Have you been seeing longer LinkedIn posts from people in your network? This isn't a coincidence. See how LinkedIn's latest algorithm update affects your feed.
2. Dan posts on LinkedIn.
Machine learning (ML)
predicts the probability of
Mary taking a "viral action" on
Dan's post.
What is a "viral action"?
3. A viral action is a react,
comment, or re-share.
Profile data, connections, completeness of
profile, etc.
# of global clicks, # of viral actions on Dan's
post, etc.
Mary's affinity for Dan's post and profiles
similar to him.
Contextual features like the time of day Dan
posted on LinkedIn.
LinkedIn's ML model also considers:
4. Mary re-shares Dan's post. This
propagates it across her network
and adds it to their LinkedIn feeds.
Viral actions have 2 types of effects:
Effect 1: Downstream
Mary comments on Dan's post. This
provides him value and encourages
him to post more frequently.
Effect 2: Upstream
Then what happens?
5. All outputs are synthesized into a single
weighted score. This ensures LinkedIn feeds
are balanced and provide value to users.
What is "dwell time"?
Then comes "dwell time"...
6. Dwell time is how long a user
spends reading a LinkedIn post.
Dan The Man
Marketing @ Company
7h •
Let's talk about posting on LinkedIn with this new
algorithm update.
Now LinkedIn is considering something called "dwell
time", or how long a user spends reading a post.
....see more
7. Why LinkedIn cares about dwell time
Actions like likes, comments, and re-shares
can be "noisy" and sporadic. For example,
liking a post then quickly exiting the page.
Not every engaged user will take an action.
Maybe they want to curate their feed a
certain way?
Re-shares without text are less valuable
than ones with text.
Why dwell time is important →
8. Dwell time is a more reliable indicator of user
engagement on a LinkedIn post. It is also more
measurable, according to LinkedIn tests.
Dwell time isn't replacing viral actions. Think
of it as another "signal" LinkedIn is using to
curate feeds.
This is why you are going to see a lot more
long-form posts on LinkedIn.
9. Thanks to Siddharth Dangi,
Senior Software Engineer at
LinkedIn, for his thorough blog
on the dwell time update!