EXPLAIN, PLEASE: Anyone Ever Had This Issue - Why & What To Do

[Mrcheaply]

So, I finally launched a Push Notification campaign on PropellerAds using BeMob tracker and a page builder app (not WordPress)

Not exactly sure what's going on, but it's already 3 days, 3 campaign variations (optimizations) and over 500 visits, plus about 60 offer clicks, and yet I'm yet to clinch that much coveted first conversion. So unenergizing.

At the meantime, here's a more pressing issue:

1)
A huge discrepancy exists between the data for "Visit" and that for "Unique Visit" (about 100% difference. In some cases much more than 300%).

2)
There's also another huge difference between offer "Clicks' and offer "Unique Clicks" (always more than 50+%).

In both cases, the "unique" number is always abnormally lower - at least that's it looks to me, (unless you want to confirm that that's in fact the norm in your own experience).

Please, what could be causing this?

Can you explain it?

Ever experienced it? How did you resolve it?

I need your help.

 

[Graybeard]

You are getting multiple visits from the same user (IP),
a 'unique' per day means a once (or the number of times) in a 24 hr clock (time zone) day;

Try so set the unique frequency to 1 or 2 per 24 hrs in the ad network interface.

Redundancy = the number of times a user might see your ad (the push creative in this case).

Why would someone click the same push ad many times and not simply dismiss it? Discuss ... (Idiot factor or ?)

 

[Mrcheaply]

You are getting multiple visits from the same user (IP),
a 'unique' per day means a once (or the number of times) in a 24 hr clock (time zone) day;

Try so set the unique frequency to 1 or 2 per 24 hrs in the ad network interface.

Redundancy = the number of times a user might see your ad (the push creative in this case).

Why would someone click the same push ad many times and not simply dismiss it? Discuss ... (Idiot factor or ?)

Thank you @Graybeard for taking the time out to respond and provide insight.

Your explanation makes absolute sense. I remember that BeMob had that 1/24 time per visit stuff set up automatically as I was putting up the campaign. Was not sure what it meant until now. So did not want to reset.

Happy I did not. At least I know now that at least 100% of the time my ad was clicked at least twice or more by each target. This does not seem right though. After all, I set the frequency capping to 1/24hrs in PropellerAds.

Why would the multiple clicks per IP take place when the frequency capping is 1 24hrs? Is this a problem from Chrome browser? (Some times too I've noticed that my tab's Push Notifications refuse to open on the browser despite clicking it hard severally. The browser opens after each click without loading the intended page, and without prompting me of any blockade).

Or, could it be that folks were actually deliberately clicking on my ads multiple times? For what purpose? Opposition research? Malice? Competition? Dumbness?

But then, I also see the same pattern on offer clicks. Why would one click on an offer repeatedly when he/she does not want to claim it? Research? Bot?

Kind of losing confidence in PropellaAds' push traffic. Not necessarily because of the zero conversion so far, but more because of the untidy data I'm seeing.

More ideas are welcome though.

 

[awesomesauce]

I've found it normal to have quite a bit of duplicate clicks, both in the tracker and in the affiliate network. Who knows what causes this, but it's most likely people clicking on your push notification 2x. Who knows why they do it.

Clicks = all the raw clicks
Unique clicks = the number of times a unique person clicked

Clicks are higher because of people clicking 2x or more.

Dealing with untidy data is life. Your tracker's data will almost never be accurate.

 

[Mrcheaply]

Thank you, @awesomesauce for your take.

Well, I just noticed now that at least 90% of those REPEAT clicks come EXCLUSIVELY from an Unknown/Empty ZoneID. Not exactly sure why this is, but that's how it is.

Aside that, the other known ZoneIDs are just fine. The number of visits is almost about 90% similar to that of the unique visits - as it should. So I'm now looking for a way to filter out the culprit - the Empty ZoneID. Anyone with any suggestion on how I can get that done is welcome.

As at now, I've only set up the Bot Filter rule to remove any (custom1) ZoneID that's Empty or Unknown. But I'm not sure if this will work since there's no predefined parameter for both. In any case, I selected the option that offered to include "empty parameters" (or something of that nature). We will see how it goes.

Another thing is that I noticed that the rate of clicks/ engagement on the landing page is simply ..... This is despite the fact that the landing page is simple, clean (white background, bold crispy fonts, nice logo) and mobile optimized, with just 1 or a maximum of 3 sentences followed with two Yes or No very clear CTA buttons. I should be seeing at leat 35% unique CTR on the lander. But the average (on the higher traffic ZoneIDs) seems to be 10% or less. Maybe I need to kill more of these ZoneIDs. BTW, some of the tiny traffic ZoneIDs record 100% CTR.

Can't wait to test this campaign on another traffic source, though.

I've found it normal to have quite a bit of duplicate clicks, both in the tracker and in the affiliate network. Who knows what causes this, but it's most likely people clicking on your push notification 2x. Who knows why they do it.

Clicks = all the raw clicks
Unique clicks = the number of times a unique person clicked

Clicks are higher because of people clicking 2x or more.

Dealing with untidy data is life. Your tracker's data will almost never be accurate.

 

[PropellerAds]

So, I finally launched a Push Notification campaign on PropellerAds using BeMob tracker and a page builder app (not WordPress)

Not exactly sure what's going on, but it's already 3 days, 3 campaign variations (optimizations) and over 500 visits, plus about 60 offer clicks, and yet I'm yet to clinch that much coveted first conversion. So unenergizing.

At the meantime, here's a more pressing issue:

1)
A huge discrepancy exists between the data for "Visit" and that for "Unique Visit" (about 100% difference. In some cases much more than 300%).

