Informed 491

Postbag.

Judi Radice Hays: feedback on birthdays

Jeremy Freeman: LinkedIn newsletters

Kathie York: Since LinkedIn likes to see engagement within the first hour of a post, is it still bad form to be the first commenter? One reason I ask: having a link in a post hurts our ‘reach.’ What are the rules, now, for placing the link in a comment we create? And what happens to that comment if someone reposts our post?

---

875 LinkedIn employees (5% of workforce) sacked!

Is AI slop really going away? We need a “report as AI” feature!

“Nonconsensual intimate imagery” warnings.

How to find the date and time of any post or comment on LinkedIn: see my free tool at jesp.me/date

Full transcript.

What is being done about AI slop on the LinkedIn platform? That’s the topic of episode 491 of the Informed podcast.

Hi, everyone. Welcome back to the Informed podcast. I’m your host, John Espirian, and this show is all about LinkedIn best practice, with weekly tips on how you can improve the way you use this platform.

We’re going to start as usual with some listener questions in the Postbag.

OK, the first item in the Postbag is a voice note that’s come from Judi Hays, who’s a member of my Espresso+ community. And this is actually feedback on the topic of birthdays on LinkedIn, which was talked about last week. Let’s take a listen to what Judi’s got to say.

Thank you for your feedback there, Judi.

Word clouds are not something I really make myself, but that’s an interesting option and I suppose it’s a bit more personalised than just sending a short text note.

Creating content around your own birthdays: if it’s a significant life event, I’ll be 50 in a year and perhaps I’ll do something around that.

So, long as it doesn’t come across as a bit of a cheesy excuse to put out something helpful. It should just be helpful regardless. But possibly that could be an interesting angle for people to take.

Interesting that you started by saying that you turn off your birthday notifications until a week before your birthday, but then you presumably wouldn’t get any approaches because of your birthday at other times of year. So, I’m wondering whether that actually does anything good for you. But anyway, it’s something for us all to think about.

As I said last time, I’ve just turned mine off because there’s just far more important things to worry about on LinkedIn than people congratulating each other for birthdays, I think. But thank you, Judi, for your feedback. OK, next question is another voice note.

It comes from my Mission Control partner for UpLift Live, Jeremy Freeman. Let’s listen to what Jeremy’s got to ask.

Yeah, thanks for that, Jeremy. Not the first time I’ve been asked recently about the power of newsletters.

I do get that the numbers can seem quite small, and I think these days it’s hard to build up a momentum around that longform content. I was fortunate to get access to newsletters on LinkedIn some years ago and built up a lot of followers very quickly. My follower count on my newsletter subscription now currently stands at almost 17,000 people.

But I do find on the relatively rare occasions that I post newsletters on my personal feed, they don’t get anywhere near the kind of visibility and engagement that they used to. So, that does feel as though it’s been dampened.

But as I’ve said before, all longform content on LinkedIn – so that’s either an article or a newsletter – all of that content gets indexed by Google and other search engines and therefore it gives you long-term discovery potential.

And now that the AI platforms are really looking at the LinkedIn domain as a trusted source of content, you’ll start to see more and more people find you because of an AI-based search. And I’m seeing that in some of my invitations.

People are saying that they found me as a LinkedIn trainer through a ChatGPT search and I still get people finding old articles that I wrote 5 or 6 years ago and interacting occasionally. So, that doesn’t really happen with shortform posts. No one’s really going to discover that kind of content years after it’s been produced.

So, if you think about it in terms of building your content estate, the long-term play means that it is really quite good still to invest some time and effort into building that LinkedIn newsletter, although the short-term numbers might not look great, I think the long-term drip effect of it is worth sticking to.

So, stick at it, Jeremy, and good luck. The other question that we’ve had for the Postbag this week is from another Espresso+ member, Kathie York. She writes in and says:

So, a few things in one there from Kathie. It’s not really the best if you’re the first commenter on your own post. Sometimes, you’ll notice that you’ve forgotten to add something to the post, in which case you’re kind of writing that first comment almost like a PS, but there’s nothing to stop you editing the post and adding to that if you need to.

Unless you’ve run out of the 3000 characters that you’re allowed on the post. But generally it’s probably best if someone else could be the first commenter on your post.

