Newbie on Pops - Third attempt. 1st very timid green. And now? (2)
Hi, parsifino!
How's it going?
I want to congratulate you on your first positive results. The way of Media Buying is not that easy at the beginning but - as soon as you understand all the logical reasoning and processes - you'll achieve great results.
I wanna give you some suggestions so as to help you with your process. I'm also gonna answer some of your questions.
First, I recommend you be more patient and gather more stats before optimization. I reckon that making your first changes based only on 4 total conversions is not the right course of action. You've gotta give your traffic some time for it to work, especially since you've picked 9 offers to test. Each of those 9 should get the same equally-signficant amount of traffic.
Another mistake you've made was choosing iOS as your targeted OS only based on 3 conversions. This clearly isn't the logical thing to do. Why? Because these three conversions could have been the product of pure luck and not the cause of a specific trend. Moreover, by choosing iOS, you avoided the OS with the highest amount of traffic and more conversions. This means you've limited your potential revenue. Yes, it was negative. Even so, you can make some other actions such as playing with the bid/targeting in order to avoid cutting the volume.
Regarding your questions:
1. If - out of the RON campaign - you got only one placement converting (based on relevant stats) – you can close it and leave only the Whitelist running (don’t forget to retest the RON after some time). I don’t know whether or not you've changed the bid for your Whitelist. Even so, always try to put the highest possible bid so as to get the first position. Why? Because you'll have more volume of quality targeted traffic. After that, you can try to open cap.
Moreover - now that you know which offer is working for that specific site – try to adapt your prelander to offer and placement.
2. You can retest the RON campaign with the only one offer (which you already know is good) in order to see how this offer behaves in other placements.
3. Normally, the first step is to find the good offer which would work for your traffic, when you have it. You can also play with other parameters.
4. If you want to achieve results in Media Buying you need to take risks and not be afraid of losing money. This is part of the game. If you really wanna know the traffic source/placement/types of offers working there etc. you need to invest. Don’t forget that'll you pay for the info.
5. Yes, sure. Try them again with this offer which is profitable for you. At least now you'll test the traffic source with the aim of testing placements, not offers.
6. For your profitable site, you definitely need to try increasing the bid. You should also try to increase the bid for your RON campaign (with the best offer only) in order to increase the traffic quality.
7. This can be related to the placements (and their number) and can be related to your position for the traffic.
I wish you good luck on your Media Buying journey!
I know you'll see those sweet profits go up like crazy!
Cheers!
Hi Lorenzo! I've come across your recent follow-alongs - hope you keep updating them!
Some comments below...
Abandoned Lander 2 (obvious)
Abandoned all traffic sources but PopCash for now.
Kept only the 3 most performing offers, in terms of conversions (payout not taken into account for now). Payouts: 0.55$, 4.30$, 6.00$
For cutting landers:
http://stmforum.com/forum/showthread...Banners-Part-1
For cutting offers:
http://stmforum.com/forum/showthread...211#post289211
When payouts are as different as the offers you're testing, you MUST take payout into account. Trust me - I love to cut corners when I can, but the offer and lander are basically your entire campaign, so make sure you don't cut anything earlier than you should.
I would suggest cutting the inferior candidate at less than 5% probability of being best (another way of saying that is, when the winner is at greater than 95% probability of being best).
Also: It would be necessary to drill down into your stats, to see if there are landers/offers that are performing especially well for certain major traffic segments. For example, some landers and offers may be converting a lot better for android, or a specific carrier. Also, some offers may convert a lot better with specific landers. Therefore, even when a split-test calculator is telling you that an offer or lander is ready for cutting based on overall stats, it would be good to drill down into your stats to make sure that the offer/lander you're about to cut, isn't an exceptionally good performer for any major traffic segments.
This is why, when I'm mass-testing offers, I like to run a lot of traffic to get a lot of conversions before I start cutting anything, just so I could drill down into various traffic segments to see whether offers are performing differently for EACH major traffic segment. For example if I see that a particular offer is doing especially well for android, or a major carrier, then I would start a separate camp for that OS/carrier to test offers to identify the best offer for that OS/carrier.
More testing from PopCash. Today Android outperfomed any other OS so this confirms that the IOS only campaigns were based on a very noob assumption and low data.
Discovered one interesting thing about placements. Almost all conversions are from 1 placement only:
Actions taken:
Whitelisted the only one profitable placement.
So yes you can completely just target that one placement to conduct your offer and lander testing.
For OS though - you could have used either IOS or Android to test your offers and landers, but keep in mind that offers and landers can perform differently for different OSs. So, if Android is doing better, you may want to pause IOS for now to get green faster. But when you want to retest IOS later, you may need to test all offers again.
My first GREEN! Please let me post it, even if it's a shame:
Conversions are 100% from mobile, 3G, and only a specific mobile carrier and double the ROI. However, I can't target the carrier on PopCash. So no more optimization can be done here.
Any idea on how to improve the earnings?- I believe that whitelisting 1 placement only is good to achieve breakeven and to recover the testing budget, but there is no room for scale now. Also, the difference with other placements was incredible. What I had to do otherwise?
- I believe that I had to test more landers from the beginning but this required twice the budget, while I wanted to see GREEN fast (needed to prove that AM can work for me too)! Ok, too greedy. But did you see any higher promise from this camp that I didn't see?
- I believe that I cut too much too fast. But to gather enough data, too high budget was needed compared to the promise of the campaign. Or not?
- Is there any room to scale to other sources, even with the big failure I experienced so far?
- Maybe attempt higher/lower different bids for the same placement on PopCash?
- Why such an higher CTR and CR for PopCash?
- anything else missing from the above?
Bids are always a good thing to test. You can start testing bids even now. Clone the camp a couple of times and set them to different bids and see which one does better.
I wouldn't try to scale again before cutting down to a best offer+lander combo.
Amy
Home > Mobile > Follow-along Campaigns