This is a bit of a rant, and a request to the great internet for answers – if you have no interest in, or knowledge of, Facebook Advertising and/or Google Analytics, you can move on.
OK, anyone left? Probably not, but I’ll press on.
In a nutshell, Facebook says my ads are getting way more clicks than my website reports seeing (via Google Analytics). In a perfect world, the FB Ad clicks and Google Analytics visitors should be nearly the same. In my world, they are different by around 85%! In other words, Facebook says that 10 people have clicked on my ad, and they charge me for those 10 clicks, but my website has seen only one or two visitors!
For more details, and to offer any insights as to why this is happening, please visit http://forum.developers.facebook.com/viewtopic.php?id=49767
And now, back to our regularly scheduled programming…
I’ve noticed a similar thing when comparing click-thrus to a client site from a referring advertising site. We were getting charged for way more click-thrus than appeared in Google Analytics. I had the advertising site send me the raw apache logs so I could do some sleuthing.
1) One of the big differences was that analytics doesn’t report robots/crawlers – not sure if that applies to FB ads.
2) Another one is if Javascript is disabled analytics doesn’t work but that’s probably pretty rare – although probably true for crawlers.
3) But still others that looked like legitimate browsers/people weren’t appearing in the analytics report. Looks like Analytics is not completely accurate and under reports visitors. It’s better to analyze your own server logs. Be interested to see if there’s more research on this.
Don’t have hard numbers to report because it was only a quick look. When I have some more time
I’ll make a more detailed report.
Hi, thanks for this post.
Have the same problem, around 80% of the clicks don’t show up in GA.
It seems like a redirecting problem of FB, so GA is not any more able to declare it as FB traffic. Check:
Keep in mind that if you target your ad to your facebook page (and not to an external URL), there’s a ‘Like’ button within your ad. Any click on it is counted for CPC purposes, but the user does not get to your page (in FB) just by doing that.
I just contacted Facebook about the same problem. My Google analytics AND my own server logs showed 21 hits came from facebook.com, 18 from apps.facebook.com, , and that only 10 of them were from the state my fb ad targeted. Facebook billed me for 85 hits. In addition, my own stats covered a date span of March 13-23, even though I didn’t take out my ad until March 16. I thought that was being pretty generous.
Question based on a comment… How would robots click on my fb ad and go to my site?
I think I figured out part of it… the analytics shows front page hits unless you go look at an individual page (on website stats, too). Total hits to my site was 85 (same number I was billed) but that numer included direct referrals and all other referrals.
Could it be that facebook is getting numbers from our own analytics and not tracking facebook ad clicks? Just seems weird that the numbers match up…
tim b.
thanks for the info
Similar situation we faced here (but not FB Ad actually).
We used an Ad network to promote our site.
their click report shows 5-10 times higher than GA visits report. (avg 200-300 click a day vs 20 – 70 visits a day)
The visit report from server log stats shows slightly lower number than click report.
we ask if there’s any robot they used to click the Ad .
Their respond: We don’t use any robot, if we used robot the click report will really-really big (10k clicks or more a day).
Now, we still check if there’s any other reason of this problem.
any idea?
Dear friend(s) from FB…
GOOD ARTICLE
i have three (3) stat checkers… web stat, webstatseo, en GA…
all they have 50/60 % less click than FB make believe.
In my add i give product for free!!! the click i do receive i have a conversion of 20% !!!!! my feeling is: stolen money!!!
how can we cat a finger behind this?
people are not all so smart to check the big boys with the toys…
is here a control system (internet POLICE??????)
THIS KIND OF BOYS CANT RUN AWAY THAT EASY….
Krys
I had exactly the same problem last time I used FB. My google analytic would show 50 entries to my site from FB while FB would bill me for 200 of them. I emailed them and they responded that Google Analytic was not accurate and they could not help me with 3rd party tracking. In my opinion they steal money and make up clicks to charge us more. I’m planning to investigate it for the next couple months. If I can prove it to them, I will sue.
What are the chances of Google Analytics being wrong?
frank
At least part of the problem would be that many FB users have HTTPS enabled. So, the FB pages they are viewing would be HTTPS. If these HTTPS users click out from an FB Ad to your sites the referrer information would not be passed to GA (or your web logs) unless you’re using HTTPS landing pages. These guys would look like direct clicks.
If you use GA tracking codes in your ads landing page URLs you could work around the HTTPS issue.
I have more clicks from facebook then all visits on GA. They are cheating but how to prove it?
I’ve got one even worse. New FB ad directs to an offer page on my site. FB says ~200 clicks. GA show ZERO. This is my 3rd FB ad and the largest discrepancy. My hosting provider is discountasp.net, so I enabled logs and smarterstats feature.
Turns out my hit count on new offer page is higher than ad clicks reported (not by much, but not ZERO as reported in GA). And my total referrals from FB is pretty close to my total ad clicks reported by FB.
So in my case it looks like GA is wrong and FB is correct. I’m new to both these tools, so it’s completely possible I am the cause of the issue. Our site is ASP.Net 4.0 C# web forms using master pages so the GA code sits in one place – the master page.
Gives me a little peace of mind that I’m not being robbed by FB!
g
I think your GA code is wrong – 0 GA visits suggests a problem – you have visited your own page, right? GA should show that visit (though it may take a few hours). Get that fixed, then let us know how your FB and GA traffic compares.
We are noticing the same thing. They can say GA is inaccurate all they want, which it rarely is, but server logs do not lie.