Discrepancy between FB PPC and Google Analytics

[aj613]

This question was answered in 2011 here: http://moz.com/community/q/facebook-ppc-number-of-clicks-according-to-fb-different-than-visits-in-analytics.

Wanted to post it to the community to see if anyone had any new thoughts in the last 3 years.

I have been running campaigns on Facebook and seeing dramatic discrepancies between Facebook and GA's numbers.  For example, I ran a Facebook ad campaign for a chiropractor where FB shows 35 clicks to the website, but GA only shows 2!

An attorney ran a Facbook promotion, got 4 clients who actually filled out a questionnaire online, but GA only showed 2 visitors exiting off the form completion page.

Is this because the users did not have JS/cookies enabled?  Something else?

What is the recommended work around?  Tracking URL?

[JasmineA]

The main discrepancy is going to still be the one specified in 2011: The attribution method of facebook & google tracking clicks.

FB is going to use first click attribution, GA is going to use last click attribution. Have you looked at the mutli-channel funnel reports & the first click attribution?

Facebook reports on ad clicks so depending upon the type of ad being run, you could just see your ad being clicked on for many purposes. Google has a specific definition for when a session starts. When you say Google only shows 2 users - do you mean in total or from Facebook?

Side note: Using tracking codes in sponsored posts doesn't actually show facebook / cpc being the referring resource because these show up as organic traffic in 3rd party website tracking cookies.

[DJ123]

People could possibly have their browsers set to "Do Not Track" which would throw it off.

Also, their are tons of plugins out there for Firefox, Chrome and IE which specifically act to block tracking.  Ghostery comes to mind, Ad Blocker Plus, etc.  Mobile browsers also can add plugins like this.

The best solution is to have multiple tracking codes/devices in place, then take all of the data together over the same period of time and analyze the numbers.  Don't forget your own server logs either!  They are a valuable source of data.  You're going to have to bring in multiple data sources and develop a range of values, similar to confidence intervals.  Check: http://en.wikipedia.org/wiki/Confidence_interval to see what I am getting at.  So you would say something like, "with 95% confidence, our real traffic lies between 1,000 to 1,400 unique visitors per day" or whatever the number is.  Of course track your real conversions and try to get data from your paying customers like asking them how they found you.  It is amazing the intel you can gather from your actual customers.  Analytics is good, but don't underestimate what customers tell you face to face.

Best of luck!

[LesleyPaone]

It could be a myriad of things. One thing I would suggest is using tracking urls that should solve, or at least shed some light on what is happening. You can use them in conjunction with AWStats to see what people are actually coming and try to figure out if it is because of a user agent or even a specific platform they might be using. It could be some facebook web phone app issue or something.