Home > Paid Traffic Sources > Mobile

40% Click Loss - Is There A Solution? (14)

01-11-2015 01:13 PM #1 priyankcmp (Member)
40% Click Loss - Is There A Solution?

Hi,

I am testing 2 apps (clickdealer) with pop traffic. But I am seeing huge click loss:

Traffic to Voluum - 15%
Voluum to offers - 25%
Total 40%

Is that normal with pop traffic?

At the moment, it's direct linking.

Do you guys think that having prelander will help in reducing such heavy click loss?

Any help is appreciated.

Regards,
Priyank.

01-11-2015 02:10 PM #2 integrity (Member)

Yes, this is absolutely normal, especially when linking to the Google Play Store.

I've tested some network links on my mobile phones and many times between the banner/landing page click and when the Google Play Store starts to appear, the screen is white for a few seconds. I can only assume it is much worse in other countries with slower mobile internet.

So what you're probably seeing is the visitor hitting back on their browser before the network tracking link loads, since they really have no idea what they're waiting for or if that white screen will eventually load into something else.

My advice is to use a lander (that loads quickly) to bridge that gap and let users know what they will be waiting for on the other end.

01-11-2015 02:34 PM #3 priyankcmp (Member)

Quote Originally Posted by integrity View Post
Yes, this is absolutely normal, especially when linking to the Google Play Store.

I've tested some network links on my mobile phones and many times between the banner/landing page click and when the Google Play Store starts to appear, the screen is white for a few seconds. I can only assume it is much worse in other countries with slower mobile internet.

So what you're probably seeing is the visitor hitting back on their browser before the network tracking link loads, since they really have no idea what they're waiting for or if that white screen will eventually load into something else.

My advice is to use a lander (that loads quickly) to bridge that gap and let users know what they will be waiting for on the other end.
Hey,

Thanks for the help.

Yeah I thought the same that using lander will help.

My first test was to see which browser people are using because one of the two apps I am running is browser app. And thankfully, only 2% people are using DU browser in the targeted country.

Thanks.

01-11-2015 02:52 PM #4 elangokp (AMC Alumnus)

Hey Priyank,

Check the unique IP count in Voluum. The networks seem to only count the number of unique IPs hitting their tracking link.

01-11-2015 04:17 PM #5 priyankcmp (Member)

Quote Originally Posted by elangokp View Post
Hey Priyank,

Check the unique IP count in Voluum. The networks seem to only count the number of unique IPs hitting their tracking link.
Hey,

I can see getting up to 4-5 visits from some of the IPs but I am running this campaign for last 4 days with frequency of 1 view / 24 hours.

However, I got your point.

To overcome this, do you think increasing freq cap to something like 1 view / 48 to 72 hours because I think that would kill the volume.

I am just trying to collect titbits which I can implement in next 2 days because I am seeing some promise in offers I am running. But this click loss is really discouraging.

Thanks.

01-11-2015 05:55 PM #6 elangokp (AMC Alumnus)

Traffic to Voluum is the actual click loss you are experiencing.
Voluum to Offers, it should not be click loss, but just the difference in the way the count is shown.
Network shows only unique IP hits. But that doesnt mean if the same IP hits 20 times, they just show the offer 1 time.
They would show the offer all 20 times. But voluum would count it as 20 hits to offer, while network would count it as 1 hit to offer.

To test this theory, export the unique IP list of the campaign from voluum.
Now compare the number of rows(one IP per row) with the network count. it should match.

Even if you set the frequency to 1 visit per 10 days. If the visits comes from a previously visited IP, the count will not increase.

But, the catch here is that usually mobile users will be turn on and off their data frequently. So, most will get unique IPs frequently.

Still, if there are many visits from same IP. Then there are 2 possibilities.

1. Bot traffic from same IP
2. Dynamic IP allocation. (The same IP which was previously assigned to one mobile, would have been allocated to another mobile during turn on and off of data. If these two visitors have reached the offer page, the visit and conversion attribution will likely go to the first visit. Still you will not miss conversion or visits. The attribution is what will be messed up)

This is in large theory, but seems to be true to me.

01-12-2015 03:41 AM #7 priyankcmp (Member)

Quote Originally Posted by elangokp View Post
Traffic to Voluum is the actual click loss you are experiencing.
Voluum to Offers, it should not be click loss, but just the difference in the way the count is shown.
Network shows only unique IP hits. But that doesnt mean if the same IP hits 20 times, they just show the offer 1 time.
They would show the offer all 20 times. But voluum would count it as 20 hits to offer, while network would count it as 1 hit to offer.

To test this theory, export the unique IP list of the campaign from voluum.
Now compare the number of rows(one IP per row) with the network count. it should match.

