Subscribe
Share
Share
Embed
Zero Click Marketing is a marketing strategy podcast about content marketing, audience research, and how brands grow when clicks matter less. Hosted by Amanda Natividad, Chief Evangelist at SparkToro, the show explores how marketers reach audiences, build influence, and earn attention in a zero-click internet. New to the show? Start with Episode 2: What Zero Click Marketing Actually Is.
[00:00:00] Hey, it's Amanda. I originally published this episode back in March, and I wanted to bring it back because since then, LinkedIn has told us more about how its feed works. The episode you're about to hear was based largely on Christopher Penn's unofficial LinkedIn algorithm guide, which synthesized a bunch of LinkedIn engineering papers and posts.
[00:00:23] And one of the big ideas I pulled from it was that what we actually call the LinkedIn algorithm is really a multi-stage system.
[00:00:32] First comes retrieval. Out of the enormous universe of possible posts, which ones should even be considered for your feed? Then comes ranking. Among those candidates, which should you actually see? Since then, LinkedIn has published more detail about its feed systems, and the basic architecture holds up. So first caveat: when you hear me say LinkedIn has two algorithms, think of that as shorthand.
[00:01:05] It would be more technically accurate to say that there are two major stages, retrieval and ranking, with multiple systems and models operating inside them. There's also one other thing I would phrase a little differently today, so the second caveat. In this episode, I talk about topic consistency, and I still think that matters.
[00:01:28] I don't mean you need to pick three content pillars and never talk about anything else. There's no evidence that LinkedIn punishes you for having multiple interests. The better way I would describe that now is that it helps to have a recognizable center of gravity. LinkedIn is trying to learn what people are interested in and which posts might be relevant to them.
[00:01:53] So consistently participating in certain professional conversations gives both the system and your audience more information about what you know, what you care about . So consistently participating in certain professional conversations gives both the system and your audience more information about what you know, what you care about, and who might want to hear from you.
[00:02:20] You can wander. You just probably want people to know what room at the party they're most likely to find you in. And then there's one separate, very recent development worth mentioning because people are talking about it a lot right now. LinkedIn has added a reporting option called Seems Like AI Slop, which lets people flag content that feels like low-quality, mass-produced AI material.
[00:02:49] And I would not interpret that as LinkedIn hates AI or even LinkedIn penalizes AI-written posts. The more useful interpretation is that LinkedIn is getting more explicit about trying to distinguish between content that contributes something useful to the feed and content that just adds more stuff to it, which actually makes the original takeaway of this episode feel even more durable.
[00:03:20] Your LinkedIn strategy probably shouldn't be about memorizing a bag of algorithm hacks. Your job is to be clear and to become understandable. Make it easy for people, and yes, probably systems too, to understand what you know, what you care about, and why they might wanna hear from you. And being easy for the algorithm to understand is useful.
[00:03:48] Being worth understanding is probably even more important, which brings us back to my extremely sophisticated theory of social media. LinkedIn is basically a party, like a real life, albeit nerdy party. Here's the original episode
[00:04:10] Amanda Natividad: I'm Amanda Natividad. Welcome to Zero Click Marketing.
[00:04:18] Hey, it's Amanda. This is an episode format I'm trying out called ZCM Field Notes. So these are short reactions or observations of what's happening in the field. Overall, these are going to be less polished, but I'm trying my best. Today, I want to talk about the unofficial LinkedIn algorithm guide by Christopher Penn.
[00:04:39] So Christopher set this out on his Substack last Friday. His Substack is almosttimely.substack.com. This LinkedIn algo guide is 150 pages, so there's a lot to get through, and we only have, like, seven minutes, so let's go. the first thing I thought was really interesting is that the report says LinkedIn doesn't really have one algorithm anymore.
[00:05:02] It has two systems working together. Distribution happens in two stages. So stage one is something called retrieval. This is the system deciding, should this post even be considered for someone's feed? That's before ranking and engagement. Then comes stage two, ranking. So now the system asks, out of the thousands of posts that could appear in someone's feed, which ones should show up first?
