Taylor Bordeneditor at LinkedIn, emailed me last week with a question she’s asking a handful of writers for a special edition of her newsletter, The Work Shift. This hypothesis is supported by data showing that entrepreneurship on LinkedIn has increased nearly 70% year over year, that more than six in ten entrepreneurs also identify as content creators, and that people who post weekly see up to 4 times more profile views, with comments generating 2.5 times more.
His question was simple: what is the lesson that changed your approach to content creation? And if you were starting your LinkedIn journey from scratch, how would you approach your first 10 posts?
I almost responded with a frame. Then I remembered why frameworks are the problem.
4 categories felt complete. Data in disagreement
Around 2009, Guy Kawasaki asked me for a few pages for his book “Enchantment: The Art of Changing Hearts, Minds, and Actions.” I’ve outlined four ways brands could create YouTube videos that will truly delight an audience: inspire viewers with emotional stories, educate them with useful information, enlighten them with documentaries, or entertain them by making them laugh.
Four seemed complete. It was clean, informative and easy to remember. I used it. Other people have used it. I even included it in an article for Search Engine Journal years later, “What is a content marketing matrix and do we need one?»
Then the data kept coming. In 2023, I wrote another SEJ article with not four, but 39 emotions – count them.
I had never connected these two until Borden’s email forced me to do so. The gap between them, 14 years and 35 emotions, is the most useful thing I’ve learned in 24 years of writing about this industry. The four-category framework wasn’t wrong when I wrote it. It was simply the size of the dataset I had access to at the time. The mistake would have been to consider it finished.
Practitioners who get stuck are those who fall in love with their framework
This is the part of my response to Borden that directly applies to anyone currently doing SEO, content marketing, or social media marketing work, not just posting on LinkedIn.
Every framework you build, every category system, every “four types of X” or “five steps of Y” is a snapshot of what the evidence showed you the day you built it. AI insights didn’t exist when most of our content frameworks were written. Neither does AI mode, Search integrated into GeminiOr Insights of AI appearing in the results of 2.5 billion users. The frames built for a 10 blue link world weren’t bad for that world. This is simply the size of the dataset that existed at the time.
The practitioners who get stuck are those who continue to apply the 2019 framework to 2026 data because the framework is familiar and the new data is impractical. Those who continue to grow are those who remain curious enough to ask themselves, “What would this framework look like if I rebuilt it today, with everything I know now that I didn’t know then?”
This is exactly the trap that many Content of the AI presentation the strategy is being implemented at the moment. The “Answer the query in 40 words at the top of the page” framework was designed for a world where the goal was to win a featured snippet. This setting was not bad for this world. But AI insights don’t reward the page that has already said it all; they reward the page a user clicks on after the presentationand they reward him for being more than the summary that sent them there. A page designed to win over the old framework is, by design, the page that no longer has anything to offer that user. The four-category model and the 40-word response model failed for the same reason; both were finished products built for a dataset that continued to grow after the deadline.
What I would say to anyone starting their first 10 posts
This is the answer I gave directly to Borden, and it’s the same advice I would give to anyone in SEO, content marketing, or social media marketing starting from scratch, on LinkedIn or elsewhere.
Find something you believe in with confidence. SO find the research that complicates things. Write honestly about the discrepancy, including the part where you were wrong or incomplete.
This single movement does three things at once. It gives you a topic (your existing belief), it gives you a hook (the data that challenges it), and it gives you credibility that a polished, unchallenged framework never can, because readers can tell the difference between someone defending a position and someone genuinely updating it.
2 steps to apply it this week
First, display the oldest “X types of Y” frame, list, or article you’ve published, the one you’re most proud of, the one that’s still cited or linked to. Find what has been published on this specific topic in the last 12 months. If a four-category framework from 2009 were to quietly become 39 by 2023, everything you wrote in 2019 or 2021 almost certainly has a a similar gap awaits in 2026 data. Don’t defend the old version. Write the article that updates it and explicitly say what changed and why.
Second, before posting anything presented as “X ways to do Y”, ask yourself whether you are presenting a snapshot or a conclusion. One snapshot says: “Here’s what the evidence shows right now, and I would expect this number to increase. » A conclusion says: “Here is the complete list. » The first framing is aging well. The second framing is the one that will have to be returned to in front of an audience, as I have just done with my own framing from 2009, in public, 14 years later.
The entrepreneurial data shared by Borden, the 70% growth, 4x profile views for weekly posters, isn’t really about LinkedIn specifically. It’s proof that more and more people are now doing what editors and SEO practitioners have always done, which is to make the public believe and find out, often quickly, whether the evidence supports that belief. The lesson is the same in all cases. Stay curious about what the data says next, especially when it disagrees with the framework you’ve already published.
More resources:
Featured image: Roman Samborskyi/Shutterstock
