open/ktg-plugin-marketplace
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity Actions

docs(linkedin-studio): Voyage remediation setup — brief + research + plan (Phase 0-3)

Browse source 
Audit-remediation Voyage project authored end-to-end this session:
- brief.md (reviewer PROCEED; validator pass) — full Phase 0-3 scope, phased,
  with success criteria refined by research
- research/01-03 — high-effort external swarm + Gemini (Topic 1); reconciled the
  external bar and corrected several audit feature-premises (no publishable model
  name/date; saves UI-visible not API-pullable; auto-publish possible-not-built;
  9:16 not mandatory; newsletter notifications deduplicated not triple; CLI crash
  = missing npm install, depth-bug latent)
- plan.md (21 steps, 7 sessions, 5 waves; validator pass; A- 88/100) — plan-critic
  REVISE (3 blockers + majors) addressed; scope-guardian ALIGNED; gemini Pass-2
  folded in 2 blind spots (git-history decision; lint stat-grep sequencing)

Execution is future sessions (one wave each) via /trekexecute, /trekreview as the
release gate. Audit report stays local until the article ships.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Kjell Tore Guttormsen 2026-05-29 19:49:27 +02:00
commit a61b818578
6 changed files with 2082 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

310
plugins/linkedin-studio/docs/remediation/research/01-linkedin-algorithm-signals.md Normal file
View file

@ -0,0 +1,310 @@
---
type: trekresearch-brief
created: 2026-05-29
question: "What does the 2026 LinkedIn feed-ranking system actually reward — comment-vs-reaction weighting, document/carousel engagement rate, external-link reach effect and first-comment status, the early-engagement window incl. delayed reinjection, and the deployed ranking model's verifiable name and date — with a source and confidence per claim?"
confidence: 0.82
dimensions: 8
mcp_servers_used: [tavily, gemini-deep-research]
local_agents_used: []
external_agents_used: [docs-researcher, community-researcher, security-researcher, contrarian-researcher, gemini-bridge]
---
# 2026 LinkedIn Feed-Ranking — Canonical Signal Statement
> Generated by trekresearch (high-effort swarm: 4 external + Gemini) on 2026-05-29.
> Topic 1 of 3 for the linkedin-studio remediation. This is the **substrate**: the
> Phase-0 fixes that reconcile the plugin's contradictory algorithm stats consume it.
## Research Question
What does the 2026 LinkedIn feed-ranking system actually reward — comment-vs-reaction
weighting, document/carousel engagement rate, external-link reach effect and the
current first-comment-workaround status, the early-engagement ("golden hour") window
incl. delayed/evergreen reinjection, and the deployed ranking model's verifiable name
and deployment date — with a primary or credible source and a confidence level per claim?
## Executive Summary
The plugin's algorithm "facts" are **directionally right but numerically indefensible**:
every specific magnitude it states (comment "15x", carousel "6.6%"/"1.92%", link
"40-50%"/"25-40%", a clean "−40-60% before distribution", "360Brew, January 2026") is
either third-party-only, self-contradictory, conflated across denominators, or — for the
model name/date — **not establishable from any primary source.** What IS defensible and
high-confidence: an LLM-based relevance-ranking system is live in 2026; the engagement
hierarchy is **saves > shares > quality comments > reactions** with **dwell-time a
top-tier signal** (the only two signals LinkedIn officially confirms by name are *dwell
time* and *topic/interest relevance*); documents/carousels are the #1 format; body links
reduce reach (magnitude contested, ~19–60% across studies, LinkedIn denies it is
*intentional*); the early window is **60–90 min** (90 is the 2026 consensus); and — the
single best-supported actionable finding — **LinkedIn now officially suppresses generic
AI "slop"** (named executive, May 2026), which directly justifies a short-form de-AI gate.
**Key caveat:** treat every number as directional and per-account-testable; encode
*ordering + sourced direction*, never hard coefficients. (Overall confidence 0.82 — high
on direction, medium on magnitude.)
