TL;DR: I exported a year of my own LinkedIn analytics and found a growth story hiding a conversion problem. Twenty thousand impressions, seven thousand people reached, and only two hundred and sixty-two new followers to show for it. Reach was never the ceiling. My profile was leaking, and I fixed it the way I would fix any product with strong traffic and weak conversion. Four leaks, and the first line of every post turned out to matter most. This experience encapsulates the reach trap my year of libkedin data.
The best thing I wrote all year was one sentence about AI catching what people fake. It out-traveled every careful, well-built post I labored over for hours. For a few days, I felt good about that. Then I opened the analytics export, and that good feeling did not last through the afternoon.
Twenty thousand impressions. Seven thousand six hundred people reached. Two hundred and sixty-two new followers.
I had been reading the top number like a scoreboard. Impressions up. Reach up. A February where everything jumped and mostly stayed jumped. If you only look at the traffic line, it reads like a win. But I build products for a living, and I know the shape of a lot of people arriving and almost none of them staying. This isn’t a traffic issue; it’s a conversion issue. I spent a year focusing on the wrong part of the funnel, feeling proud of the wrong accomplishments.
Here is the good-faith version of how anyone lands here, because I did. LinkedIn shows you impressions every time you open the app. It congratulates you on your reach. The whole surface is built to make you feel at the top of the funnel, because that’s what keeps you posting. Nobody is trying to trick you. The incentive just quietly points you at the number that is easiest to grow and least connected to anything you actually want.
So if you have been writing into that same machine and wondering why your follower count crawls while your view count climbs, let me take one explanation off the table first. You have not lost your voice. Your ideas are not the problem. The reach is proof that people show up. What is missing is everything that happens after they arrive, and that part is mechanical, which means it is fixable.
I have come at conversion before from the other direction, in the Product Onion, where the argument was that you build from the inside out, and that the surface exists only to earn the next layer. This is the same idea pointed to my own page. So I audited it the way I would audit any product with strong traffic and weak conversion. I found four leaks.
Leak one: the hook that starts too slow
The sharpest pattern in my data is boring to say and hard to live. The posts that won led with their best line, and they led with it immediately.
My top five posts all open the same way. A contrarian or curious first sentence, before the reader ever taps “see more.” One asks what if the screen you are designing is already obsolete. One says stop iterating and start shipping. The one that traveled furthest all year just says AI can now smell a fake. None of them warm up. None of them set the context by clearing their throat. The argument hits immediately, and the reader decides to keep going before the fold hides anything.
This is the same discipline as the first screen of an app. You do not get to explain the value once someone is three taps deep. You earn the second tap with the first screen or you lose them. A LinkedIn post is a product with exactly one onboarding moment, and it is the opening line.
The AI angle matters here, and it matters specifically, because that is what my readers keep rewarding right now and the hooks almost write themselves. When I put AI next to something concrete and human, the food you cook, the customer you are trying to reach, the fake someone is trying to pass off, the first line does the work on its own. The posts that framed AI as a force acting on a world the reader already lives in beat everything else I made. That is not a coincidence I want to admire from a distance. It is a template I want to reuse on purpose.
The fix is unglamorous. Write the post, delete the first two sentences, and see if it starts stronger. It almost always does.
Leak two: the headline that hedges
A headline that lists everything you are is a headline that positions you as nothing. Mine tried to be polite by juggling three identities simultaneously, but it ended up resembling a business card soaked in rain.
The data has an opinion about who is actually in the room. Roughly three in five people who follow me are senior or above. Founders and CEOs show up more than their share. They sit in software, IT services, and consulting, clustered first in Austin and then across the usual coastal cities. That is not a consumer-food audience, however much I love talking about spice. It is a product-and-tech audience that keeps coming back for one thing, which is AI seen through the eyes of someone actually building.
So the headline should say that and stop apologizing. Product leader writing about AI and product, with the company I am building and this journal as the proof rather than the pitch. Positioning is about communicating your value proposition, which defines the one main benefit you offer to a specific person. I previously approached mine like a résumé, a list of credentials, but lists tend to fail to engage or convert.
Leak three: the profile with no front door
When a new visitor lands on my page, curious because a post caught their attention, the page does little with that curiosity. No obvious next step. No proof of the good work up front. No door marked start here. In product terms, I had shipped a landing page with no onboarding and then wondered why activation was flat.
The Featured section is the onboarding, and I had left it empty. It should hold the three posts I already know turn attention into more attention, sitting right where a first-time visitor will see them. Next to them, it should hold one clear, working path into the journal, because a follow is a weak promise and an email address is a real one. This is build to learn, build to earn applied to a profile. The early layers exist to learn what lands, and the outer layer, the part a stranger sees first, is built to earn the subscription.
One honest note that belongs in the body and not somewhere softer. My own signup path was broken for part of this year, and I wrote a whole post admitting it. A great top-of-funnel poured into a broken form is the most expensive mistake on this list because you pay for it with the exact people who wanted in.