2)
There's also another huge difference between offer "Clicks' and offer "Unique Clicks" (always more than 50+%).

In both cases, the "unique" number is always abnormally lower - at least that's it looks to me, (unless you want to confirm that that's in fact the norm in your own experience).

Please, what could be causing this?

Can you explain it?

Ever experienced it? How did you resolve it?

I need your help.

Hi!

Did you show this to the BeMob support?

 

[Graybeard]

Well, I just noticed now that at least 90% of those REPEAT clicks come EXCLUSIVELY from an Unknown/Empty ZoneID. Not exactly sure why this is, but that's how it is.

That is suspicious. I would ask for a detailed explanation ...

The ad server would(should) have all those *zombie zone's* IP addresses logged ...

 

[Mrcheaply]

Exactly my thoughts too. However, one of the support staff @PropellerAds told me that the "ZoneID was missing probably because I used Broker traffic". I'm not sure what "Broker traffic" means on their platform. I did not see an option like that while setting up the campaigns. Of course, I asked the lady for clarifications, but I'm yet to hear back from her.

That is suspicious. I would ask for a detailed explanation ...

The ad server would(should) have all those *zombie zone's* IP addresses logged ...

 

[Mrcheaply]

Hi!

Did you show this to the BeMob support?

Thank you for checking in here .

Yes I did, in fact, one of them logged into my account, reviewed the affected campaign and made some suggestions.

But regarding the discrepancy here's what they said:
__________
"Concerning discrepancy, it looks like you experience some kind of prefetching because your visits double.
Analyzing your reports, you may see that you have doubled visits in relation to unique visits.
In this case I advise you to synchronize the time zones of your BeMob and traffic source accounts. Also you may compare stats from tracker and traffic source' side and see the difference."
___________
I will soon do the two things they recommended. But I now know that those repeat clicks were coming from an unknown ZoneID (a fact I assume my friends at BeMob were not yet aware of when they made those recommendations).

Again, thank you.

 

[awesomesauce]

Filtering anything in your tracker isn't the solution, as you'll still be paying for those clicks. You just won't see them in your tracker.

Regarding your landing page... Maybe it's not as good as you think it is? There could be some kind of a disconnect between your ad and landing page causing low CTR. Don't know without seeing everything.

 

[Mrcheaply]

Okay, finally got one conversion ($3.20) after a total of $80+ spend on PropellerAds. I've duplicated and optimized this campaign 5 times. The 5th try was what got the conversion at -86% (negative) ROI. This is after about 350 ad clicks and 37 prelander CTR (to multi offers). The prelander CTR for the 5th optimization is less than 5%. I've done one major optimization on the prelander too.

To be clear, starting from the first ad campaign data on this offer, I've eliminated a couple of cold zoneids, streamlined device carriers to a few engaging options, targeted the ad on specific cities at about a total of 10 different hours of the day. The vertical is finance.

Thing is, is there anything I'm doing wrong? Is there still a meaningful way to optimize this campaign further?

@Graybeard, @awesomesauce, @PropellerAds.

 

[Mrcheaply]

Thank you @Graybeard, well the landing page has multiple offers, ranging from basic loan to unsecured bad credit loan and disability benefits, etc.

Generally, I think the right audience would be the poor folks looking for a quick bailout fund or something. That's actually the angle I'm using for the ads which I think is nicely in sync with the above the fold offer on the landing page. The rest of the offers are just mildly listed below the fold.

That being said, can you please give me more detailed insight on how to effectively track and decode the "IP group attributes"? I'm using the @BeMob tracker.

Finally, the conversion action is supposed to be one page submit (Name, email, phone).

What does 'finance' mean to you?
Loans to poor people to offers of high-end financial speculation. <<< the only thing the have in common is money
st.

What you need to know is the productivity of IP groups or identify the traffic as people, not the exact person, but the IP group's attributes.

Until you get enough data back from the landing page, both of those that bought and those who abandoned or attempted to buy then were declined for some reason -- you are just pulling on straws. And just tossing your money at it.

$3.20? what does the converted customer do for $3.20?

 

[Graybeard]

What does 'finance' mean to you?
Loans to poor people to offers of high-end financial speculation. <<< the only thing the have in common is money
st.

What you need to know is the productivity of IP groups or identify the traffic as people, not the exact person, but the IP group's attributes.

Until you get enough data back from the landing page, both of those that bought and those who abandoned or attempted to buy then were declined for some reason -- you are just pulling on straws. And just tossing your money at it.

$3.20? what does the converted customer do for $3.20?

 

[Graybeard]

OK; so you target 'customer' lives in a lower income area. If it is in the United States I can match zip codes with income distribution levels that might be better than 30% distribution of the income level medians that you are looking for.
Like 40% in this zipcode are families of 2.07 with $28K yr AGI taxpayer unit (family) income. Or, $300K if you are marketing new BMWs
I can match the IP address the traffic is using to that zip code with about 50% - 70% accuracy.

So, you need to either whitelist the traffic by zip code or IP address when you buy it; or buy a lot of traffic and sort it to various demographics and put offers that will be of interest to that demographic (hopefully the right ones) in front of those people. Note I said PEOPLE -- traffic buys nothing People do.

BeMob or any other tracker that I have seen cannot do this.

You need to have a traffic buying method or a traffic funnel set up (or both).

So-- what were all the IPs Seen so far? And the IPs seen from the most current campaign. How big is the sample? 1000 IPs or 10,000.

I have to program something like this for my own use anyway That's my motivation to test the principle against actual data.

Have you found a way to reach the right people are are you just slapping ads on the Internet and hoping for the best?