How does the Linkedin Algorithm work?
How does the Linkedin Algorithm work?
It's a question I have seen asked a lot recently, along with how you should structure content.
After digging around a little, Linkedin’s engineers have been very open about how the content algorithm works. As ever quality content wins out. Here is why:
Firstly you need to remember Linkedin views its members as exactly that, members! You are able to personalise your feed which helps filter content. It was a point Mark Williams (Mr Linkedin) raised when I asked what tips he could give to making the most out of the Linkedin experience.
“I would say that the engagement opportunities on your home page have increased considerably but there is still a lot of ‘crap’ blocking your view. I have spent time unfollowing people who autopost content and promotional posts and conversely followed those that are highly engaged. The result is a feed that always provides interesting and useful engagement opportunities and the more I engage, the better my feed gets”
This is also true of the algorithm when you post content on Linkedin:
Uploading content
Right at the outset your text, images and videos are being classified as they are typed or uploaded, your text in short form and long is being processed to determine what grade of content you are producing.
At creation, the content is classified into:
- Spam
- Low Quality
- Clear
Audience Reach
As your content is served to your network predictions are made using Hadoop Clusters (WTF is that I hear you cry - Hadoop is an open source, Java-based programming framework that supports the processing and storage of extremely large data sets in a distributed computing environment)
Basically, your content is A/B tested and the size of the audience you reach initially is proportional to if your content is Spam, Low Quality or Clear. It is one of the reasons that content is able to stay around longer on Linkedin where is disappears far quicker on other networks, you are essentially drip feeding to your network.
We can also make some assumptions that this helps protect against those who are building large networks (although it may not stop the downloading of emails and the familiar line within a newsletter - you are receiving this email as we are connected on Linkedin)
This process is done by the programs (Kafka, Rest.li,liblinear, and TensorFlow) as a lot of content can fall into a grey area it can be manually checked.
Based on engagement and reaction of this sample determines how big your reach is and how many of your connections contacts will see the post by serving into news feeds outside your network.
What content should we post?
Every SocialDay delegate knows that quality content is the way to go and the virtues of creating content that need to be engaging and relevant to the audience do not need to be preached in this article.
Linkedin has changed alot over the last 6 months. The introduction of native video and a resurgence of attitude. It now feels like more than just a place for recruiter to find candidates and members are behaving more like true versions of themselves. The past attitudes are changing and engagement continues to increase, it's a great place to put across your opinion after all it's still a forum from business. That does not mean you can't mix it up. Chris “Kubby” Kubbernus says:
"My biggest tip is to try and break Linkedin. The old rules of LinkedIn and ultra conservative professionalism is dying. I am not saying be unprofessional. I am saying, press the envelope and see the results"
With the use of Native video B2B marketers now have a great way to stand out on a highly targeted platform. Be creative and don’t be afraid to let the personalities behind the brands shine.
We can assume that the more engagement that you get the wider your reach into new networks through those who engage with your content will be, this is based on the Hadoop clusters. We could not find any official figures that rank the levels of engagement, anecdotally many feel that comments are worth more and this would be supported by the Linkedin SSI (Social Selling Index) which actively encourages you to:
- Build a personal brand
- Connect with relevant people
- Share content that is relevant to your network
- Build relationships (talk to people)
Sharing content is a key pillar to building SSI scores, so creating the right content for brand or business that is shareable is a win win.
Some great articles and case studies exist about how to write content, and in particular headlines, you should check out Neil Patel’s blogs on this as it offers some interesting facts about how to make your content stand out (not just visually)
Long Form
Another area to look at is the use of longform posts, this used to be a separate feature but now is served alongside content in the news feed (write article button on the status update) This is an excellent way to get thought leadership and deeper thinking scene.
Talking to Kirstie Smith (Make & Believe Digital)
“The next thing on the list for me on LinkedIn, is writing and sharing more articles. It’s really easy to use the LinkedIn articles platform and I’m reaching more people on LinkedIn than I am through my blog (website). I personally see, and then read, a lot of LinkedIn articles appear in Google search”
It’s all in the detail
To make the most of Linkedin before all else make sure your profile is up to date if you are personal branding and likewise for company pages and showcase pages. You’ve done the hard part writing the content, don't fall foul by having outdated info or incomplete info on your pages.
Finally make the most of the stats that linkedin provides, not many platforms will allow you to see who has been engaging with your content. This useful feature means you can ascertain if your content is hitting the right audience, and you can see by the reach if it falls in tp spam, low or clear.
We have many sessions at SocialDay 2018 covering a host of topics, have you got your ticket yet? grab yours today ? here