My number one LinkedIn tip for the longest time has been not to share links directly in LinkedIn posts.
But now things have changed.
- Introduction
- The data: testing links in posts
- Why did this change happen?
- Link shortening
- Link-click experiment
- Let’s wrap up
Introduction.
A few people contacted me in late May 2020 to say that LinkedIn was no longer suppressing the reach of posts containing external links.
Until then, including an external link in a post would mean that post would have very little visibility with your followers and extended network. Why?
Because anything that was set up to take people’s attention away from LinkedIn (as external links surely do) wouldn’t be good for LinkedIn.
And because this is their house with their rules, LinkedIn could happily stop such content from being seen.
But that seems to have changed around mid May 2020.
The data: testing links in posts.
As an eager experimenter, I wanted to test this myself. So, I popped on my white coat and goggles and headed into the LinkedIn Lab.
I started using links in short-form LinkedIn posts in early June 2020. I wanted to know what the visibility would be for such posts.
Would LinkedIn’s algorithm trample all over the reach of this content as it had done for the past couple of years?
Below is a list of short-form LinkedIn posts including links, along with their content type and view count.
Post | Type | Views |
---|---|---|
Jobseeking tips from a non-expert | Text | 22,119 |
How do I contact LinkedIn for help? | Text | 9452 |
Selfie tips for iPhone users | Doc | 6653 |
Friday Shout Hari Haralambiev | Image | 4318 |
Link shortening for URLs in LinkedIn posts | Text | 20,085 |
Tips for better article headlines | Text | 7044 |
More free LinkedIn courses until 30 June 2020 | Text | 19,914 |
Friday Shout Synne Lindén | Image | 5813 |
Average | 11,925 |
Note that this test relates only to LinkedIn posts and not to articles, as links in articles have never been an issue. Each view count was recorded 2 weeks after I published the corresponding post.
(See my article about LinkedIn view counts if you’re not sure how they work.)
How do these figures differ from before?
Compare the figures in the table above with what happened in December 2019, when I accidentally put a link into an image post.
That image post – LinkedIn’s analytics suck. SHIELD’s don’t. – performed terribly by my usual standards:
- 566 views
- 9 reactions
- 11 comments
That’s 8 times lower than my average for image posts.
Even worse, the terrible post performance happened on a Saturday, which is normally my second best day of the week for engagement. This is bettered only by Sunday posts – so much for people not being on LinkedIn on the weekends, eh?
The same sort of terrible viewing figures used to happen before I made the change to the write-post-edit method for sharing links about 2 years ago.
Updated stats for January 2021 to August 2021.
I analysed 172 posts I’ve made since the start of 2021 and found the following average stats.
📖🔗 Text/image/document posts – links
- 6824 views
- 120 reactions
- 107 comments
📖⛔️ Text/image/document posts – no links
- 6761 views
- 100 reactions
- 105 comments
🎬🔗 Video posts – links
- 1903 views
- 90 reactions
- 92 comments
🎬⛔️ Video posts – no links
- 2128 views
- 104 reactions
- 96 comments
It looks as though engagement on video posts suffers a little if the posts contain a link but there’s very little difference for non-video posts.
Almost all my video posts are “native” videos, not links to YouTube or Vimeo.
Don’t stuff lots of your posts with links. Use them judiciously and remember that people probably don’t want to be pointed off the platform unless it’s essential. Serve people where they are if you can.
You can place links in comments instead if you wish, but note that those comments might be buried in the noise if there’s lots of discussion on the post.
Why did this change happen?
It makes sense to me that LinkedIn wouldn’t want to promote content that has the potential to take users off the platform.
Therefore, I was never surprised that link-based posts were suppressed from being spread through the network.
So, why the change? Why is it now OK for the algorithm to happily promote posts containing links? And will that change last?
I don’t know for certain but it looks as though the change is related to the LinkedIn Engineering team’s May 2020 blog post about dwell time, which is now a ranking factor for the visibility of posts.
