Read on Twitter
We’ve been digging into LinkedIn organic performance recently, mostly looking at Company Pages.
We began with a conceptual model of what makes a post show up (or not) in a follower’s feed.
(THREAD)
We began with a conceptual model of what makes a post show up (or not) in a follower’s feed.
(THREAD)
The most important thing to understand is that LinkedIn’s editorial algorithm is probably attempting to optimise for “the most interesting or useful content for each individual feed.”
The business goal is “increase visit frequency.”
The business goal is “increase visit frequency.”
The algorithm must choose from ALL the content produced recently by every single one of your connections, as well as all the Company Pages you follow.
(NB: a majority of LinkedIn users still don’t post much. And those who do are frequently what @broderick calls “hustle goblins”)
(NB: a majority of LinkedIn users still don’t post much. And those who do are frequently what @broderick calls “hustle goblins”)
LinkedIn isn’t parsing the content of the posts.
Instead, it selects content by interpreting some very simple signals. Our best guesses at these are as follows:
Instead, it selects content by interpreting some very simple signals. Our best guesses at these are as follows:
What’s your relationship with the post’s author? What’s the recency & frequency of engagement?
(Since people don’t post that often, it’s hard to design an experiment to demonstrate this.
But a simple test might be:
1. Follow 2 new pages
2. See if their posts appear in your feed
3. A/B test - Like & Comment on one, ignore the other)
But a simple test might be:
1. Follow 2 new pages
2. See if their posts appear in your feed
3. A/B test - Like & Comment on one, ignore the other)
Another signal: how much engagement has the post received? More engagement means it’s likely to be more interesting.
More engagement from people in your own network may mean it’s really interesting.
More engagement from people in your own network may mean it’s really interesting.
One of the side effects of these signals is that content from a company you work for will likely score high.
1. LinkedIn automatically follows your employer’s page for you.
2. You are likely to have engaged with your employer’s posts
3. Your colleagues have also engaged...
1. LinkedIn automatically follows your employer’s page for you.
2. You are likely to have engaged with your employer’s posts
3. Your colleagues have also engaged...
This leads to some odd artefacts. It’s hard to know who’s exposed to a post, but we CAN see who’s engaged. We collected these data from 2 pages.
On one, we found that more than half of post engagement was internal.
On one, we found that more than half of post engagement was internal.
On another page (a top law firm) we looked more closely at each individual.
We found that external engagement primarily came from competitor firms, suppliers, and jobseekers.
We found that external engagement primarily came from competitor firms, suppliers, and jobseekers.
We believe that page following by non-employees might broadly be thought of as “aspirational.”
That is, people follow companies for whom they might like to work, not those they’d like to employ.
That is, people follow companies for whom they might like to work, not those they’d like to employ.
When we DO see external engagement, we believe that this is more likely to come from the edges of employees’ networks.
External audiences are exposed to company content through the engagement of employees.
External audiences are exposed to company content through the engagement of employees.
If this is the case, then we might better understand a Company Page to be a tool through which a company can support its employees by feeding them content for their personal networks.
This conceptual model could help define a better content strategy.
This conceptual model could help define a better content strategy.
We get another clue when we look at what kind of engagement drives visibility.
Unsurprisingly, we do see median impressions rise with sharing activity.
Unsurprisingly, we do see median impressions rise with sharing activity.
But again and again, we think we see evidence that commenting counts for more than sharing.
Our data sets aren’t very robust, because comments on Professional Services firms posts can be like hens’ teeth.
Our data sets aren’t very robust, because comments on Professional Services firms posts can be like hens’ teeth.
We think that in the LinkedIn may be up-weighting comments. We have a lot of anecdotal evidence to support this.
We believe that LinkedIn rewards “audience investment.”
We believe that LinkedIn rewards “audience investment.”
It costs almost nothing to “Like” a post.
It costs more to click on a post. That’s an indicator of curiosity.
And we see evidence that LinkedIn rewards posts that maintain clickthrough rates as they’re exposed to wider and wider audiences.
It costs more to click on a post. That’s an indicator of curiosity.
And we see evidence that LinkedIn rewards posts that maintain clickthrough rates as they’re exposed to wider and wider audiences.
Sharing a post would seem to be relatively high investment. The user is explicitly associating themselves with the content in some way.
We suspect though that a Share from an internal audience member isn’t as interesting as a share from a stranger.
We suspect though that a Share from an internal audience member isn’t as interesting as a share from a stranger.
A Comment, however? That may be a big thing. It’s actually “original content”.
Take a look at your LinkedIn feed. It’s likely that you’ll see (near the top) “[your 1st degree contact] commented on this.”
For LinkedIn’s algorithm, that’s potentially 2x the interest for you.
Take a look at your LinkedIn feed. It’s likely that you’ll see (near the top) “[your 1st degree contact] commented on this.”
For LinkedIn’s algorithm, that’s potentially 2x the interest for you.
We produce a monthly newsletter for our clients where we write up our thoughts & research in a considered, concise format.
Like THIS post if you’d be interested in being added to the list.
Like THIS post if you’d be interested in being added to the list.