## Dimensions
### D1. Deployed ranking model — name & date — Confidence: high (on the negative claim)
**External findings:**
- The arXiv paper *"360Brew: A Decoder-only Foundation Model…"* (2501.16450) is dated
**2025-01-27**, self-labels as a **"research pre-production model" (V1.0, 150B params)**
claiming *offline* parity only, and was **withdrawn 2025-08-23** (submitter lacked
license rights). It is neither a deployment announcement nor a clean citable artifact.
[arXiv 2501.16450]
- LinkedIn's own 2026 communications describe a live LLM-based feed system but the
**production name is not reliably establishable**: the docs + contrarian agents both
read a LinkedIn Engineering post ("Generative Recommender / GR", attributed to Hristo
Danchev, 2026-03-12); the independent Gemini pass **flagged a third-party citation of
that same post as possibly fabricated** (Danchev's verifiable authorship is on AWS
OpenSearch work). So even the "GR" name carries a provenance question.
- "January 2026" as a deployment date appears in **no** primary source; it is third-party
extrapolation from the paper's Jan-**2025** date.
**Contradictions:** docs/contrarian treat the GR engineering blog as primary; Gemini
casts doubt on its provenance. **Conservative resolution:** assert neither name nor date.
An LLM relevance-ranking system is live (high confidence); its *deployed name* and
*go-live date* are **not publishable as fact**.
### D2. Comment vs reaction weighting + saves/dwell hierarchy — Confidence: high (ordering) / medium (magnitude)
**External findings:**
- "Comment = 15x a like" is **unverified folklore** — no primary source; meet-lea labels
it "industry estimate, original source unclear." Sources span 2x–15x with no anchor.
AuthoredUp's NLP-quality-scored analysis puts the real comment-vs-like effect **~2x**.
[authoredup.com/blog/linkedin-algorithm; meet-lea]
- Convergent across AuthoredUp + Vertebrae + van der Blom (1.8M): **a save ≈ 5x a like,
≈ 2x a comment** — saves are the top signal (and a follow-graph signal: saving a post
gives the author's next post ~80% feed-appearance odds). The plugin's stray "5x" is the
**saves** number mis-assigned to comments.
- **Officially confirmed (the only two named):** *dwell time* is a ranking signal
(LinkedIn Eng "Understanding feed dwell time" 2020; "Leveraging Dwell Time" /
Auto-Normalized-Long-Dwell model 2024); LinkedIn describes active (like/comment/share)
vs passive (click/skip/long-dwell) tasks but **assigns no weights**. [linkedin.com/blog/engineering/feed/leveraging-dwell-time-to-improve-member-experiences-on-the-linkedin-feed]
**Resolution (for the canonical statement):** order is **saves > shares > quality
comments > reactions/likes**, with **dwell-time top-tier**; comment ≈ 2x like
(quality-weighted, single-vendor). Drop "15x" and the comment-"5x" entirely.
### D3. Document/carousel engagement rate — Confidence: high (format rank) / medium (number)
**External findings:**
- Three independent large-N studies agree documents/carousels are **#1**: Socialinsider
(1.3M) native document **7.00%** (multi-image 6.80%), Buffer (2M) carousel **21.77%**
median, Metricool (673K) **49.52%**. The 7 vs 21.77 vs 49.52 spread is a
**denominator/methodology artifact**, not disagreement about the winner.
[socialinsider.io/social-media-benchmarks/linkedin; buffer.com/resources/data-best-content-format-social-media/; metricool.com/linkedin-trends/]
- The "6.6%" is a **stale 2024 multi-image** figure (now ~6.45% multi-image / ~7.00%
document) — and LinkedIn removed native carousels Dec 2023, so "carousel" = PDF document
post; the multi-image↔document conflation is real.
- **The plugin's "1.92%" is NOT a carousel rate** — it matches the **personal-profile
per-post baseline** (Metricool personal 2.60% / company 1.74%; AuthoredUp 2.10–2.67%).
