions before I made a decision.
Placements: none were sucking too much out of the budget, so no killing.

Instead added in 3 new landers. and killed direct linked because ROI was shittiest, and bumped spend
Next day:
Spend 25$, grab 4 conversions only on the new landers.
In the morning 2 were even green! later in day not so much though. But promising so far with much better ctrs!.



As a result of this, I removed the worst performing lander. It did not have the lowest ctr, but it did have the worst ROI of -85 vs the remainder which ranged from -70 to -10

I killed two placements. One that had one conversion but at -70 ROI, and another that had spent almost 1.5 payout.

more data soon....

---

I need to ask. Why do you write: „I'm seeing the beginning of some statistical difference“?

Your data is FAR away from statistical significance. If you really care about this, use a tool like this: http://drpete.co/split-test-calculator. You don't always need statistical significance, but you should understand the difference and their consequences.

Or am I the one getting it wrong?

Edit: Check this one out: http://abtestguide.com/calc/. There is a real chance that your worst performing lander (the one you kicked) actually performs a better than your best performing lander (the one with the highest ROI in your screenshot).

---

I may be incorrect? In the last image I posted, the lander in the second position vs 5th position
These are my thoughts on why the 2nd is better than the fifth

Lander A had 252 views, 57 clicks, and 2 conversions for a CTR of 22.6% and a cr of 3.51%
Lander B had 1024 views 174 clicks, and 1 conversion for a ctr of 17 % and a cr of 0.57%

I used http://splittester.com/
to compare both the ctrs and the crs (using conversions instead of clicks)

when ctrs were compared, A was 95% certain to be better
when crs were compared A was 85% certain to be better

So I thought those ads were significantly different enough to kill one? was I incorrect?

---

Hey! I am really much interested to know few things.

1. Which traffic source you're using now?
2. Which GEO you're doing now?
3. You are doing quiz landers?

Thank you.

---

Spend more.

---

Yes you'll have to spend way more - the clicks do not matter much, you have to look at the conversions to make the decision which lander performs better.
I've had one lander with 4 times higher CTR than another one, and in the end the lander with the 4 times LOWER CTR ended up generating a better ROI than the one with the high CTR.

---

You're right, if you look at the clickrate of your creatives you have enough data, but you do not based on their conversion rate. This is obviously not 100% applicable to affiliate marketing, but in "real life" I would never base a decision on an a/b-test, if one of the following things is not true:

1. Every variation got > 100 conversions.
2. The data is statistical significant.

---

so it really could be 2 variables here.

Traffic could be shit , doesn't look like it cause you are getting some conversions with it.

and your landing pages. My guess is your LP needs some serious tweaking. I'd run 2 test with 2 completely different LP's and maybe even one with a JS alert redirect only because you saw some success DL'ing. From the 2 completely different LP's you can see which performs better, the one that performs better start tweaking it for higher CTR's (alerts, OS call out, device, model, date call outs ect.) Then split test it against another insanely different LP. hope that helps.

---

This is really spot on. CTR as a metric really doesn't mean much and gets far too much hype. It's all about the relative revenues / earnings per view...

---

I am realizing this now, thanks guys! No more cutting of anything till I have more conversions. Unfortunately I cannot seem to spend more than 15 to 20 a day on this campaign without massively upping my bids, so I will let it run for a while. I upped daily budget to 50 without it increasing traffic. Very competitive geo for traffic with my targeting I think. I am currently in bid position 15 bidding around 0.017 per popunder. I can up it to 0.1 and only hit position 13.


Thanks! I did fuck up and base too much on ctr and not enough on conversions.

Thanks for the insight into your process! very valuable. So when running a camp you would suggest only 2 or 3 landers, wait till significant data and then kill the worst ones. When you make variations on the best LP, do you prefer to go all out and make like 6 different ones to split test, or just have 3 or so running at a time? Then, once you have the best one narrowed down split test something new and totally different into the rotation.

Could you provide the script for the JS alert redirect? I am not familiar with the code for this. thanks! Also, check your PMs, sent you a message about mobile ad scout.




So my traffic source does not let me separate android phone from tablet. As a result 20% of my traffic cannot be converted as this is a pin sub. So I talked to my AM, and she suggested I redirect the tablet to DU speed booster. Made up 4 landers and angles for it and got the redirect going.

Also found the main offer on another network for a 2.00 payout instead of 1.7, so I split test it into the rotation.

First day with the redirect going


Offers: 2 (+1 redirect)
Payout:1.7 &2.0 ( &0.23)
spend: 14.80
conversions: 4 (3 main, 1 redirect)
Rev: 5.33
ROI: -63%

2nd day with redirect going things went bad. None of my main offer converted, but 4 $0.23 conversions for DU

Offers: 2 (+1 redirect)
Payout:1.7 &2.0 ( &0.23)
spend: 25.24
conversions: 4 (4 redirect)
Rev: 0.92
ROI: -96%

Let it keep going, and right after this day was over, in the space of the first few minutes of day 3 with the redirect added in, I had 3 conversions, 2@2.00 and 1 @1.7. Literally with 0.5 spent so far at that point. Then nothing for the rest of the day yet. More data soon, but being at +180% ROI for the first half of the day is making me happy

Also picked out my best 2 landers and targeted the placements that converted. Same bid with a spend of 20$ a day. After 3 days it has spent 4$ and made 1 conversion at 1.7

So I am not sure how to continue at this point. I can't seem to cheaply scale my traffic. I have contacted a few other popunder traffic sources, and nobody has big volume in this geo, yet my AM says some guys are doing over 5k in revenue a day, so I am not sure where they are getting their traffic from.

Anybody have a popunder traffic source for ZA that has a huge number of impressions a day? 40k across devices is what everyone seems to have, and after carrier and device targeting this is cut down hugely.


Finally, I need to improve my record keeping.

I had been just entering my bid for pops as my cpc in Voluum. However I have found that what I bid is not necessarily what I get. Oddly enough when I bumped my bids up to 0.021, the rate I got them at was actually 0.005. Voluum's data when I enter 0.021 is then off significantly. And i get very different rates daily, so i am not sure how best to keep track of this. I think I will have to use excell, or are there any Voluum tricks I have missed?

---

You don't pay what you bid because the traffic source most likely charges you only what you need to pay in order to get that traffic. Let me give you an example:

Bidder 1: 0,1 / Redirect
Bidder 2: 0,2/ Redirect

Bidder 1 is on the 2nd position, but Bidder 2 pays only 0,11 for the first position. How much more he needs to pay depends on the traffic source, that at least is all I know about this.

---

Well I have not been updating this thread, but still taking everything into consideration above and I have reached profitable! small volume but it is promising so far!

Also, anyone know what is up with CTRs over 100% for a lander that is just a JS pop that redirects to the offer when the OK button is clicked?

---

It's literally that. You're forcing them to click through the offer. Whether they like it or not.

---

Home > Mobile > Follow-along Campaigns