In terms of links, I’ve said before, if you are going to include a link, try not to in the first instance, but if you have to, then it’s much better to put it into the post because then at least it has a chance of being seen, because the algorithms will try to hide comments that contain links. And certainly if you repost something, the comments don’t really come alongside for the ride.

So, that comment containing a link will probably be lost when someone reposts with thoughts. So, yes, I’d avoid that altogether if possible.

The ideal is to get lots of early comments, but not from yourself, and try to avoid links in those comments if you can.

OK, the main topic is about whether AI slop is being dealt with effectively on the LinkedIn platform.

Well, I’ll start first by saying that LinkedIn have announced that they’re getting rid of, I think it’s 875 employees, which is about 5% of the workforce. So, that seems to be one of the first acts in charge by the new CEO, Dan Shapero.

Feels like a significant cull and I believe that a lot of those employees are engineers. I wonder what that means. Does it mean that AI is going to be doing more of the heavy lifting at LinkedIn so that they can save some money? Wouldn’t surprise me at all.

In which case, how does that tie in to the fact that LinkedIn are trying harder than ever apparently to get rid of AI slop with fewer people doing that work.

Well, a few LinkedIn employees have actually posted about this and the latest news is that there’s now a 3-step attack on AI slop.

So, LinkedIn are going to be working harder to detect generic AI written posts and comments in the feed, to try and crack down on AI automation tools that result in loads of spam comments. We’ve all probably seen those.

And also to crack down on anything that looks like attention bait, especially videos with random footage and just really generic basic business wisdom that is just there for engagement. I don’t know how whether that’s going to be effective, and I think LinkedIn has said this sort of thing in the past, so I’ll believe it when I see it.

I don’t get that much AI slop, I suppose, in my feed, but then I work quite hard to try and curate what I see by unfollowing people or muting people or not connecting with people in the first instance if I think their content isn’t going to be of interest to me in my feed. And so it’s natural that I wouldn’t see as much as maybe the average LinkedIn user does.

But even then I still see some posts that get hundreds and hundreds of comments and it’s very obvious that some of those comments are just not from real people: they’ve just been automated in some way. So, I hope that LinkedIn is true to its word and can reduce these things.

What I’d really love to see is a very obvious, easily accessible button, link, or something like that on posts and comments that says “report this as AI”, just to make it really easy for us humans to say look, this is wrong and if LinkedIn get enough reports about posts and comments, then they can just, you know, adjust their algorithms and not show more of that stuff and maybe penalise those creators in the first instance. Only time will tell whether this works out. But that is the latest news from LinkedIn that they are trying to fight against AI slop.

A slightly odd addendum to that point about AI slop is that I think LinkedIn systems are having a bit of a wobbly moment because in the past couple of days numerous members of my Espresso+community have told me that they’re getting warnings via email from LinkedIn about comments and posts that were written some time ago that have now been flagged for what LinkedIn call “nonconsensual intimate imagery”. And everyone who has check this has found that they haven’t posted anything remotely inappropriate.

So, LinkedIn’s AI detection systems obviously not quite working the way they are intended because these are false positives. And I’m sure LinkedIn’s support department has been flooded with complaints about that.

So, if you are affected, I don’t really think there’s probably much for you to do and hopefully the situation will right itself. But yes, one to look out for.

Last point for this week is to mention that I have shared a tool that I created some time ago that will let you see when any LinkedIn post or comment was published on the LinkedIn platform. So, the exact date and the exact time down to the second of when that thing went out.

LinkedIn did say relatively recently that they were working on putting that data direct into post so that everyone could see it, but they haven’t done that. So, I’ve made my own thing. And in fact, I made this actually a few months ago and shared it with the Espresso+ community.

But I thought it was useful enough that I should share it with everyone. If you want to go to jesp.me/date, then you can see this free tool. There’s no signup needed, there’s no cookies, there’s no tracking, none of that.

It’s an internet tool, the way the internet should be: totally free at the point of access. So, go and use that. Maybe bookmark it and it will let you just. If you copy the link to any post or any comment, paste it into this tool, it’ll give you the exact date and time when that was put out, so you might find that useful.

OK, I’ll leave it at that for now, as usual. If you have any questions or feedback, please drop me a line. I’d love to hear your voice notes if possible, and I’ll catch you all again next week.