The plugin mixed a format benchmark with a personal-profile baseline.
**Resolution:** documents/carousels = top format (high confidence). For a number use
**~7% (Socialinsider, conservative, company-page per-impression)**; never present 1.92%
as a carousel figure; state the format-vs-account-type distinction.
### D4. External-link reach effect + first-comment status — Confidence: medium (effect) / low (intent, first-comment)
**External findings:**
- A body-link reach reduction is real and observational. The most rigorous source
(Ordinal, 900K posts, Mann-Whitney p<0.001) shows it **changed over time: 5% (2023) →
35% (2024) → 42% (2025) → ~38% (2026 YTD)**, 37-month avg 26.5%. van der Blom reports a
milder **~18.8% median**; DigitalApplied/Gemini cite **~60%**. So the plugin's "40-50%"
≈ the 2024-25 peak and "25-40%" ≈ the long-run average — **both partial views of one
moving number.** [tryordinal.com/blog/linkedin-link-penalty-study]
- **LinkedIn denies an *intentional* penalty** (Sr. Director Product, reported Aug 2025):
no penalty "if the post leads with value"; the effect is engagement-driven, not a flat
tax. The observed reach gap is real **regardless of intent**. [threads.com/@mattnavarra/post/DOWa_61Cown/]
- First-comment workaround is **genuinely contested**: Ordinal data leans "still
net-positive but reduced (~−5 to −10%)"; multiple 2026 blogs claim it's now detected as
"bridge behavior" and throttled — but that claim is **practitioner-only, no large-N
backing.** The one officially-confirmed principle: what gets limited is
**off-platform-funnel intent + thin standalone value**, *regardless of link location*.
**Resolution:** state it as a **correlational reach reduction (~38% in 2026, contested
band ~19–60%, LinkedIn disputes intent)**, not a hard penalty. Reframe first-comment as
**neither a magic fix nor a confirmed penalty** — lead with standalone value; native
formats are the durable answer. Drop the precise % from the enforcing hook.
### D5. Early-engagement window + evergreen reinjection — Confidence: high (60-90 min) / low (24-72h timing)
**External findings:**
- 2026 consensus has widened from "strict 60 min" to **60–90 min** (90 is van der Blom's
current figure), with the **first 15–30 min** the highest-leverage sub-window and ~70%
of reach decided in it. [buffer.com/resources/linkedin-algorithm/; expandi.io/blog/best-time-to-post-on-linkedin/]
- Evergreen resurfacing is **real in direction** (the 2026 relevance model resurfaces
strong-save / high-dwell posts days-to-weeks later on viewer intent; AuthoredUp: posts
now live 2–3 weeks vs days) — but **no large-N source confirms a specific "24–72h
reinjection" rule**; it is intent-driven and irregular.
**Resolution:** "**60–90 min golden window; first 15–30 min highest-leverage**"; describe
evergreen as "**can resurface days-to-weeks later on intent-match**", not a fixed 24–72h
second wave. The plugin both over-indexes the strict first hour AND omits evergreen — fix
both.
### D6. Profile/topic relevance as a ranking input — Confidence: high (signal) / none (the −40-60% figure)
**External findings:**
- **Officially confirmed (qualitatively):** topic/interest relevance drives distribution,
including beyond your network — Tim Jurka (Head of Feed AI, 2025-08-11): "Exceptional
content may even be distributed broadly … to members interested in the type of content
you post, even if they don't follow you." 2026 comms add an Interest Picker + "relevant
to your interests, not a popularity contest." [linkedin.com/pulse/how-does-linkedin-feed-work-tim-jurka-oxraf]
- **No primary source** states any **−40-60% reach reduction** for off-topic content, nor
a discrete "validation-before-distribution gate" with a number. That figure is
third-party.
**Resolution:** keep "profile/topic alignment is a real ranking input" (sourced
direction); **drop the "−40-60% before anyone sees it" figure** entirely.
