Content Strategy
I pulled my LinkedIn analytics for the past four weeks and ranked every post two ways: by engagement and by impressions. I expected the top of each list to look about the same. Instead, the post with the most reach barely cracked engagement, and the post with the most comments and shares barely cracked reach.
One post, about exposure management for chief information security officers (CISOs), pulled 24,394 impressions and 78 engagements. Another post, about being named a Distinguished Analyst for 2027, pulled 1,191 engagements on 6,211 impressions, a quarter of the reach and fifteen times the interaction.
Impressions Reward the Topic. Engagement Rewards the Person.
LinkedIn's algorithm decides who sees a post before anyone reads it. It favors posts that match what a professional audience already searches for: cybersecurity and AI infrastructure, or a named vendor in the news. Those posts travel outside my own network and rack up impressions from people who have never followed me.
Engagement runs on a different signal. A comment or a share is a person choosing to react, and people react to other people. The posts that pulled the most engagement were personal: a visit to Asana's San Francisco headquarters, a congratulations note for a colleague, a reflection on research and career growth. None of those posts cracked the impressions top ten.
A Few Posts Do Both
The Distinguished Analyst post landed first in engagement and fourth in impressions. An identity infrastructure and AI security post landed sixth in engagement and second in impressions. A post on agentic AI in enterprise work management landed seventh in engagement and third in impressions.
A comment or a share is a person choosing to react, and people react to other people.
What ties those posts together is specificity. Each one named a real event or a real person, not a general observation about the industry. That specificity works for the algorithm and for a reader deciding whether to comment.
Two Metrics, Not One
I stopped grading every post against both numbers. The topical, technical posts earn their keep through reach: they put the account in front of people who don't already follow me. The personal and relationship posts earn their keep through engagement: they deepen the relationship with people who already do.
One metric chases new eyes. The other chases people already paying attention. Both belong in the calendar. Neither should be judged by the other's yardstick.
The Full List, Ranked
Two rankings, pulled from the same four weeks of posts.
Top by Engagement
| 01 | Being named a Distinguished Analyst for 2027 | 1,191 eng. |
| 02 | A reflection on research and career growth | 518 eng. |
| 03 | A day at Asana's San Francisco headquarters | 163 eng. |
| 04 | Dynatrace's push into AI observability | 134 eng. |
| 05 | A congratulations note for a colleague | 132 eng. |
| 06 | Identity infrastructure and AI security | 126 eng. |
| 07 | Agentic AI in enterprise work management | 122 eng. |
| 08 | Cloudflare's FedRAMP public sector authorization | 118 eng. |
| 09 | A Talking Headless Live episode on analyst relations | 98 eng. |
| 10 | Exposure management for CISOs | 78 eng. |
Top by Impressions
| 01 | Exposure management for CISOs | 24,394 impr. |
| 02 | Identity infrastructure and AI security | 8,060 impr. |
| 03 | Agentic AI in enterprise work management | 6,328 impr. |
| 04 | Being named a Distinguished Analyst for 2027 | 6,211 impr. |
| 05 | A day at Asana's San Francisco headquarters | 5,615 impr. |
| 06 | Forward deployed engineering in enterprise AI strategy | 5,053 impr. |
| 07 | Adobe's CEO transition | 5,032 impr. |
| 08 | A reflection on research and career growth | 4,986 impr. |
| 09 | GIS and AI strategy for CIOs | 4,916 impr. |
| 10 | Data protection and cyber resilience | 4,202 impr. |
What to Do Monday
| 01 | Pull your own LinkedIn analytics export and rank posts by engagement and by impressions separately. Look for where the two lists disagree. |
| 02 | When you want reach, write about a named topic or vendor your network already searches for. |
| 03 | When you want engagement, write about a specific person or moment instead of a general industry observation. |
| 04 | Cite any number you pull from another source in MLA format, right where you use it. |
I ran the same comparison on Shashi.co, ranking posts by on-site pageviews and by Google AI Overview citations. The two lists disagree there too.