Adding PUSH to native funnels to increase ROI - Let's see what happens! (22)
Adding PUSH to native funnels to increase ROI - Let's see what happens!
Background: I work with a small team promoting various verticals on native platforms. What I've found is that profits are relatively hard to crack and maintain in every vertical - we pretty much have to have the right ads, best lander, top offers, and great placements to approach the green zone. Even then its hard to keep campaigns above water and profitable.
So now I'm focusing on bumping up the roi on the backend, and I've decided to focus on PUSH.
In this follow along I want to see what kind of increases in roi that we can get, what are the best practices, is it all worth it, etc. I'm challenging myself to stick to it for 30 days.
I'll be focusing on 3 verticals that we have running:
Mainstream dating
Solar
Diabetes
All of these verticals we optimized to profitability at one time or another...but that mostly depends on how the top offer is performing.
I'll try to share as much info as possible here. Any inputs/suggestions etc are greatly appreciated
hey mate,
out of curiosity: what geos are you guys running?
good luck! Looking forward to updates
Originally Posted by alfo1324
I like geos with higher competition since there are more offers available. Also less quality problems.
So what exactly are you about to try? To add a push notification subscribe pop, to build a list and try to monetize it down the road?
Or do you have any other setup in mind?
Originally Posted by matuloo
Ah just realized there are 2 "jacks" playing with the same idea at the same time. Here is the other thread if anyone wants to have a look: https://stmforum.com/forum/showthrea...ative-lp-epc-s
Did you already set this up, would like to see what your initial results looked like. I'm pretty sure the initial CVR will go down a bit, but the follow up sequence could make up for it. BTW: based on the info I have from some guys who work with push, the value of the subscriber goes down with time, so make sure you have some kind of sequence ready from the start as you need to start monetizing them straight away.
Still setting push up....
I signed up with Push engage. I started the setup process but they suspended my account because my main domain url wasn't resolving. ugh.
For the push sign up box there are a few options. There is 'single step,' generic-looking opt-in. This one you can't customize at all. In Chrome browsers it just say "[website name] wants to show notifications" Then the buttons are ALLOW / BLOCK.
The 2-step sign up box is very customizable and a leaves us marketers a lot of options to encourage users to sign up.
HOWEVER, pushengage recommends the one-step form, and I asked a few people and they all agree. Seems like there maybe a big drop-off in the subscriber rate using two step.
What I did notice from the single opt-in box and from all the push-spam that I'm receiving is this: the domain name [aka sender] seems like a good space to not only have as a brand, but also to convey a message in itself.
Here are examples of some diabetes related push optin +messages:
Looking at those, I feel like a .com is not necessary at all since a push-er doesn't feel like a regular site. I also noticed that a lot of pushes come from a short, good domain name from one of those unique extensions:
.news
.blog
.guru
.expert
etc...
I decided to start the push project with our diabetes funnel. So I picked up a (IMO) pretty sweet diabetes.[extension] domain name last night. Today I'm setting up ssl for it since single optin requires https.
Our landing pages will also have to be on the domain name in order to send out pushes.
I contacted pushengage but if they don't unsuspend my account relatively soon I'll just start a new one haha
Here are my initial results from collecting push traffic.
Prior to this push test, I had about 4-5 lp's in my diabetes funnel that I had run and tested out.
I took the top performing lp, made a copy, and didn't change anything to this copy except add push opt-ins. I then sent half of my diabetes traffic to each of these two pages.
At first I made a mistake and left "subscription overlay" on. This immediately triggered a near-blackout on the lp and encouraged the user to accept the opt-in.
It sorta looked like this:
With this lp I got 10 subscriptions at about a 10% subscription rate. Meaning 10% of my visitors from Native and (other) push sources clicked "ALLOW".
However, during this run the lp ctr tested at 61% (lp-WITHOUT-push-optin) to 46% (lp-WITH-push-optin).
I figured this was too big of a drop in lpctr so I turned off the subscription overlay feature and increased the delay of the opti-in box from 0 to 2 seconds.
Here are the landing page results over the next week:
To me the test results look good. The "drop" in lp-ctr is statistically insignificant. Interestingly, the lp with the push notifications optin received a higher sales rate (7 vs 2) but this can be totally random. ...Or maybe the user is more engaged to watch the VSL after clicking "ALLOW" ?? Who knows...
As soon as I started collecting push sign ups, I made it a habit to send one push notification as the first action item of every work day....so since I'm a total night owl most of these pushes were sent early afternoon. (I'm US based and so is this audience)
I tried a mix of dl and lp destinations. I am tracking these sent pushes as a separate traffic source in
***note to self - add some way to track individual push notifications...probably as a custom variable***
Here are the results at push engage:
What is not depicted is the number of unsubscribers: 6
So this 2nd version of the push-collecting lp is getting 5.3% optin. Not bad I guess.
Out of the 39 push-visitors, we got no sales (yet).
Conclusion:
IMO the optin rate looks fine. The best news probably is that the lp-ctr virtually DID NOT drop on this second push-collecting lp, so I will move 100% of my initial diabetes traffic to collect push.
I won't make any new versions of push-collecting lp's for now. I just want to collect more optins and see how this subscriber list performs.
I'll just continue sending out different messages and repeat the better performing ones.
I'll also add push collecting to the other funnels and will share results
PS: the lp ctr dropped a lot between the first run and the second run because the sources of the traffic are not stable and perform differently day to day.