### D7. Buzzword penalty — Confidence: high (that it is NOT a measured ranking mechanic)
**External findings:**
- **No primary source** ties specific words to a measured reach penalty. Evidence is
either editorial/clarity advice (Inc.) or unmeasured vendor assertion (linkboost
"LLMs throttle corporate speak"). A semantic-relevance ranker *may* indirectly favor
specific over generic phrasing — inferred, not confirmed. [inc.com/...buzzwords; linkboost.co/blog]
**Resolution:** keep buzzword-avoidance as **editorial guidance**, not a "reduces reach"
ranking claim. (The plugin already enforces a buzzword list via a hook — keep the list,
fix the *justification*.)
### D8. AI-content down-rank — Confidence: high (officially confirmed) — *the build-justifying finding*
**External findings:**
- **Officially confirmed, named executive:** LinkedIn VP & Executive Editor Laura
Lorenzetti (2026-05-19) confirmed an active program targeting (1) generic AI-written
posts/comments, (2) automation tools, (3) attention-bait video. Mechanism: ML models
trained on thousands of human-annotated posts distinguish "original thinking" from
"posts lacking substance"; **low-quality-flagged posts are reach-suppressed (reportedly
down to first-degree connections), not deleted.** [entrepreneur.com/business-news/linkedin-is-fighting-back-against-ai-slop-and-ai-comments]
- Corroborated: Jobanputra (Feed) — "we actively detect and limit the reach of spammy or
low-quality content, including bot-generated posts." Originality.ai (8,795 posts):
likely-AI posts saw **45% less engagement** (correlational). [prdaily.com/...guardians-of-the-feed; originality.ai/blog/ai-content-published-linkedin]
- Also officially confirmed and relevant: **engagement-pod crackdown** (VP Product
Gyanda Sachdeva, 2026-02-16 — auto-comments demoted out of "Most Relevant", scoped to
own network, repeat offenders restricted). [socialmediatoday.com/news/linkedin-outlines-more-measures-to-combat-engagement-pods/812290/]
**Resolution:** **build the short-form de-AI / differentiation gate** — it targets an
officially-confirmed suppression surface. Enforce the signals LinkedIn *named* (personal
substance, original thinking, concrete specifics, genuine voice), not an unverified SEO
"tell-list."
## External Knowledge
### Best Practice (official / primary)
Only two ranking signals are officially named: **dwell time** and **topic/interest
relevance**. LinkedIn officially **denies an intentional link penalty** and officially
**confirms an AI-slop down-rank** + **engagement-pod enforcement**. Everything else
(coefficients, multipliers, windows) is third-party.
### Alternatives / contrarian
The contrarian pass refuted 6 of 7 plugin claims **on magnitude/naming, not direction**:
the strategic advice (favor native formats, prompt quality comments, write with
substance, expect link posts to underperform, post when the audience is active) survives;
the specific numbers and the "360Brew, Jan 2026" branding do not. Two need **outright
correction**: the model name/date, and the "no analytics API → CSV only" premise (see D9
in Topic 2 — Member Post Analytics API launched 2025-07-08).
### Known issues
Numbers rot: every magnitude is observational and moves year-to-year (link penalty
5%→42%→38%; carousel 6.6%→6.45%). A fabricated citation ("Hristo Danchev / Mar-12-2026")
is actively circulating — do not propagate any single named-source deployment claim
without first-hand re-verification.
## Gemini Second Opinion
Independent ~22-min deep-research pass (27 grounding sources). Agreements with the swarm:
360Brew is a Jan-**2025** pre-production paper, not a confirmed 2026 production system;
saves/dwell primacy; carousel #1 with methodology-driven rate spread; 90-min window;
**per-post Saves ARE visible in the native UI for your own posts**; a Member Post
Analytics API exists but is gated behind Community Management API approval (not
self-serve). Unique contribution: independently flagged the "Hristo Danchev / March 2026
engineering post" citation as likely **fabricated**, which is *why* this brief refuses to
publish any deployed-model name even though two of the swarm agents cited "GR."
