Anyone notice a huge amount of click discrepancies on some mobile networks?
It seems I’m getting invisible clicks that my tracking is registering. Is anyone getting this problem?
Had a huge problem with Admobile. Ran 2 campaigns into Australia, got 100s of clicks, but about 40% were coming from the same IP address.
When I questioned them about it – all they could say was that the phones must have been roaming to the same IP address – yeah right.
That was my first and last mobile campaign.
Im not noticing it on some of the top tier ones like Jumptap. Right now Adfonic is giving me around 50% of clicks that’s not registering…
im glad people besides myself are commenting on this.
I have brought this up before and it gets ignored with the comment "I know x is making bank"
I have heard that jumptap is better than ad mob. admob was the one that I had the most ‘fraud clicks’
so, do we need to test and report each network here ?
I think this is a good idea….there will be some changes coming soon to the forums……I believe a mobile section is in the works……due to the nature of mobile traffic, will prob have a big sticky at the top covering these exact points…..and more…
maybe even with some tools to help in working out the issues with mobile…….stay tuned
Yeah, I had an issue with Jumptap where a lot of clicks didn’t register on a campaign I tested few days back.
Thanks for the info guys, was about to give them a shot tomorrow… Will rethink it now or be conservative with my testing.
Anybody have anything good to say about jumptap or admob?
i’ve been doing a bit of testing on mobile. here’s my take on some of the ad networks:
mojiva: lots of traffic, but conversions sucked.
adfonic: click discrepancies around 50% a lot of the times (i.e., i only see half the clicks they’re charging me). i also got some (unwanted) international clicks.
admoda: pretty good traffic, but not much volume.
decktrade: pretty good traffic, but can’t seem to get much volume lately.
buzzcity: pretty good traffic, but not much volume for u.s and targeting isn’t as robust as say admob.
inmobi: haven’t tested much lately, because their new interface sucks. heard they have a lot of international traffic.
jumptap: good traffic. i see click discrepancies around 20-40% depending on carrier. e.g., with metropcs, i usually don’t see discrepancies, but with alltel or tmobile, it’s about 30%.
admob: good traffic. good thing about admob is that android handset usage is growing and if you target just androids your clicks are a little cheaper…but you also get a lot of shitty app traffic…and there doesn’t seem to be an option to exclude app traffic.
all in all i think jumptap and admob is the way to go. but i think, really, the way to get mobile campaigns very profitable is to get an ad rep at those networks. that way they can exclude/blacklist sites for you. in almost every case there were a handful of sites that just ate up my budget without a single conversion. and there was no way to get those campaigns profitable without blocking those shitty sites.
Wow thanks for that, that deserves a double thanks. Saves a lot of people time and money.
I’m only working with Jumptap now, and I don’t have any problems.
Just my 5cents to this. I was concentrating on mobile for more than a year now. All the mentioned issues sound very familiar to me. After a lot of arguments with almost all major networks I have to come very reluctantly to the conclusion, that in most cases it is not clickfraud that happens there.
Because of me spending quite a bit of money with them and arguing very hard, they actually provided me with clicklogs to prove that these clicks have happened. The problem here is that the tracking is by far not as good as on the net. It is very common, that a click that comes from let’s say admob and is a legitimate one will be filtered out at the advertiser’s end ( mostly by the cpa network), because it does not match their direct requirements,for example you have done all your targeting right,operating system,carrier,even handset and then their system detects that this customer has a prepaid plan,they might filter your click out.You lose.
Therefore it will not show up in your dashboard. Another problem is, that in many cases a carrier(metro,cricket,etc.)will put several handsets on one IP address.So it might look like you are getting the same one over and over.But they can be all individual users.And the last problem is, (what that far noone can really explain) that clicks simply disappear.
Several networks have proven to me that these clicks have actually happened at their end and even the big boys have not found a satisfactory solution to this problem. They call it "breakage". It can be different from country to country. Where I live, in Australia it is the worst, up to 85% at some times.
excellent post dieter.
this makes a lot more sense then just ‘clickfraud’
I do say that some of that is happening in that many people do click ads that are poorly places (near navigation zones for apps) and that will come across as a ‘legit’ click but since the end user did not mean to do that action. not really a ‘good’ click
the other one is with CPL. I have listened to actual recordings of the callers and heard how many in a block of an hour were ‘butt dialing’ [fun conversations you hear at times] or some child playing with parents phone [or some really baby sounded adults having fun]
so add those to the list you mentioned.
Now, that we know the issues.
How do we ‘clean it up’ and make it workable/profitable ?
Yeah, just to follow up with what Dieter is saying, we see on average a 30%-40% click breakage on our offers, and that is with the Server-to-Server tracking we do with our advertisers. We found that that is MUCH MORE ACURATE than regular advertiser pixel tracking.
This is due to the IP issue that Dieter was talking about plus the fact that many handsets will not render certain banners (IE/ some feature phones cannot display image banners). This is something we are just getting used to with mobile traffic.
Aside: we do NOT lose conversions with S2S tracking. Just clicks.
totally awesome share JAYON
not lets figure out with the ‘slack’ how to make it profitable !