Home > Mobile > Follow-along Campaigns

---

First Mobile PPV (pops) Dating Campaign in cheap traffic Geos (20)

---

Hi All,

As I'm new the Mobile Marketing arena, I have set myself a fairly modest target of reaching
$100 profit per day with my first mobile pops campaign(s).

After reading a little in STM about Mobile and PPV traffic it seems that its a good plan
to try the cheaper traffic GEO's and offers with low payouts to gather data quickly so
that optimisation can happen quite quickly.

So, to begin just 2 simple dating campaigns.

Campaign 1
GEO: India
Traffic Source: ZeroPark Mobile Pops set to RON (Currently bidding at 0.0003, this shows an average bid position of 4)
Offer: Standard Dating offer, payout $0.45 (have found the offer on 2 Affiliate networks so am A/B testing these two).
Landers : 4 Landers (2 ripped and 2 of my own)

Campaign 2
GEO: Africa
Traffic Source: ZeroPark Mobile Pops set to RON.(Currently bidding at 0.0004, this shows an average bid position of 3)
Offer: Standard Dating offer, payout $0.45 (have found the offer on 2 Affiliate networks so am A/B testing these two).
Landers : 4 Landers (2 ripped and 2 of my own)

Results:

I didn't have Voluum tracking setup correctly yesterday (this has been fixed now) so, I just have to type out the results.
So, for yesterday.

Campaign 1 India:

Spend = $0.45
Visits = 4474
Clicks = 242 so CTR of 5.41%
Conversions = 1
Revenue = $0.45
Profit/Loss = $0

Campaign 2 Africa

Spend = $18
Visits = 19542
Clicks = 2393 so CTR of 12.24%
Conversions = 2
Revenue = $0.90
Profit/Loss = -$17.10

Will add today's results at the end of today, but any thoughts/opinions so far would be appreciated.

---

If you're looking for any extra cheap run of world desktop/mobile pops please let me know. We have 80-100mm a day available

---

Great start! Looking forward to following-along!

With so few conversions, you can't cut any offers or landers yet. You really need to just be patient and run more traffic to get an idea on whether the offers are even promising enough to continue.

Also, in addition to split-testing the same offer from different networks, be sure to also include different offers (but still the same geo and offer type so you can use the same landers).

Also be sure to check each offer's requirements to see what OSs and carriers they'll each accept, and just target those. When you have more conversions you'll be able to drill down into stats to identify the best-converting traffic segments and focus on those.

Looking forward to your next update!


Amy

---

@vortex

Thanks for the positive feedback, much appreciated. I will follow your advice and update this one when I have more data.

---

Since you're running on zeropark, you also need to get rid of all the many bot placements there are, read this and use it : http://stmforum.com/forum/showthread...n-Any-Campaign

As Vortex suggested, its too soon to cut landers and offers, focus on cutting under-performing and BOT placements first.

Is this adult or mainstream dating? Pops are generally not the best traffic type for dating, so don't be surprised to get quality complaints pretty soon. Not saying it definitely will be a problem, but usually it is.

---

@matuloo thanks for feedback.This is mainstream dating.
I am currently implementing that Bot-test lander, hopefully it can save me wasted cash.

You say Pops are not the best traffic for dating. What traffic source would you advise? Would Mobile display be
a better option?

---

The best for dating, when it comes to quality is still desktop traffic. But, its also the most expensive, so the numbers will eventually balance out. When comparing pops and display, display has usually higher quality, yes. I dont like to generalize too much, but such is my experience, pop traffic almost always resulted in quality issues. But you can always mix it with more expensive quality display traffic. At the end of the day, its about finding what works for you

---

For dating offers, I've heard that traffic from dating site membership areas can convert really well. After all, people that are already paying members are more likely to subscribe to other dating sites as well.

Etology and Greenpointads have told me that they have dating site traffic. Please send me a PM if you're interested and I'll refer you.


Amy

---

Members area traffic is very good, whatever niche we touch, but the prices are pretty insane too

---

Okay so what is Members area traffic exactly? Does this mean that my pops could show when a user is logged into another
dating site? That could lead to a good conversion rate for sure.

---

Doesnt have to be just dating site, pretty much any site where a membership (paid) is needed to get into. So for example a banner on top of this page would be members area traffic too

This will be more banner than pops tho. They dont want to annoy the paying member more than needed.

If you want to try this, prepare for a few obstacles - its not self serve, there will be more rules than normally, you will need to take some volume and not a 1 day test, and the price will be higher than anything you have seen so far

---

Oops I just realised after reading Vortex's excellent pops guide again that I had forgotten about this:

"Test multiple bids from the start. Some traffic sources will provide an "average" or "recommended" bid, but bidding at the average will not always give you the best value for your money. Make copies of the new campaign so you have 3-5, and set a different bid for each ranging from low to average to high, set a low budget for each (10-20x payout should be enough). Run traffic, see which bid gives the best ROI, and use that camp to test your landers and offers."

So, is it still a good idea for me to try this method even though Im a few days into the campaigns now?

---

Since you already started, I would continue for some more time, try to figure out the best offer/LP combo and then take those and run them at multiple bid levels.