## Synthesis
Three insights emerge only from triangulation:
1. **The plugin's contradictions are mostly denominator/era artifacts, not errors of
fact.** "40-50% vs 25-40%" = the same link number at peak vs average; "6.6% vs 1.92%"
= a format benchmark vs a personal-profile baseline; "15x vs 5x" = a folklore comment
figure vs the real *saves* figure mis-assigned. The fix is therefore **one canonical
statement that names the era, the denominator, and the account type** — not a hunt for
"the right number." This is the single most important design instruction for Phase 0.2.
2. **Encode ordering + officially-named signals, not coefficients.** The only durable,
defensible spine is: *dwell + topic-relevance are the two officially-named signals;
saves > shares > quality-comments > reactions is the engagement order; documents are
the top format.* Every coefficient must carry a source + confidence + "directional,
test per account" caveat. A `references/algorithm-signals-reference.md` rebuilt around
*named signals + ordering + per-claim source column* makes the contradictions
structurally impossible to reintroduce.
3. **The two highest-confidence findings each map to a Phase-2 build decision.** The
officially-confirmed **AI-slop down-rank** justifies the **short-form de-AI gate**
(D8); the officially-confirmed **link-intent principle** (value-first, location-
secondary) rewrites the link advice (D4). Both are now grounded in *named-executive*
sources, not vendor blogs — the strongest evidence in the whole pass.
## Open Questions
- **Deployed model name/date** — unresolvable from open sources and partly contaminated
by a fabricated citation. *Carry as: do not assert; state "an LLM relevance model is
live in 2026" only.* No further research will likely fix this before publication.
- **Link-penalty exact magnitude & first-comment status** — genuinely contested
(~19–60%; first-comment net-positive vs detected). *Carry as a range + "test per
account"; do not hard-code.*
- **Member Post Analytics API self-serve depth** — answered enough here to act, but is the
primary subject of **Topic 2** (verify gating + saves-UI before writing boundary prose).
## Recommendation
For the Phase-0 "reconcile to one sourced statement" step, adopt this canonical spine and
make every command/agent cite it:
1. **Model:** "An LLM-based relevance-ranking system is live on LinkedIn in 2026." **No
name, no date.** Remove "360Brew" and "January 2026" from CLAUDE.md/README/profile.
2. **Signals (officially named):** dwell time; topic/interest relevance. **Engagement
order:** saves > shares > quality comments > reactions; likes ≈ 1x baseline. No
coefficients without a source column; comment ≈ 2x like is the most defensible single
figure (medium).
3. **Format:** documents/carousels are the top organic format (~7%, Socialinsider,
company-page per-impression). Delete the 1.92% carousel claim (it's a personal-profile
baseline). Native video #2 and *declining*.
4. **Links:** correlational reach reduction (~38% in 2026; contested ~19–60%); LinkedIn
denies intentional penalty; value-first matters more than link location; first-comment
is a hedge, not a fix. Soften the enforcing hook from a hard % mechanic.
5. **Timing:** 60–90 min early window (first 15–30 min highest-leverage); add evergreen
resurfacing (days-to-weeks, intent-driven); drop the strict-60-min fixation and the
"24–72h reinjection" precision.
6. **Profile/topic:** real ranking input (keep); **drop the −40-60% figure.**
7. **Buzzwords:** editorial guidance only (keep the list, fix the "reduces reach" claim).
8. **Build the de-AI gate** (D8, officially-confirmed surface) and **reframe link advice
around intent** (D4). Both are Phase-2 builds with named-executive backing.