Dwell time is the measure of how long you spend looking at a piece of content. I discuss this in my June 2020 slot on the LinkedIn Sofa:
Say you see a post with a few lines of text visible and then you click the “see more” link to read the rest.
If it’s engaging, high-quality stuff, you might sit there for a little while to take it all in. But if it’s rubbish, you scroll on immediately or click away altogether.
Naturally, LinkedIn will want to present you with content that maximises your dwell time, because the more time you spend on the platform, the more valuable the platform becomes to you.
And from LinkedIn’s point of view, you’ll be more likely to pay for a premium package or consume ads if you hang out there longer.
Two types of content that will help people to stick around and increase the dwell time of your posts are videos and document posts.
For example, I recently shared a quick intro video to myself on LinkedIn:
It was only 40 seconds long but I knew it would give me decent dwell time and likely get a lot of engagement through comments (and it did).
Similarly, I like creating document posts, because they get great dwell time and engagement. Some examples:
But how does this relate to links in posts?
Well, if LinkedIn think that showing you a helpful link that’s part of an engaging piece of content could increase your overall dwell time, that might override their fear of you clicking the link and disappearing to some other place on the web.
Such a change in attitude would mean that there would be no great value in reducing the visibility of posts containing links.
On top of that, a recent change to the LinkedIn user interface on mobile is that external links in posts are now displayed differently when tapped.
This means the external content is loaded into the LinkedIn mobile app but that the original post remains viewable at the bottom of the screen while the external content is displayed.
This blended way of displaying content might help LinkedIn track how viewers engage with the external content while giving those viewers an easy way to get back to the original LinkedIn post.
Given that ~60% of LinkedIn sessions take place through the mobile app, LinkedIn will probably become more comfortable with displaying external content this way.
If that’s true, perhaps there won’t be a need to suppress posts containing links in the future. Or at least not until we reach a content saturation point – and we’re definitely not there yet.
The new CEO theory.
I’m into the realms of guesswork here, but it’s also possible that the change to allow link-based posts to succeed has been encouraged by the new LinkedIn CEO, Ryan Roslansky, who took over from Jeff Weiner on 1 June 2020.
I think a smart CEO would make moves to maximise the punching power of organic content on the LinkedIn feed, as that would surely attract more content creators. Perhaps that’s what’s going on here.
As I mentioned above, I don’t think this sort of promotion can last forever, as we’ll eventually reach a content saturation point. At the time of writing, LinkedIn has 690 million members (the latest figures are published on the LinkedIn Pressroom). That’s a lot but is still a drop in the ocean compared with Facebook.
Perhaps when we reach the content saturation point on LinkedIn, my good old write-post-edit method will kick back into action again.
Link shortening.
If your post includes an external link longer than 26 characters, LinkedIn will change it in the published post to lnkd.in
format.
This happens on publication, not during composition.
This shortening isn’t great because readers can’t see where the link will take them, unless they use a free service such as getlinkinfo.com to check the destination address.
A workaround that may help is to remove the http://
or https://
protocol from the start of links placed in posts, as that saves part of the 26-character budget. You can usually remove www
from your URLs, too.
Here’s an example:
espirian.co.uk/call/1234
– 24 charsespirian.co.uk/call/12345
– 25 charsespirian.co.uk/call/123456
– 26 chars (all good so far)https://lnkd.in/e9inqRE
– 27 chars (shortened!)
You might want to use a free link-shortening service such as Bitly to better manage what’s displayed to your audience. This lets you customise your short URLs and gives you visibility of how often your links have been clicked.
Link-click experiment.
In November and December 2021, I did a small experiment to see what proportion of people clicked the links in my LinkedIn posts.
The results weren’t good!
Let’s wrap up.
It looks as though it’s once again OK to include links in short-form LinkedIn posts.
Don’t use this insight to batter your audience with external links. They’ll still prefer to read or watch content without having to go anywhere else. If you can serve people where they are, that’s always going to be the best approach.