• Background Image

    The Facebook Link Shim



The Facebook Link Shim

Are you driving traffic using Facebook and using Google Analytics to analyse that traffic?

If you are – then you are likely to see multiple referrals like the ones below:

• m.facebook.com / referral
• l.facebook.com / referral
• lm.facebook.com / referral
• web.facebook.com / referral
• touch.facebook.com / referral
• facebook.com / referral
• mobile.facebook / referral
• business.facebook.com / referral

What are these Facebook referrals?

Facebook referrals come from a powerful Facebook tool called ‘the Link Shim’ that has been built to help protect the privacy of its users or protect its users from malicious websites, content and links.

It is really a tool Facebook uses internally to manage links.

For example, when you see a l.facebook or lm.facebook it means that someone passed through a Link Shim before landing on your site (the m is just for mobile). This person would have seen a warning and asked if they wanted to continue to a website or to cancel.

Mostly, these are triggered because of privacy not because of a malicious site. You can always open one of your links from Facebook and it will show you if a site has been flagged as it will show a warning.

Do these split up referrals help you?

These multiple referrals don’t help you in any way to analyse your data and really only split your data. When your data is split – it makes it more difficult to analyse your Facebook ads and posts.

For example, if you want to segment your referrals by location e.g. cities, it is going to be more difficult to do that with multiple referrals. You really need one referral for each of your marketing platforms like Facebook, LinkedIn and Twitter.

To tidy things up you can indeed combine all versions of FB referral within your Google Analytics and you also need to know which ones to combine. We suggest combining the following which we know come from real users:

• l.facebook.com
• lm
• web
• touch
• mobile

Others, like business.facebook.com, are likely to have another source and do not come from real users.