6 things about LinkedIn B2B marketers should know

Events can go one of two ways: you either come back buzzing with ideas that actually change how you work, or you come back with a notebook full of things you already knew, dressed up in new slides. Uplift Live was the former.

Here’s what I found most useful, surprising, and immediately applicable. If you use LinkedIn as a marketing and sales channel, there’s something here for you.

Use DMs to cut through (it boosts you on feeds)

John Espirian’s session on LinkedIn direct messages was, for me, the single most immediately actionable talk of the day. The core argument: DMs have been quietly doing something that posts cannot.

2 things stood out:

• DMs are immune to the algorithm: When you send a message, it lands – no feed ranking, no suppression. You have a higher chance of being seen than in any other LinkedIn format.
• DMs give you an algorithmic boost on feeds: This is something I hadn’t fully appreciated. Using DMs actively makes your other content more visible, not just your conversations.

On the practical side, the easiest way to build a DM habit is to send a note with every connection request (since free LinkedIn limits the number of notes you can add to invitations). This creates a natural cadence without requiring a separate strategy.

A note on InMails – they’re useful, but only if you use them to be genuinely helpful and specific. The association with cold outreach spam means anything generic gets mentally filed under “noise.”

It’s also worth flagging that DMs aren’t private in the way WhatsApp is. But there is an upside, in that you can download your full sent and received message archive and analyse what has worked. That is, as John pointed out, a genuinely good use case for AI.

Use brain power

Phill Agnew’s talk was a highlight. He presented evidence from his own podcast-building experience (and his Nudge Podcast is one of my favourites) that behavioural science nudges apply directly to LinkedIn content and advertising – and work well.

The nudges he discussed included:

• Costly signalling: Demonstrating effort or investment as a proxy for quality or commitment
• The halo effect: Association with trusted names or platforms increases perceived credibility
• Input bias: People value things more when they perceive effort has gone into them
• Social proof: Evidence that others find something valuable
• Avoiding reactance: Giving people a sense of autonomy reduces resistance to requests people instinctively feel when they sense their freedom of choice is being constrained. You can sidestep it by signalling that the person is “free to refuse” – which, counterintuitively, makes them more likely to say yes

I had a recent personal encounter with this principle. My son used it on me when asking for extra video game time. He made it clear I could absolutely say no. I said yes – and gave him 30 extra minutes. Effective nudge.

The broader point – and this applies directly to LinkedIn content – is that the principles persuasion that researchers have spent decades studying don’t stop working just because the medium is social media. If anything, they become more useful in shaping short-form content because you have fewer words to work with – and each one has to earn its place.

AI helps, but not with comments

Heather Murray’s session on using AI effectively for LinkedIn content was probably the most practical grounding of the day – and a useful corrective to the assumption that AI makes content creation faster.

The key points:

• AI isn’t quick: Setting up an agent, project or prompt framework that produces genuinely usable content requires hours of upfront investment. You can’t cut corners on this and expect good results
• Tell-tale AI mannerisms actively hurt your LinkedIn visibility: Phrases like ‘The best part?’ and ‘If X, then Y’ constructions are now so associated with AI-generated content that they can signal inauthenticity immediately – to readers and, potentially, to the algorithm
• You still need to edit: Even with a well-configured AI setup, copy-paste isn’t a viable workflow. The human in the loop remains essential
• And – obvious, but worth saying – don’t be the person who does AI-generated comments

This tracks with what I see in B2B content right now. AI raises the floor – it makes it easier to produce something serviceable. But serviceable isn’t the goal. The goal is content that actually cuts through: that sounds like a specific person, makes a specific point, and gives the reader something they didn’t have before. That still requires human insight.

LinkedIn isn’t one platform, it’s many

Mic Adam and Guy Strijbosch gave an eye-opening session on how LinkedIn use varies across cultures. If you operate internationally – or are thinking about it – this matters more than most people realise.

The highlights:

• There’s enormous variation in how different cultures use LinkedIn: In Brazil, content is more personal and resembles Facebook use more than it does in Europe. What works in one market can fall flat – or worse, seem tone-deaf – in another
• Never mix languages in a single post: It confuses the algorithm and typically delivers worse results than posting separately in each language
• Consider separate profiles for different language markets: Even if all your actual content is in English
• Use the right salutations and titles for the cultural context: People do this for in-person meetings and email but overlook it entirely on LinkedIn

If your audience spans markets, your LinkedIn strategy needs to reflect that. A single content approach won’t serve multiple cultural contexts well.

Your goal is relatability, not polish

Nicole Osborne built on the behavioural science theme by discussing the pratfall effect – the idea imperfection (a stumble, an admission, a moment of honesty about what didn’t go to plan) actually makes you more likeable, not less.

On LinkedIn, the application is straightforward. Content that’s too polished can read as inauthentic, which doesn’t build trust or leave a lasting impression. The slightly rough edge, the admission you’re still working something out – are what makes a post memorable.

This doesn’t mean showcasing manufacturing vulnerabilities or performance issues. It means being a real person rather than an over-curated one. Dave Harland made a related point about humour on LinkedIn, that it can play a role in making content land but it needs to be authentic. Because forced wit is immediately recognisable – and immediately off-putting.

LinkedIn is much higher up the funnel than most people treat it

Tony Restell and Brenda Meller made a point that I suspect many people in the room needed to hear: you only win more business through LinkedIn by integrating it with your broader sales approach.

LinkedIn is a trust-building tool, not a closing tool. The people who try to use it for direct sales conversion – pitching in DMs, leading with offers, measuring results against immediate pipeline – are measuring the wrong thing and wondering why it doesn’t work.

LinkedIn builds awareness and keeps you front of mind. You become the person someone thinks of when a particular problem surfaces, which might take months. But when that moment arrives and your name comes to mind, the relationship is already half-built.

Therefore, your LinkedIn content strategy needs to be patient, focused on consistent, useful, authentic posting over time.

Trust does the heavy lifting

What struck me about Uplift Live is how consistent the underlying message was, and it’s not rocket science: the fundamentals of human connection – trust, authenticity, usefulness, patience – cut through the algorithm and AI.

The people who treat LinkedIn as a distribution mechanism for content will get middling results. The people who focus on the relationship – with what the person on the other side of the phone/browser needs to think or feel or know – and work backwards from there will get the results that make LinkedIn worth the time.