Home > Paid Traffic Sources > Mobile

---

PopAds: Bids, Throttling and Time Targeting (6)

---

Hello fellow money stackers!

I have some little questions regards PopAds, I hope you can clarify these questions.

1. Today I created a new campaign in PopAds with a test budged of 15$. I set the bid at 0.003 and for that bid PopAds offered me 40'000 impressions. I though PopAds is smart enough to split the budget through out the day, but I was wrong. In the first two hours PopAds drained 8$ of my budget. With this speed my budget would be 0 in just two more hours. Automatic Throttling could be solution but the minimum spending is 3$ an hour. Too much for this test. I could set manual throttling but I wonder how would that work. I need just 12% of the offered impressions.
(target GEO is in different time zone)

Question #1 How do you go about this issue? Do you just set manual throttling or is there something I am missing here?

2. I noticed that I start to get a lot more traffic towards the night time. I guess thats because there is a lot less competition at night time so there is more traffic for cheaper prices. And here this question relates with the first question. If I let that 15$ budget to go for the whole 24h time period, PopAds will spend most of it in the night. Solution to this would be to create second campaign with different(lower) bid for the night time. But is that how I should do it?
(target GEO is in different time zone)


Question #2 Is it somehow possible to split that money evenly so I could get the better picture of what hours do better?

Hope someone could clarify this up.

---

I honestly think $15 dollars is too little to test. The throttling request I have set is at "Disabled".

For me, I would rather save up some money a do a whole days of test with a reasonable bid target RON. Say you set aside $300-400 dollars to test a campaign for 2-3 days. Within those 2-3 days you will this data that could give you a good idea of what to really target:

- OS-Browser: Which Browser is performing particularly well?
- Hour of Day: Which hours of the day perform the best for you? Find the best hours and then create a campaign based off of this. I would stick to weekdays testing instead of weekends testing.
- Placements: Which placements are draining the majority of your budget, find those placements and zap them out of your campaign.

But, before doing all of this, there is something you should take note of and this is important...... Follow the steps(Other members on this forum may have different approaches, but this has worked for me).

1) Find the best offer that works for you. Find one of the best volume based offers on your network and take that offer and see if other affiliate networks have the same.
2) Rip some landers and test them. Make sure your landers are optimized. Make sure they load fast. Ideal speed would be less than 300ms-1sec. Anything over that is tooooo long of a load time.
3) Find the best winning lander, then use that one lander and approach your campaign with those guidelines from above.

Good Luck!

---

Question #1 How do you go about this issue? Do you just set manual throttling or is there something I am missing here?

What are you testing exactly though? My suggestions is to do multiple tests at once, so then you need a bigger budget anyway. An easy example is if you test multiple LPs and offers at once, you see which is the best combintation and then you can spend more to cut placements and others.

Question #2 Is it somehow possible to split that money evenly so I could get the better picture of what hours do better?

Day parting is really at the end of most tests anyway. You should only test that once you are maxing out on daily spend, else it doesn't tell you what you want to find out. As soon as you up your budget, hours might perform very differently.

---

Ok, I am doing offers that have payouts between 0.5-2$, and read (Mobile Cookbook) that there is no point of spending more than 4x the payout on each lander. So if I am doing 1$ offer and 5 landers, then thats 20$ per campaign, per day. I don't see why I would have to waste
300$, if the offer and landers are not having any conversions after 20$ or 50$. Lets take CPM 2$, for 20$ I could buy 10000 impressions, if there is no conversions after 10000 impressions, them there is big chance that situation wont be much different after 100000. At least thats what is written on a lot of STM tutorials. I understand that if you have 20$ offer, then obviously testing budget will be different.

And thank you very much for those steps. I am 99% doing what you are preaching here!

What am I testing?
I am just following The Mobile Cookbook by caurmen.

I have an offer (or two similar(around 1$ payout)), 4-6 landing pages. I create 4 campaigns low bid and mid bid RON + Cellular and RON + WiFi and set aside 15-20$ budget for each. So total about 80$ on testing. The problem is that my mid bid campaigns eat that 20$ budget in no time.

---

By what are you testing I mean, are you looking for a result to see the best LP, the best offer, the best placements, etc?

The more things you want to see the "best of", the more money you have to spend.

Your $20 amount sounds about right for an iteration. What I would do is spend it in whatever time the traffic source spends it, and draw conclusion for the best LP + offer combination. You won't see green at this stage, you are just trying to find out the strongest combination here, hopefully around -50% or better, and then you can improve the placements and other factors. If you use all combinations to drop placements, you will cut a lot of traffic. I think it's best to find the best creatives and offer before cutting traffic. It gets more expensive to begin with but should have higher returns in the long run.

You should still cut obvious bot placements though, that have a lot of visitors and no clicks.

All this said, the truth is that you might have to spend much more than this to draw conclusions. Also, high bid could make a big difference compared to mid. What I like to do is do the first iteration at a very high bid and see if some conversions come in. The logic there is that a high bid will bring some of the best quality traffic and if that doesn't bring any conversions at all, it's highly unlikely a lower bid does better. I don't expect a good ROI out of it, it's more to give the offer a quick test, since if the offer brings 0 conversions, there's no data to improve on.

---

In the begging I am looking kind of at the big picture. If the offer/lander/bid combination is showing any promise. If it is, I am looking at how did landers perform. Which offer did better with what landers and so on. If there is 0 conversions after 20000 impressions (20$) for a 0.5$ offer, then there is obviously something wrong, and I cant see why would I need to continue investing in this offer.

Ok, I'll top up my budget a bit and try to do this. Lets see how it works out! Thanks for a suggestion!

---

Home > Paid Traffic Sources > Mobile