[00:05:30] So ranking That's more of what we usually think of as the algorithm, and this evaluates signals like engagement patterns, interaction history, dwell time, relationship strength. But the key idea is this: if retrieval fails, then ranking never even happens. And when I read that, it actually reminded me of something that I've already been telling people about social media, which is social media is a party.
[00:05:59] When you walk into a party, you don't climb onto the coffee table and start shouting over everyone. I mean, at least you shouldn't, right? Instead, you do what normal humans do. You walk in, you find your friends first, you settle into conversations. You know, then you, you walk the room, you meet someone new, you join existing conversations.
[00:06:20] And after you've warmed up a little bit, that's kind of when you start your own conversations. And that's normal, sociable party behavior. And that's kind of how social media works, too. And this is where Christopher Penn's report clicked for me because for years, the advice around LinkedIn, and I think most social media, has been stuff like post at the right time, warm up the feed, comment right before you publish your own content, and all of that.
[00:06:46] And I'm not saying that stuff does nothing, but this report argues that before any of that can help you, LinkedIn first has to understand who you are, what you talk about, and who should care. That's the retrieval part. That's the do you even get invited into the conversation part. Which is why I think the party analogy works.
[00:07:09] At a party, people don't need your full life story, but they do need to get a rough sense of, you know, what your deal is, right? So for LinkedIn, it's, "Oh, Amanda, marketing, audience research, zero-click marketing. Okay, got it." So that context helps people know where to place you in the room. And according to this guide, LinkedIn is doing something pretty similar.
[00:07:31] So it may be using your profile, your post language, and your engagement behavior to figure out your professional identity and what conversations you belong in. That also means your profile matters more than a lot of people think Not just because humans look at it, but because the system may be using it to understand what topics you're associated with, what expertise you signal, and what kind of audience is most likely to care about your posts.
[00:08:00] And then there's the whole topic consistency, which I think is one of the most useful takeaways in the whole report. If one day you post about marketing, the next day you post about parenting, the next day about cryptocurrency, then personal productivity, sourdough bread, the system, and honestly your audience, might struggle to know who your content is actually for.
[00:08:27] Again, think about the party. If someone keeps manically jumping between totally unrelated conversations, people start wondering, "Wait, what's your thing? What do you even care about? And why are you trying to sell me your NFT? NFTs aren't even a thing right now." But if someone consistently contributes to the same conversations, people start associating them with those topics, and so do the systems.
[00:08:56] other interesting things from Penn's report, LinkedIn is constantly learning patterns like who interacts with your posts, how often, and for how long, so their system is trying to learn your audience graph over time. Relevance matters more than timing, so it matters less that you optimize for the time that your audience is online and more about whether your post is relevant to them.
[00:09:18] And then also your own daily regular engagement is important. LinkedIn systems track specific behaviors that constitute what they call professional interactions. These behaviors include long dwell time, so the time spent reading, not just scrolling. It also includes reactions, comments, and reposts. Now, I want to be clear about something.
[00:09:42] This report is not official LinkedIn guidance. It's an interpretation of more than 30 current LinkedIn engineering publications, which includes research papers on their blog. So I wouldn't treat it like gospel, but the broader idea still makes sense. Modern recommendation systems are looking for predictable relevance.
[00:10:03] And that brings us back to zero-click marketing. The most important thing you can do is to be easy to understand. If people and systems can quickly understand who you are, who you help, what you talk about, and why your perspective matters, then your content becomes easier to distribute and easier to amplify.
[00:10:25] So if you want to think about a simple strategy for LinkedIn, think about the party. Don't walk in shouting. Walk in and join the room. Talk to your friends. Introduce yourself to new people. Join existing conversations. And over time, people will start listening when you start one of your own. And if this analysis is even partially right, the real LinkedIn strategy isn't gaming the algorithm.
[00:10:51] It's becoming easy for the algorithm to understand. That's today's ZCM Field Note. Thank you for listening. more to come on zero-click marketing tomorrow, or maybe the day after. I have not recorded the next episode yet, so subscribe so you don't miss it. Bye, friends.