Less than half of Facebook ad clicks are visible in Google Analytics

Question

We have a website that we advertise via Facebook, but in Google Analytics only 30-50% of the clicks appear. (The page is not really public yet, so I cannot post a link here)

We already did several test runs and tried different optimizations. Unfortunately none of the optimization resulted in a massive change. (Optimizations like improving page load time, in particular loaded JS like jQuery assets via CDN, etc...)

Right now we have several hypotheses why this happens:

• the page is hosted in Virginia, however the site is for Brazil, so latency might be too high
• in general it takes too long until the page appears, so people might close the page before it's fully loaded
• it's a one-page app, so it might be slow on old computers
• too much AJAX, in particular we use Facebook Connect
• people might use adblockers

We ruled out that Facebook gives us the wrong number of clicks, we already take the number of actual website clicks.

Benchmark results from webpagetest.org using São Paolo/Chrome/Cable:

• Load time (First View: 3.9s-6.7s; Repeat View 1.9s-2.8s)
• First Byte (First View: 0.5s-1.6s; Repeat View 1s-2s)
• Start Render (First View: 2.3s-5.4s; Repeat View 1.4s-2.2s)
• Document Complete (First View: 3.9s-6.7s; Repeat View 2s-2.8s)
• Fully Complete (First View: 5s-8s; Repeat View 2.9s - 3.6s)

(Yesterday we changed the hosting provider, but I think the numbers before should have been similar.)

What could be the problem? To us it is very mysterious, we would assume that the click counts from Facebook would differ from Analytics' numbers by less than 5%. Seeing more than 50% of the users not reaching our website, makes us believe there is a single big problem that we are not aware of yet.

Solution

We have a website that we advertise via Facebook but in Google Analytics only 30-50% of the clicks appear.

This is quite common with ad programs, including Facebook ads as can be seen here. As covered there:

Facebook Ads reports clicks. Many third-party reporting packages report visits or page views, which may not correspond directly to Facebook’s click tracking.

Some of the other reasons for discrepancies between ad clicks and Google Analytics reports are:

• Users can click on ads more than once, particularly if your site is slow to load.

• JavaScript, cookies, or images might have been disabled in their browsers.

• If the user leaves (bounces) from the landing page before the tracking code executes, then the click will not be associated with the visit. This might also be hastened by a slow loading page, and problems with compatibility or viewing it.

• Facebook statistics might reflect different time zones and daily intervals than Google Analytics reporting.

• Filters in Google Analytics may remove some of the data from your reports - check for those.

Note that Facebook also recommends comparing the results from more than one analytics package:

We strongly recommend installing multiple statistics packages and comparing them to each other (as well as to Facebook Ads reporting) to get a more complete picture of site traffic.

Lastly, this Google Analytics guide regarding ad discrepancies might also be helpful to you.

Comments

[Stephen Ostermiller]

See also: Huge difference between Facebook Ad Click figures and Apache log requests

[ashleydc]

Thanks, we have collected a lot of evidence for this. On the one hand we have counted no. requests to the landing page and no. unique IPs and the number of unique IPs was significantly lower than the former. (Evidence for double-clicks) Also we improved landing page performance and the "success rate" increased by 50%.

[dan]

So it sounds like loading speed of the landing page and double clicks were the major issues. Thanks for the updates - I'm sure this will be helpful to others.