We are back! (With a lot of questions :) ) (4)
Hi there,
We are back! Me and my buddy decided to give it one good last chance.
The last year we have learned a lot gathered some information and we definitely got more insight that ever before.
Still, there are a lot of questions left. Would you guys be so kind to answer them? We pinky swear we bring some value in return.
Regarding Decisive/DSP/RTB/MOBILE COOKBOOK.
1) The nice thing about DSP/RTB and especially Decisive is that the data is very transparent. Especially compared to blind networks like Leadbolt/Inmobi (own inventory). However, we also noticed that the quality and volume is usually bigger on these. Fortunately, the data seems to cheaper.
Question: We still have a lot to learn. Do you guys still recommend to start on Decisive anyways? Or start right away on Leadbolt/Inmobi?
2) After running some data, (with setting the CPA goal to 1,5 the payout), it is quite obvious to see that like 100 placeholders get a single win. I suppose you have to bid higher to get to that (probably higher quality) inventory?
Question: If we put the CPA goal to 3x the payout, does this mean we also bid higher?
3) We still have to optimize our creatives. So to get this straight: you use the SMART CPM to investigate the right bid for that placement? In addition, you first start with CPC and when the CTR is higher than 1% you switch to CPM? That is, after SMART CPM, you start with CPC?
4) In the cookbook on Stack that money, they recommend you to create breakout campaigns from well performing placements (sites/apps).
Question: Does this mean you have to blacklist that placement from your main campaign? (to explore further).
More general questions:
5) What confuses me tho, is that we also noticed that when you put a blinking arrow on your banner >>, the CTR gets higher but this does not increase the CVR per definition. On the other hand, a slight text animation does increase the CTR and CVR. That is, using an arrow on your banner, does cost you more on a CPC based payment. But obviously a higher CTR means also cheaper clicks.
Question: Use a blinking arrow on your call to action or not? (or is purely dependent on CPC/CPM)? No animation means also smaller file size, does this have any effect?
6) In the beginning, you usually get a lot of conversions, however, soon it gets less, I expect that a lot of placement holders (app/site) have such a low volume, that when we bid on them, we automaticly buy all the available volume. Another option is to let it run for a few days, to see which placements are stable (after 4 days or so).
Question: What should the minimum daily available bids be to be a viable placement? Or what is the best approach to this?
7) About banner blindness.
Question: Do you have to change your banners every 3 days, or is having 5 in rotation a sufficient idea? What is the best approach?
8) When we use What Runs Where for one dollar. We found a similar campaign for a blind network. Which gives all placements. The AM said he would give all the Site _ IDs of those placements, so we can litterally copy the campaign, evil, yes.
Question: Do you think this is a good idea to start off? Or is bidding against someone else a silly idea?
9) What kind of surprises me, is that usually on this forum, some traffic sources are commonly named: Decisive, Tapit, Inmobi, Leadbolt etc. However, I never hear of: Adx, AppNexus, Improve Digital, Microsoft Ad Exch, Rubicon, which seem pretty decent traffic sources as well.
Question: Why are those traffic sources not commonly named? Too much competition?
Seriously, SERIOUSLY! Thanks a million.
Cheers
PIM
You can also send me a pm if you prefer it that way.
1) The nice thing about DSP/RTB and especially Decisive is that the data is very transparent. Especially compared to blind networks like Leadbolt/Inmobi (own inventory). However, we also noticed that the quality and volume is usually bigger on these. Fortunately, the data seems to cheaper.
Question: We still have a lot to learn. Do you guys still recommend to start on Decisive anyways? Or start right away on Leadbolt/Inmobi?
For the sake of learning I would stick to an RTB, it could go2mobi, avazudsp, decisive, just any RTB really.
2) After running some data, (with setting the CPA goal to 1,5 the payout), it is quite obvious to see that like 100 placeholders get a single win. I suppose you have to bid higher to get to that (probably higher quality) inventory?
Question: If we put the CPA goal to 3x the payout, does this mean we also bid higher?
Yes.
3) We still have to optimize our creatives. So to get this straight: you use the SMART CPM to investigate the right bid for that placement? In addition, you first start with CPC and when the CTR is higher than 1% you switch to CPM? That is, after SMART CPM, you start with CPC?
There is no one way to do this. I usually start with Smart CPM then switch to manual when I want more volume, or predictability in terms of bids.
4) In the cookbook on Stack that money, they recommend you to create breakout campaigns from well performing placements (sites/apps).
Question: Does this mean you have to blacklist that placement from your main campaign? (to explore further).
Yes.
More general questions:
5) What confuses me tho, is that we also noticed that when you put a blinking arrow on your banner >>, the CTR gets higher but this does not increase the CVR per definition. On the other hand, a slight text animation does increase the CTR and CVR. That is, using an arrow on your banner, does cost you more on a CPC based payment. But obviously a higher CTR means also cheaper clicks.
Question: Use a blinking arrow on your call to action or not? (or is purely dependent on CPC/CPM)? No animation means also smaller file size, does this have any effect?
You shouldn't be bidding CPC on an RTB.
6) In the beginning, you usually get a lot of conversions, however, soon it gets less, I expect that a lot of placement holders (app/site) have such a low volume, that when we bid on them, we automaticly buy all the available volume. Another option is to let it run for a few days, to see which placements are stable (after 4 days or so).
Question: What should the minimum daily available bids be to be a viable placement? Or what is the best approach to this?
If you are spending $5/day on 1000 placements that ain't to bad. You shouldn't shape your campaign around a couple of placements, unless you have a mega placement and in that case you shouldn't run into your initial issue.
7) About banner blindness.
Question: Do you have to change your banners every 3 days, or is having 5 in rotation a sufficient idea? What is the best approach?
I've run the same banners for 6 months at a time. There are no set rules here, keep an eye on your CTRs if they drop majorly then test some new banners.
8) When we use what runs where for one dollar. We found a similar campaign for a blind network. Which gives all placements. The AM said he would give all the Site _ IDs of those placements, so we can litterally copy the campaign, evil, yes.
Question: Do you think this is a good idea to start off? Or is bidding against someone else a silly idea?
Depends if you want to learn a process, or just focus on making money. 
Why not try to deconstruct the campaign, understand why and how they came to the point that it is at now. Then try to recreate that with your own twist.
9) What kind of surprises me, is that usually on this forum, some traffic sources are commonly named: Decisive, Tapit, Inmobi, Leadbolt etc. However, I never hear of: Adx, AppNexus, Improve Digital, Microsoft Ad Exch, Rubicon, which seem pretty decent traffic sources as well.
The first ones are the most popular, it's not to say others don't use the second batch you mentioned.
Home > The Newbie Zone > Questions and Answers