As for the offers : I would also cut the 2 in both India and Indonesia, in Nigeria you dont have enough data yet. Since the offer performed so differently, you cant really cut landers yet because the data is largely affected by the fact that some of those offers are pretty much dead. So cut the worst offers and run some more. Dont touch LPs yet.

How about bots? Did you identify some bot placements already? These can skew data a lot too.

---

@matuloo

Thanks for the feedback. Yes I did identify some Bot placements and have paused those. Hopefully that can help to improve ROI.
Yes good idea, I will try and work out the best lander/offer combos and then run those at staged bids.

---

Have been working more on all this. Will post results soon.

But I have another quick question.

Once, I have a funnel that is working fairly well. i.e Landing page/Offer combination and Optimised traffic targetting such as Carrier/Device/Browser/TimeofDay etc that is (hopefully) giving me consistent positive ROI. Will this formula then just be able to rolled across other traffic networks. Or does the whole optimisation process have to start from the beginning again when a new traffic source is tried?

---

You cannot roll it over just like that, because of one thing, there will be different placements on the new traffic source.

Truth is, with a proven funnel, it will be way easier, because large part of the legitimate placements will work with the same funnel. You will still have to catch the bot placements again. And the negative side of a new source, some legitimate placements will not work and you will need to optimize them again if you want to run on those too.

BTW: dont focus on the dayparting part too much when transferring campaign to a new source, its very dependent on the placements and will most likely be different from source to source.

---

You have 1000s of clicks without conversions but then there are some with 1 o 2 conversions, there is really no significant difference between those just yet. All it takes is 1 or 2 lucky conversions and it would all look differently. 3 to 1 looks clear but its not, multiply it by 3 and you get 9 to 3 and its all clear now, because it can no longer be a coincidence. You need more data for sure.

The whole idea of dayparting is not something Im a fan of. Im pretty much not using it at all. Im looking for campaigns that end up positive when in running in 24/7 regime. Its a lot about effectivity too, I dont have time to follow dozens of campaign and pausing/unpausing them several times during the day

You really should focus on maximizing the funnel for now and getting rid of bots. Only resort to dayparting when you identify pockets of profit in a campaign, that you otherwise cant run as it would loose money no matter what you do.

---

Just thinking ahead a little. Once I have a funnel that I am happy with for my mobile pops campaigns. Does it then make sense to duplicate that
into a desktop pops campaign? If so then would Run of Network traffic or a more targeted keyword based campaign be a good option (still talking
Zeropark for the traffic source).

---

Nice stats-crunching and observations! And all that green makes the camp look very promising.

I have a question for you: When you drill down to Android -> landers are you seeing green for any lander?

If the answer is yes, then please compare lander stats for Android and Windows Phone, and observe lander performance trends, i.e. are the same landers doing well/badly for both OSs? Or are certain landers doing well on one OS but not the other?

If you see the same trend, then just keep running both OSs and keep cutting landers.

If you're seeing different trends, then set up a separate camp for Windows Phone so that you're running the two OSs separately, and cut landers for each separately.

And if you're not seeing any green lander for Android, then consider pausing that for now if you want to save budget and focus on Windows Phone. Of course if your budget allows, by all means keep Android running - maybe wait until the lander testing yields a winner and test new offers to get that ROI up!

At any rate, I wouldn't recommend cutting carriers at this point because 1)you haven't spent enough on each carrier and 2)after you cut underperforming landers, some of the currently-red carriers may turn green.

You're making wonderful progress! Keep it up!!



Amy

---

Oops forgot to answer some of your questions...

Regarding why offers perform differently for different carriers: It's definitely not because the audience is different! I mean you can have the same visitor using data on their phone when they're outside, only to switch over to using wifi when they're at work/home.

The biggest cause of the difference in conversion rates between carrier and wifi traffic for SOME offers, is the fact that these offers will provide different conversion flows for each traffic type. So when a visitor is using cell data (i.e. carrier) to reach the offer page, they may be able to just tap once or enter a phone number and be subscribed to complete a conversion. Whereas if the visitor was using wifi to reach the offer page, they would be required to send back an SMS or submit a pin - I'm sure you can guess how many visitors will bail instead of continuing to complete the conversion.

As for why some carriers convert better for certain offers - I actually haven't thought much about this so have no idea. If you ever find out please let me know. I prefer to make decisions based on stats alone without thinking about the root causes too much, which may or may not be a good thing.

Regarding the optimization strategy you proposed: Sounds sensible! Just be careful that you don't start setting up a bunch of camps to each target a small traffic segment. That way it will take forever for anything to reach stat sig and you'll end up babysitting a bunch of camps each making pennies a day. As much as I'm a fan of granular testing, I'm always careful not to go past the point of diminishing returns. We each have a finite amount of time, and over-optimizing is a real danger when it comes to analyzing stats and managing camps. Just a heads up.

I find that a good balance is to identify traffic segments that are a)giving you the most traffic volume and/or b)giving you high conversion rates, and then just focusing on those. For these major traffic segments, testing offers and landers for them separately may be justified.



Amy

---

Home > Mobile > Follow-along Campaigns