PPS: just about all of this initial diabetes traffic is desktop. Its's 75% Chrome users
Nice- very interesting man!
I will admit my Push experiment kind of dropped off before I got much data...
Need to try again and see what happens!
I did sign up for some push notifications on competitors' lp's though so that was interesting seeing what they're sending. Seems like mostly links to their existing native funnels or else direct links to various ecom and lead-gen offers.
Please keep updating this thread with your results going forward!
I'm a newbie, this thread give me a lot of helpful information, thank you a lot.
Hope to see your update, I will follow it
Here are results after week 2:
The current diabetes lp that is collecting opt-ins saw 5,468 visitors so far. These have been media buys on push and native sources.
The total subscribers from this lp are 205. So the Opt-in rate is 3.7%.
Unsubscribers have been 14.
I sent out 15 messages total. I stopped direct linking to vsl's and 6 of the last messages linked to the same lp that they originally signed up on. The idea was to try to get them to watch the vsl again. Good idea/bad idea ?? I don't know...haha
But here are results from these 6 messages: 145 clicks, 45 lp-clicks. So the lp-click-through rate is 30%
For comparison purposes, this lp has a lp-ctr of 60% from visitors that came from white-list native sources. White-list push media buy visitors do about 10% lp-ctr.
So at first glance, the traffic quality from these push subscribers (30% lp-ctr) seems to be somewhere between native and generic-push media buys.
0 sales so far. I think. We did get a phantom $35 sale of one diabetes offer during the time frame, but the click didn't track correctly.
The lp-ctr from my push list has been decreasing, as expected, since some of these people are just seeing the same page over and over. I will be adding new lp's and offers into rotation.
I don't want to get too spammy so I also started sending out info about recent diabetes research. It's only been one message so far - I just direct linked it to an outside medical article. I didn't even run it through my tracker.
What else...
So far the push-ad-%ctr has been up and down. I havn't seen any clear patterns.
I am currently on the $29/mo business plan at pushengage.
Originally Posted by matuloo
Was actually something that Caurmen(RIP) had mentioned testing and speaking on some time ago - adding opt-ins for emails before the offer didn't seem to negatively impact performance, which gave me the idea to test it.
Opt-in rate was also around ~10%.
If I had to guess, I'd assume people just automatically click the accept/ok button out of habit as it's a system message prompt and still something new to many users. Plus the user has already interacted with your creative by clicking on your ad, so they're more committed to following through, unlike interrupt based marketing like say pops or redirects where I could see adding such a prompt negatively impacting LP CTR by triggering it right away.
Week 3 summary:
More members, more pushes, more clicks, no sales.
Total members are 273 +29 who unsubscribed.
The ad click rate has been slightly lower recently but I suspect that it largely matters on the quality of the ad. I try to include a 'large' image in the push and emoticons in the headline.
These are all still from the diabetes funnel.
It's been almost 3 months of sending push's to my diabetes list and results havn't been great. Here's the basic data:
I sent notifications for Diabetes-related clickbank info products, diabetes nutra, and medical insurance plans. The only revenue that we got was from the medical insurance plans, and those were cpl. Revenue = $4.50 total
Not very promising...
I still plan on adding another list for mainstream dating campaigns. Perhaps that'll be better. If not, I may call off the push experiment, as it seems to be a better use of my time working on ads, lp's, offers, etc.
I wouldn't be too discouraged that is still a relatively tiny list of users. Keep building and testing new offers.
I tested something similiar and stopped it.
Sure it's nice to add push subscribers on top for "free" (well it's probably not really free as there are opportunity costs).
But then you have to send a lot of push notifications to get something tiny back.
I had low CTR rates (2-3%). So even 1000 subscribers generate maybe 25 clicks which would cost about 1$ when buying the push notifications on a network.
Of course there are others who are very successful with this, but it didn't become a gold mine for me 
I want to close out this thread.
So this experiment was a bit of a dud.
I ran some more push-friendly ads [mainstream dating] to my push [diabetes] list and results have been bleak here too. Few clicks, even less lp-clicks, and even less signups (0)
Just like with the diabetes push ads, the quality of the traffic seems pretty weak, compared to native ad buys.
My internet is slow here and I can't sign into Pushengage, but here are the approx stats:
Total push optins - about 1000
Total ads sent - about 50
Total revenue - About $5
So the "ROI Increase" that I was seeking for my original media buys is way less than 1%. Overall that's not worth pursuing further because I can catch an extra 10-20% roi by just testing more ads or keep optimizing the lander.
So that's what I'll do...stick to optimizng the native ad buy funnels in other ways [landers, pics, headlines, offers, sources, etc]
I'm quite surprised by the super low revenue... $5 from 1000 push subscriptions is very low indeed. I know that building push lists is about huge volumes, so 1000 isn't really a big list by any means, but still, I'd expect more.
I understand you probably don't want to spend much time on this thread anymore, but could you please tell me how many ads you actually sent out? Maybe you were not aggressive enough with the promo?
I'm surprised by that too, 50 ads sent out for $5 seems whack. Could be the niche and native for sure though, I know some people doing well adding push to their adult buys and then just heavily pushing more dating/games to it.
Also like matuloo mentioned, the big money in push is from uber big lists.
Anyway appreciate you taking the effort, not all experiments work out as expected this game always bends my thinking
Hey man, It's not always working at first. Maybe you got a group of users that could be saturated. Maybe think of a different angle to target these people. I wouldn't give up so soon, you got 1K users in hand, it's still something that you can make the most of.
Home > Push Traffic >