## Sources
| # | Source | Type | Quality | Used in |
|---|--------|------|---------|---------|
| 1 | [arXiv 2501.16450 — 360Brew (withdrawn 2025-08-23)](https://arxiv.org/abs/2501.16450) | official | high | D1 |
| 2 | [LinkedIn Eng — Engineering the next-gen Feed (provenance contested)](https://www.linkedin.com/blog/engineering/feed/engineering-the-next-generation-of-linkedins-feed) | official(?) | low | D1 |
| 3 | [LinkedIn Eng — Leveraging Dwell Time (2024-10-01)](https://www.linkedin.com/blog/engineering/feed/leveraging-dwell-time-to-improve-member-experiences-on-the-linkedin-feed) | official | high | D2 |
| 4 | [Tim Jurka — How Does the LinkedIn Feed Work? (2025-08-11)](https://www.linkedin.com/pulse/how-does-linkedin-feed-work-tim-jurka-oxraf) | official | high | D6 |
| 5 | [AuthoredUp — LinkedIn Algorithm (621K posts)](https://authoredup.com/blog/linkedin-algorithm) | community | medium | D2, D3, D5 |
| 6 | [Socialinsider — LinkedIn benchmarks (1.3M)](https://www.socialinsider.io/social-media-benchmarks/linkedin) | community | medium | D3 |
| 7 | [Buffer — Best Content Format (2M+)](https://buffer.com/resources/data-best-content-format-social-media/) | community | medium | D3 |
| 8 | [Metricool — 2026 LinkedIn study (673K)](https://metricool.com/linkedin-trends/) | community | medium | D3 |
| 9 | [Ordinal — Link Penalty Study (900K, p<0.001)](https://www.tryordinal.com/blog/linkedin-link-penalty-study) | community | medium-high | D4 |
| 10 | [Threads/Matt Navarra — LinkedIn denies intentional link penalty](https://www.threads.com/@mattnavarra/post/DOWa_61Cown/) | official (relayed) | medium | D4 |
| 11 | [Entrepreneur — LinkedIn fights AI slop (Lorenzetti, 2026-05-19)](https://www.entrepreneur.com/business-news/linkedin-is-fighting-back-against-ai-slop-and-ai-comments) | official (reported) | high | D8 |
| 12 | [PR Daily — Guardians of the Feed (Jobanputra)](https://www.prdaily.com/what-works-and-doesnt-on-linkedin-according-to-guardians-of-the-feed/) | official (reported) | medium-high | D4, D8 |
| 13 | [Social Media Today — engagement-pod crackdown (Sachdeva, 2026-02-16)](https://www.socialmediatoday.com/news/linkedin-outlines-more-measures-to-combat-engagement-pods/812290/) | official (reported) | high | D8 |
| 14 | [Originality.ai — AI content on LinkedIn (45% gap)](https://originality.ai/blog/ai-content-published-linkedin) | community | medium | D8 |
| 15 | [van der Blom — Algorithm Insights 2025 (1.8M)](https://www.scribd.com/document/984921783/Algorithm-Insights-Report-2025-chapter-1-Richard-Van-der-Blom) | community | medium | D2, D4, D5 |
| 16 | [meet-lea — LinkedIn Algorithm Explained 2026](https://meet-lea.com/en/blog/linkedin-algorithm-explained) | community | low-medium | D2 |
| 17 | [Microsoft Learn — Member Post Statistics API](https://learn.microsoft.com/en-us/linkedin/marketing/community-management/members/post-statistics?view=li-lms-2025-11) | official | high | D2/Topic-2 |
| 18 | [Inc. — buzzwords to scrub](https://www.inc.com/amy-george/14-buzzwords-to-scrub-from-your-linkedin-page-right-now.html) | community | low | D7 |

218
plugins/linkedin-studio/docs/remediation/research/02-analytics-publish-boundaries.md Normal file
View file

@ -0,0 +1,218 @@
---
type: trekresearch-brief
created: 2026-05-29
question: "As of 2026, can a personal LinkedIn profile self-serve post-level analytics via an API, are per-post saves visible in the native UI, and can a personal profile auto-publish via any API — with the exact constraints for a solo user without partner/company-page access?"
confidence: 0.86
dimensions: 4
mcp_servers_used: [tavily]
local_agents_used: []
external_agents_used: [docs-researcher, community-researcher, contrarian-researcher]
---
# Personal-Profile Analytics + Auto-Publish Boundaries (2026)
> Generated by trekresearch (standard external swarm: docs + community + contrarian; no
> Gemini — scoped to Topic 1) on 2026-05-29. Topic 2 of 3 for the linkedin-studio
> remediation. Feeds Phase-1 honest boundary statements + the Phase-2 saves honesty-fix.