Leak four: the cadence you can’t feel
There is a clean before and after in my chart, and its name is February. Before it, I posted when I felt like it, and the numbers stayed small. From February on, I posted with a rhythm, and my reach roughly tripled and then mostly held steady. Nearly all of the year’s follower growth landed after that shift.
Cadence is release cadence. A product that ships on a rhythm builds a trust that unpredictable bursts never earn, even when the bursts are good on their own. Consistency beats intensity. A single viral post does less for a page than a schedule that the audience can start to expect. My data says my best two days are Monday and Thursday, which is specific enough to act on, so I am going to act on it instead of admiring the chart.
What I’m doing this week
I rewrote the headline first, because it is the cheapest change with the widest blast radius, and I made it say one true sentence instead of three hedged ones. I am pinning the three proven posts to Featured with a single fixed subscribe link beside them, and I am not touching that link again until I have watched it work. I fixed the signup path, which I should have done the day I wrote the post admitting it was broken. And I am holding a Monday and Thursday rhythm, Monday for a framework, Thursday for an AI-and-your-world take, because those are the two shapes my readers keep rewarding.
None of this is a growth hack. It is the plainest product work there is. Stop celebrating the visits and start fixing the funnel. The reach was never the achievement. It was raw material, and I had been letting it run through my fingers.
What to steal
I ran this whole audit with Claude, not by hand. If you want to point it at your own page, here is the exact prompt. Export your LinkedIn analytics first: go to your profile, open Analytics, and download the spreadsheet. It includes everything the prompt needs: your impressions, your daily engagement, your top posts, your follower growth, and your audience demographics. Drop that file in and paste this:
You are a product manager auditing my LinkedIn presence like a product with a funnel.
I'm attaching my LinkedIn analytics export (the spreadsheet LinkedIn gives creators:
impressions, engagement by date, top posts, followers, and demographics). Read it and
tell me where my profile is leaking. Be blunt, and don't flatter the numbers.
Work through it in this order:
1. Funnel math. Compare total impressions and people reached against new followers.
Tell me plainly whether I have a reach problem or a conversion problem, and show
the numbers you used.
2. The hook. Take my top posts by impressions and by engagement and find the pattern
in the ones that won. What do their opening lines have in common? Give me the
repeatable shape, not a compliment.
3. Topics. Group my posts into themes, then rank the themes by average impressions and
average engagement. Tell me which topic my audience actually rewards and which one
I should stop leading with.
4. Positioning. From my demographics (seniority, industry, location), tell me who is
actually in the room. Then draft one headline: a single sentence that names the one
thing I'm about for that specific person, with every hedge removed.
5. Cadence. From the engagement-by-date data, tell me my best two days to post.
Finish with a prioritized fix list: the biggest leaks between reach and conversion,
cheapest change first.
Change one thing before you run it. Tell it what your conversion actually is. Mine is newsletter subscribers, so I asked it to treat a subscription as the win. Yours might be a follow, a reply, a demo request, or a job note. Name that up front, and the whole audit points to the right thing.
Related reading from PMJ
- The Product Onion Framework: Why Most Businesses Build from the Outside In. The inside-out argument this whole audit borrows, where the surface only exists to earn the next layer.
- Build to Learn, Build to Earn. Why the early layers of anything are built to learn and the outer layer is built to earn, applied here to a profile.
- What If the Screen You’re Designing Is Already Obsolete?. The AI-angle post that proves a first line can carry a whole piece.
- AI Can Now Smell a Fake. My furthest-traveling post of the year, and the clearest case for AI-next-to-something-human as a hook.
FAQ
What is the reach trap? It is treating impressions and reach as the goal when they are only the top of the funnel. High reach with low follower or subscriber growth is not a growth story. It’s a conversion issue disguised as a growth narrative.
Isn’t reach still worth chasing? Reach is the raw material, so it matters, but it is not the finish line. If people are already arriving and not staying, more arrivals will not fix the situation. Spend the next unit of effort on conversion, not on more impressions.
How do I know if my hook is too slow? Read only your first sentence, the part visible before “see more,” and ask whether it would make a stranger tap to read the rest. If the real argument shows up in sentence three, cut the first two and start there.
Why lead with an AI angle specifically? For my audience of product and tech people, AI, when paired with something concrete and human, is the content that travels furthest right now. It is not a law of the universe. It is what my data says about my readers, and the point is to find and lean into whatever your own version of that is.
I’m not a founder with a newsletter. Does this still apply? Yes. Swap subscriber for whatever your real conversion is, a follow, a reply, a portfolio click, a job note. The audit does not change. Find where interested people arrive, then fix the step where they leave.
What’s the one thing I should do this week? Rewrite your headline into a single sentence that names the one thing you are about for one kind of person, and delete every hedge. It is the cheapest change with the widest reach, and it makes every other fix easier.
If you want the frameworks I actually use for this kind of work, grab the free Startup PM Toolkit, then reply and tell me which leak you found first on your own page.
Dan
Leave a Reply