Home > Paid Traffic Sources > POP / PPV / Redirect

---

We deliver much less Push opt-in prompts than popunders we pay for. Help please! (10)

---

Hi guys!

We're running one of our pops campaigns with the push list building code included in it.

Today, when looking at the analytics report of our push list building platform, we realized something bad:

the amount of opt-in native prompts that we are delivering are much, much lower than the popunders we are paying for.

Some numbers we got:

Traffic source 1: We paid for 5,000 popunders, our push platform analytics shows only 1,000 prompts

Traffic source 2: We bought 30,000 popunders, our push platforms shows only 5,000 prompts.

In theory or in a perfect world, we would have delivered the same amount of opt-in prompts as pops we paid for. I was expecting some difference between the 2, since people close the pops before they load... but this seems way too high

In theory our landing pages load fast, with our loading times being around 300ms according to pingdom tools.

This is a big issue since the main goal of this campaign is to collect push subscribers.

Questions:

1) Would the lower amount of opt-in prompts be an issue caused due to them appearing too late so people close the pop before they appear?

2) Or are there other potential areas where we might have misconfigured something (we took iphones out of the targeting, and no desktop either).

3) Is there any technical implementations we could do to check why is this happening or where is the leak?

4) Any tricks to make the opt-in prompt appear (a lot) faster?

Thanks in advance!!

---

I would say most of your problems are related to 2 things:

1. Clickloss: part of the traffic will get lost from a number of reasons and POPs are notorious for this. I wrote a guide on this, so won't repeat myself, read it here: https://stmforum.com/forum/showthrea...finitive-guide

2. PUSH prompts will not fire on all OS's or devices: Try to check what % of successful prompts you see on various devices and operation systems. I'm pretty sure you will see that there is a bigger loss on some. Focus on traffic that gives you the best results. I see that you do not target iPhones, but there might still be some in the mix and some browsers can block these prompts too.

Then there are other factors like BOTs that naturally cant see the prompts, users on slow connections... this will lower the success rate further.

What solution are you using to collect the push subscribers btw? There might be some issues with it too.

---

Hey Matuloo! Thank you so much for stopping by to help

I just read your thread, really helpful!

1. About clickloss:

The numbers of the reported pops we bought are from the tracker, so for example, in the Traffic Source 1, those 5000 popunders we bought, is according to Binom (so clickloss between the traffic source and the tracker is taken out of the equation).
I guess related to click loss we still have these problems:

A) smart bots who pass through the redirect rules but then don't get the prompt
B) fast fingers who aren't fast enough to stop the impression from showing at the tracker, but are fast enough to close the landing page before the native prompt has loaded

In general I would expect a lot of case B happening. But damn! 80% of the impressions not loading the prompt fast enough is a bit sad, lol.
It specially sucks because I believe showing the prompt immediately would help a lot with the opt-in rate, since a lot of people seem to click "allow" automatically as they get the prompt, without thinking much.

2. About devices, OS, browsers:

As of now we don't have a way of seeing which devices or OS are delivering successfully the opt-in prompts, since that would need to be reported by the push platform and they don't seem to offer that level of detail.
BUT what we can do is to check in Binom side if there's some devices/OS/browsers/etc. with a dramatically lower proportion of opt-ins, versus the others, for the amount of impressions they got. That would be a symptom of less opt-in prompts being delivered.
Unfortunately after spending some time doing so, I cannot find any anomalies or money drains over there.

Which push solution:
We are using Pushengage, have you heard anything about them?

In general everything points to the fact of the native prompt not loading fast enough, at least fast enough before we show something that engages the user, so he has an incentive not to close the pop. In that case the only solutions I can think of are:

1) Asking pushengage support where are their servers hosted. If their reply something that might be an issue, it would be worth to split test other services.

2) Making the opt-in load before everything else, is there any trick for this? Lots of people just click "allow" automatically before thinking much, so that would help.

3) Making our landing pages show something "catchy" fast. Jono Alderson talks a bit about this here: https://www.youtube.com/watch?v=S8_a...ature=youtu.be
Certainly there's room for improvement here, but unfortunately, our options here are limited since like I said, our main goal with this campaign is to get push subs. You can imagine what type of landing pages we are using

4) Building my own push solution with optimized hosting so the prompt loads faster? Unfortunately this sounds out of budget right now, I have heard some high prices for custom solutions. However if someone reading this is a developer and can do it for much less please hit me up

That's all I can think of for now. Thanks again for helping!

---

Just one question to clarify something: when talking about the loss, is this between the amount of traffic recorded by Binom and the amount of "optin popups" shown? Or are we talking about successful subscribes?

One more tip: this could also be related to the GEO you run in, the loss is very high in some less developed countries, in Asia for example, this can be huge.

The first thing I would do, would actually be to run a split test with another service. There are several providers of this service out there, so try to run a test with some other one, to make sure the problem isn't related to pushengage.

What is actually the page where you are collecting the subscribers like? Is it a full "website" that has to load fully and then the prompt jumps up? If that's the case, you could try using a very simple page, maybe just some "warning" or entry note with some minor graphics to catch the eye, so it loads fast and doesn't delay the push prompt.

---

Yeah, the loss happens between the amount of traffic recorded by Binom, and the amount of "optin popups" shown, not the subscribers.

So Binom would say we bought 30k popups, but the push platform would say only 5k people received the optin popup.

Yeah, I agree that's the first test we should do

The landing page we were using is one of those "loading" landers with a few elements (i like to call them the ghost landers :P). I think there's definitely some split tests to be done on this end, to find out if some simple element would catch more attention and therefore give more time for the optin prompt to show up before the user closes the window.

---

Ok clear, just wanted to make sure we're not talking about conversions as that would totally change things

I would really run a split test with some other service before doing anything else.

You could probably improve things with a faster loading lander, but the most important part is how fast the optin pop shows. Is there some setting where you can modify this? I've seen some platform where I could actually modify the time that needs to pass before the pop shows, this could have a big impact. So the most important question: does the pop show maybe only after the LP is loaded in full? In case it is, making a faster loading LP would definitely help.

I'm looking at onesignal config now and looks like there are even a few types of the optin prompts, some require https which could further influence the success rate. Maybe try to look at such options within pushengage too.

---

Definitely! Will look at how it all really works when it comes to optin prompts and if i find answers to some of the questions we threw in here will come and update

Thanks!

---

Let me know what you find out, I'm really curious about the cause of this problem.

---

Hi,
If you want to increase your push subscribers list (which is what I'm doing right now - pretty the same as you) make sure to target Android and chrome browsers -
and devices - phone, computer, tablet.

I did the same mistake and worse when I tried to target Iphone users (they are not supporting push notifications at all).

---

Yup, this would definitely have an impact, but the OP stated that they took iphones out of the targeting, so I guess they didn't target ipads either.

But would be nice to get an update here, I'm quite interested about the end of this too

---

Home > Paid Traffic Sources > POP / PPV / Redirect