Chapter 03 · Top performing posts · read 2026-08-20
Twenty posts, read off the live feed, sorted by what they actually earned.
Not a roundup. Every row was opened and read on 20 August 2026. Sort by the last column and the category’s preference is unambiguous: it rewards mechanism and people, and it ignores product copy.
9 Instagram posts
11 LinkedIn posts
Sortable
Filterable by format
The board
Every post we read, in one table
Filter by platform or format with the chips; sort any numeric column by clicking its header. The gold rows are the two ceiling posts — the highest-performing brand post and the highest-performing personal post we found.
Three new directors, a Chief Commercial Officer and an advisory-board appointment
Text + link
746
11
14.7
Read off the live posts and profiles on 2026-08-20. ‘Engagements’ is likes plus comments on Instagram and reactions plus comments on LinkedIn, where both were rendered; where a comment count was not shown, the reaction or like count alone is used, which understates those rows. ‘Per 1k followers’ is only computed where the account’s follower count is public — independent injector and practice accounts are left blank rather than guessed.
The pattern
Four findings that hold across both platforms
Mechanism beats results. People beat pages. Product copy loses to everything.
The spread inside a single brand’s own account is larger than the spread between brands. Merz’s science reel earned 12.1 per thousand; Merz’s trade reel earned 2.0. Same account, same quarter, 6× apart. What changed was who the copy was addressed to.
The ceiling is an animation, and it is 4.7× the next brand post. Galderma’s three-layer cross-section at 56.8 per thousand against Merz’s science reel at 12.1. Both are “science” content. One shows the mechanism; the other asserts it.
Individuals out-earn the organisations they work for. Galderma’s chair, writing in the first person about a symposium, pulled 225 engagements. A Wyss scientist introducing her own programme pulled 190 on video. Both beat most of what their institutions posted — and thought-leader distribution runs about 1.6× the engagement rate of the same content on a brand page.
Absolute engagement in this category is small, and that is the good news. MiMedx’s best post in seven months: 236. Aroa Biosurgery’s: 32. Conexeu’s Nasdaq film already did 82. The regenerative-tissue field is not a crowded auction — it is an under-competed one, where a genuinely good post is visible immediately.
Static is over. Every single-image post we read finished in single digits: 7, 6. Both had substantive science in the caption. The format, not the content, was the failure.
Where the two platforms genuinely differ
Instagram
LinkedIn
What it ranks
Completion rate and watch time
Dwell time and comment depth
Best format, 2026 data
Reels for reach (~60–70% of the mix); carousels for saves at 0.50% engagement against Reels’ 0.48%
Native document posts at ~7.00%, then multi-image at ~6.45%, then video at ~6.00%
Who wins here
Independent injectors and patients
Named individuals with credentials
Conexeu’s realistic job
Own the mechanism visual. Nobody in the regenerative ring is even trying.
Own the preclinical-progress narrative and convert the founders into the channel.
Healthcare / pharma / biotech benchmark
1.3% Facebook, format-dependent on IG
3.3% engagement
Conexeu today
695 followers, 34-day cadence gaps
746 followers, median post at ~3.9% — already above benchmark
Format benchmarks from published 2026 datasets (Socialinsider Q2 2026; Buffer’s 52-million-post analysis; Social Media Today on LinkedIn document posts; LinkedIn industry benchmark sets). Conexeu figures read live on 2026-08-20. Note that published benchmarks disagree materially on the carousel-versus-Reel gap — Socialinsider puts it at 0.02 points, Buffer at more than double — because they measure against different denominators. Both agree on the direction.
The three to study
Three posts worth copying the structure of
Instagram
@aestheticsbygalderma
Reel · animated cross-section · 4 Sep 2025
Let’s go deep with the first proven regenerative biostimulator. Sculptra works across all 3 layers of the skin.
Likes
4.3K
Comments
37
Per 1k
56.8
Why it worked
The verb is go deep, and the film delivers on it literally — one continuous descent through named anatomical layers. The promise and the execution are the same gesture.
It gives the finding away rather than teasing it. No link, no “learn more at the link in bio” carrying the payload.
4.3K likes against 37 comments is the signature of a post people watched to the end rather than argued about. Watch time is what got it distributed.
Conexeu’s version
This is Format 01, and Conexeu’s version is stronger on the merits: a scaffold being populated by host cells is a more visually legible event than collagen stimulation, because there is an actual structure on screen. Same descent, three named layers, ending on the lattice filling in.
The menopause symposium at AMWC this morning was a full room, with a clear focus on evidence.
Reactions
225
Format
Text
Assets
None
Why it worked
No video, no design, no budget. A named senior scientist reporting what he saw in a room that morning, then citing the survey behind it — 4,300 women, five continents.
The structure is: observation, then the evidence gap it closes, then what that changes clinically. It reads as someone thinking, not someone announcing.
It also does something a brand page structurally cannot: it takes a position. Clinicians “lacked a robust evidence base” is a mild criticism of the field, and mild criticism is what makes a post worth reacting to.
Conexeu’s version
Dr Chavez-Munoz has a stronger version of exactly this available every month: what the Wake Forest programme is actually showing, what the field still cannot answer about post-mastectomy reconstruction, and what changes if a scaffold works. Zero production cost, posted from her own profile.
Bio-stimulators aren’t about INSTANT volume, they’re about teaching your skin to act young again.
Likes
525
Comments
25
Budget
None
Why it worked
One injector with no brand backing beat @revanesse (31k) 8× and @merzaesthetics_us’s trade reel (40.9k) 6.6× on the same subject.
The whole post is a single reframe delivered in the first sentence, and the sentence names the misconception before correcting it — “aren’t about instant volume”.
25 comments on 525 likes is a much higher argument rate than the Galderma animation. Reframes get discussed; demonstrations get watched. A programme needs both.
Conexeu’s version
This is Format 04 and it is the cheapest thing on this page to test. Conexeu’s reframe is one step larger than hers: not “biostimulators are misunderstood” but “there is a third job that neither filling nor stimulating does”.
One correction to the row above: the comment-to-like ratio comparison is between @beckerinjectorbrooke’s reel and the Galderma animation, not against the category as a whole — comment counts were not rendered on several rows in the read set.