> **Primary-sourced from Microsoft Learn (LinkedIn's canonical dev docs).**
## Research Question
As of 2026, can a personal LinkedIn profile self-serve post-level analytics via an API,
are per-post saves visible in the native UI, and can a personal profile auto-publish via
any API — with the exact constraints for a solo user without partner/company-page access?
## Executive Summary
**The audit's own boundary assumptions are partly wrong, and fixing them naively would
replace one false claim with another.** Three findings, all primary-sourced and
high-confidence: (1) a personal profile **CAN auto-publish self-serve** — `w_member_social`
is an Open Permission via the free "Share on LinkedIn" product, publishing immediately at
~150 posts/member/day; the plugin's clipboard-stop is a **design + Terms-of-Service
choice, not an API impossibility**. (2) Per-post **saves ARE visible** in native post
analytics (rolled out ~Sept 2025, count-only, no saver identity) **and** exposed via the
API's `POST_SAVE` metric (since version li-lms-2026-04) — so "saves aren't trackable" is
**stale/false**; the honest line is "visible in the UI, not self-serve via API." (3)
Post-level analytics via API **exist** (`memberCreatorPostAnalytics` / `r_member_postAnalytics`)
but sit behind the **vetted Community Management API** (verified organization + associated
LinkedIn Page + use-case review) — **not self-serve for a solo creator**; CSV export is the
practical floor, not the only technical path. **Key caveat:** every boundary statement
must be *dated* ("as of 2026-05") — this surface changed fast (saves went UI→API inside
~12 months).
## Dimensions
### D1. Post-level analytics API for personal profiles — Confidence: high
**External findings:**
- The endpoint exists: `GET /rest/memberCreatorPostAnalytics` ("Member Post Statistics"),
permission `r_member_postAnalytics` (versions ≥ li-lms-202506). Finders `q=entity`
(one post) and `q=me` (aggregated). Metrics: `IMPRESSION`, `MEMBERS_REACHED`, `RESHARE`,
`REACTION`, `COMMENT` (since 2025-06) and — added **li-lms-2026-04** — `POST_SAVE`,
`POST_SEND`, `LINK_CLICKS`, `PREMIUM_CTA_CLICKS`, `FOLLOWER_GAINED_FROM_CONTENT`,
`PROFILE_VIEW_FROM_CONTENT`. (Video metrics are a separate endpoint
`memberCreatorVideoAnalytics`; follower count is `memberFollowersCount` /
`r_member_profileAnalytics`.) [learn.microsoft.com/.../members/post-statistics?view=li-lms-2026-05]
- **The access gate is the crux:** `r_member_postAnalytics` is listed **exclusively under
the Community Management API** (a "Vetted Product") — never in the consumer Open
Permissions. Community Management approval requires an approved use case, **verified
organization**, verified domain, **an app verified by a LinkedIn Page associated with
the same organization**, and (Standard tier) a privacy policy + screencast review.
Dev-tier rate limits: 500 calls/app/24h, 100/member/24h. [learn.microsoft.com/.../increasing-access; .../community-management-app-review]
- The 2025 "LinkedIn opened Member Post Analytics to individuals" headline is true **only
through approved partner platforms** (Metricool/Buffer/Hootsuite-class) that hold the
approval — not by a creator calling the API directly. LinkedIn is also *tightening*:
the read scope `r_member_social` (Member Post Management) is flatly **closed** — "not
accepting access requests at this time due to resource constraints."
**Contradictions:** "CSV is the only way" (plugin/audit) is wrong at the *capability*
level (an API exists) but right at the *practical* level for a solo dev (the API is
org-vetted). **Resolution:** state "exists but partner-gated; not self-serve; CSV is the
practical floor for a solo creator," not "no API exists."