Even if you set the frequency to 1 visit per 10 days. If the visits comes from a previously visited IP, the count will not increase.

But, the catch here is that usually mobile users will be turn on and off their data frequently. So, most will get unique IPs frequently.

Still, if there are many visits from same IP. Then there are 2 possibilities.

1. Bot traffic from same IP
2. Dynamic IP allocation. (The same IP which was previously assigned to one mobile, would have been allocated to another mobile during turn on and off of data. If these two visitors have reached the offer page, the visit and conversion attribution will likely go to the first visit. Still you will not miss conversion or visits. The attribution is what will be messed up)

This is in large theory, but seems to be true to me.
Hey elangokp,

Very well explained mate. Thanks.

I downloaded IP list and it's showing 18200 unique IPs whereas network is showing 14900 clicks. So that's still loss of 18-20% from voluum to network.

Thanks.

01-12-2015 05:08 AM #8 elangokp (AMC Alumnus)

You are welcome. Did you check for the same duration (or better whole duration of the campaign) . Also check if there are any international redirects that is happening.

01-12-2015 07:50 AM #9 priyankcmp (Member)

Quote Originally Posted by elangokp View Post
You are welcome. Did you check for the same duration (or better whole duration of the campaign) . Also check if there are any international redirects that is happening.
Yeah I have checked in same duration. And I have counted clicks going to default international offer.

Will be running this with lander in next 24-48 hours once my landers are ready. That will give me clear Idea.

Do you think running in GEO like India can be a problem because of slow internet speed?

Regards,
Priyank.

01-12-2015 08:00 AM #10 bbrock32 (Administrator)

Most networks pay the tracking platform per click so that's why only unique IP clicks are counted.

What I would do is run your page through a load test tool and if that's fine I wouldn't worry much about the discrepancy with the network.

01-13-2015 07:35 AM #11 satori (Member)

I've experienced even higher click loss with pops.

01-15-2015 02:42 AM #12 zeno (Administrator)

Quote Originally Posted by elangokp View Post
Traffic to Voluum is the actual click loss you are experiencing.
Voluum to Offers, it should not be click loss, but just the difference in the way the count is shown.
You will absolutely get click loss from Voluum to the offers.

Remember that it's the act of a user's device loading a URL in each case.

Voluum load balances to something like 8 server locations around the world. Virtually all networks go to one or two.

Ergo, the inherent click loss you will get from user connectivity and drop-off from waiting will generally be higher from Voluum > Network than from Traffic source > Voluum, though this depends on the traffic source as the way the traffic gets sent in will alter the 'true' number of clicks that you actually get (e.g. banner clicks will be more reliable than pops which may get blocked and aren't associated with a user action, traffic source may overcharge, etc.).

01-15-2015 05:34 AM #13 elangokp (AMC Alumnus)

Quote Originally Posted by zeno View Post
You will absolutely get click loss from Voluum to the offers.

Remember that it's the act of a user's device loading a URL in each case.

Voluum load balances to something like 8 server locations around the world. Virtually all networks go to one or two.

Ergo, the inherent click loss you will get from user connectivity and drop-off from waiting will generally be higher from Voluum > Network than from Traffic source > Voluum, though this depends on the traffic source as the way the traffic gets sent in will alter the 'true' number of clicks that you actually get (e.g. banner clicks will be more reliable than pops which may get blocked and aren't associated with a user action, traffic source may overcharge, etc.).
Yes, that is true, but what percentage do you think the click loss will be from Voluum to Offer?

In my case i found something like 40% to 50% click loss, but then i found out that the unique IP count was closely matching the visit count in the network.

Then came to know that networks only count unique IP clicks. (very bad for measuring conversion rates. This factor alone is enough for messing up the EPCs shared by the networks, putting aside other factors like traffic quality from different affiliates)

If we are facing a loss between the unique IP count and network visit count. Then its the genuine click loss.

what do you think?

01-18-2015 02:26 AM #14 zeno (Administrator)

If your click count in Voluum matched up nicely with the gross click count (including dupes!) at the network, then you're very lucky.

This usually only happens when your users are very close to your affiliate network tracking servers and have excellent connectivity, e.g. in US/CA and Western Europe.

Don't worry about how the networks count clicks and work out EPCs, how they get charged on the back-end, etc. Just focus on gross clicks coming out of your tracker and gross clicks recorded at the network.

Use this to assess click loss. And confirm with testing that the same IP can convert twice, then you're golden!

Note: it's well known that mobile carriers aggregate users to the same IP, so for an app install, you can get many conversions on 1 IP - the reporting shows these as non-unique but device identifiers confirm that they are indeed unique, so can convert independently. One would hope.

Home > Paid Traffic Sources > Mobile