Hello Guys,
As there are lot of questions on push notifications, wanted to share my experience with that, as we think that this traffic sources is rapidly growing, but there are still some things to mention, to fix or to work with.
Generally we are working with many networks now who support with Push, and if you can suggest more, I`m ready to test them as well.
Currently we are working with Prop, ZP, Richpush, AdMaven, Adsterra and Datspush as well. Daily we are spending on pushes around 2000$ or more and testing a lot. So here are things we consider the most important one and share our experience.
1. with Pushes, we are changing texts and icons quite often. It is required in every 2-3 days usually, as ctr is most important here to get cheap clicks. The most important factors are icons and Headings in Pushes as far we see.
2. Changing and testing offers as well in several days.
3. We tried to duplicate working campaigns but if the network has no big amount of Pushes, it can kill campaigns pretty fast and overall profit can be low. Also, duplicating with exactly same settings and texts, can be pretty bad idea. Especially if you are working with high budgets.
4. There is problem also with Tier 3 countries, as many network still has minimum bid of 0.01 and its quite high for many Tier 3 countries. We are still asking to networks, to give us minimum bid possibility at least 0.005 to be profitable. So if networks will give this opportunity in general, it will be better for all.
5. We tried to bid higher, to see if we get better quality traffic, but it doesnt work for us.
Thanks for sharing these valuable information, push traffic is exploding, so I believe many members will find this valuable!
Based on my "limited" tests, I can agree with pretty much all the points you made ... frequent changes of the ads are extremely important... I've never worked with any other traffic type that would require so frequent changes 
And yes, the prices for developing GEOs should definitely be lower.
What type of offers do you see working well on Push?
I have signed up to a few push subscriptions and I mainly see the following:
- UK IP > Crypto offers
- Spanish IP > Diet and Dating
Thank you for sharing your valuable experience, ptgtate! Push is becoming the new POPs. Many tests, case studies will be published. Lots of money and time is waiting to be burn into push!
LeadCloak
thanks for sharing @ptgtate! I'm agree with all the points.
I add a few more:
- Badge image its the smaller image (sometimes not configurable in some networks) but having a good one can change everything.
- subscriber age its a key indicator of high CR, so, working with networks that can optimize based on that it's very important (after 1 month the CR decrease dramatically)
- its very important to work with networks that have their own databases of subscribers
- those who click once, usually click again (about 60%) (remarketing?)
- Verticals that worked for us: Casino, Sweepstakes, CC sweepstakes, COD, CPI
All this info comes from our Beta Self Service for push notifications. Would be nice to have feedback from all of you!
Push self service: https://push.billymob.com
Best,
I tried the ones that you commented in your post, but unfortunately, as far as I seen, none of them you can filter/optimize by subscriber age, correct me if I'm wrong.
Hey 
Thanks for all the info 
But it's not like pop, there is a lot less trash there, so let's say on tier 3 you can get a lot more clicks with lower bids, but on Tier 1, without a decent rate, you won't get many clicks.
But I totally agree, Changing creatives is key here. You'll see awesome CTR% at first, but the user is saturated to a state he sees the same image over and over he won't click it. Were adding daily over 2.2 new push subscribers, so the inventory always is different. but you do want to take advantage of the users in the higher frequencies, meaning the not fresh users that haven't clicked your push.
I have run over 7-figures on push this year.
This is what is important and what is not...
What is not important:
- badge (thoroughly tested and even the best one was so marginal its hard to tell the real performance impact - I default mine to an message icon or get creative but dont waste alot of time)
- image (in most cases the users will never open the dropdown on their android to look at this image, while there has been some cases on mobile that this has helped it is mostly a benefit of windows desktop push - I focus on mobile.)
- don't waste time looking for tips on how to run push, jump in now and figure things out on your own because by the time you learn all the shared tips I will have drained every push users wallet - you'll be too late 
What is important:
- Headline and Icon are the biggest influencers by far.
- Different message copy will perform differently and the time of day has a difference especially for some offers.
- Having a proven lander/offer combination from the beginning will definitely be a huge leap forward.
- You will have issues with burnout if you are just hammering 1-2-3 offers and messages. You need to have a series of offers/messages/icons in constant rotation to slow down fatique but it will eventually happen once you or all competitors have exhausted the audience. (*Tip - Just keep the creatives and messages shocking, think more like native banner ads.)
- Some ad networks are just way over prices and have no clue wtf they are doing. Keep in mind that not only is this format of ads new for you, its also new for them. They are willing to sell inventory especially in the lower tier countries were its not always as easy to make profitable campaigns - get a rep, negotiate.
- A small test does not determine the true results. For lower payout offers, you can run a small test. For higher payout offers you need to run a big test or do several smaller tests in one day, like 2-3 times.
Other things to note:
- Paying more does not effect the quality of the traffic/users. All push users are real users, some ad networks favor certain types of traffic aka: streaming, adult, etc. So keep that in mind with your offer ideas when promoting on those networks. And pay the cheapest you can so you dont waste money.
- Getting a high CTR DOES NOT equal success in most cases. High CTR messages with low focus = no results = high cpc costs = you lose. But they do work sometimes.
- Instead, for simple conversion flows like lead gen, etc you can test broad high CTR geared messages but the results will be inconsistent, sometimes good and sometimes bad. Instead, focus on tailored messages that offer and pre-qualify the users with the message you are sending. It will result in lower opens (lower costs on cpc basis too) but higher CR and more often than not you can make it work well.
And.. Another big note about ad networks:
- You always hear these ad networks boast of "we are adding 1,2,3 millions of new subs per day/week/month or whatever. This is very short lived boasting. What I mean is, they will plateau on their new user acquisition and that growth will no longer happen. It will simple sustain the current level of churn they are getting from all the messages going out daily. And for the ad networks who do not properly manage their user base they will actually be losing more users daily than they acquire. Be careful of these networks because you will find it hard to make campaigns work at some point due to all the advertisers and the users being overly exhausted.
I was recently a speaker at the Arab Affiliate Summit in Morocco and I shared some insights and thoughts on Pop and Push traffic this year. Here are the slides I had for my speech, you are welcome to view them but they may not be available for too long. They do not convey all my message just highlight the points I talked about.
View Slides - 7-Figures w/Pop and Push
Cheers,
Andrew
Solid shar MrPayne, thanks 
Great post Mr Payne!
You always rock bro!