### D2. Per-post saves visibility — Confidence: high
**External findings:**
- **Native UI:** the official LinkedIn Help page on post analytics lists, under Social
Engagement: Reactions, Comments, Reposts, **Saves** ("number of times members saved
your post"), Sends — rolled out ~**Sept 2025**. **Count only, never saver identity.**
Rollout was phased and historical backfill limited (older posts may show no saves).
[linkedin.com/help/linkedin/answer/a516971; socialmediatoday.com/news/linkedin-adds-save-and-send-data-to-content-insights/759828/]
- **API:** `POST_SAVE` exposed via `memberCreatorPostAnalytics` from li-lms-2026-04 — but
behind the same Community-Management gate as D1 (not self-serve).
**Resolution (sharpens the Q3 honesty-fix):** the honest downgrade is **NOT** "saves
can't be tracked." It is: *"saves are visible in your native LinkedIn post analytics
(since Sept 2025, count-only) but there is no self-serve API to pull them, so this tool
does not auto-ingest them — read them in LinkedIn directly."* The operator's decision (no
manual-entry feature) stands and is defensible: the number is human-readable but
programmatically out of reach + would be hand-typed and instantly stale.
### D3. Auto-publish from a personal profile — Confidence: high (capability) / the ToS line is the real boundary
**External findings:**
- **A personal profile CAN auto-publish, self-serve.** `w_member_social` is an **Open
Permission** (no approval), granted by adding the free **"Share on LinkedIn"** product
(filed under `.../integrations/self-serve/`). `POST /v2/ugcPosts` (or `/rest/posts`)
with `lifecycleState: PUBLISHED` publishes **immediately, no human-in-the-loop**. Rate
limit ~150 requests/member/day. Self-serve content types: text, image, video,
article/URL share. [learn.microsoft.com/.../share-on-linkedin; .../getting-access]
- **Genuine limitations:** organic **carousels (ad-style) are NOT available** via API for
a personal profile (sponsored only) — use **MultiImage** for swipeable images; **member
document/PDF** posts and the modern `/rest/posts` *member* path are doc-ambiguous (don't
promise without testing). You also **cannot self-serve read your own posts back**
(`r_member_social` is restricted/closed) — capture the post URN from the publish
response header instead.
- **Operational friction (real, first-hand):** 3-legged OAuth (one interactive auth);
60-day access tokens / ~365-day refresh tokens that invalidate on revoke **or scope
change** (`invalid_grant` footgun); a common `403 me.GET.NO_VERSION` trap (use
`/userinfo` `sub`, not `/v2/me`); a placeholder company-page link required at app setup
even for personal-only posting; "outdated docs everywhere." [marcusnoble.co.uk/2025-02-02-posting-to-linkedin-via-the-api/; github.com/linkedin-developers/linkedin-api-js-client/issues/35]
- **The real boundary is ToS, not capability:** `w_member_social` being "open" does not
mean automated/scheduled posting is within LinkedIn's API Terms. The legitimate basis
for the plugin's clipboard-stop is **(a) per-user OAuth + token-refresh operational
overhead and (b) LinkedIn's Terms posture on automated posting** — not "the API can't
do it."
**Resolution:** the plugin must **not** write "cannot auto-publish." Honest framing:
*"auto-publish to a personal profile is technically possible and self-serve (Share on
LinkedIn / `w_member_social`); the plugin deliberately stops at the clipboard — OAuth +
60-day-token overhead and LinkedIn's Terms on automated posting make human-in-the-loop
the safer default. A choice, not a wall."* (Verify current Platform/Marketing API Terms
language on automated personal posting before finalizing the wording.)
### D4. Dwell time exportability — Confidence: medium-high (organic boundary holds, with carve-outs)
**External findings:**
- For **organic personal posts**, there is **no creator-facing dwell metric** — not in the
UI Help metric list, not in `memberCreatorPostAnalytics`. Dwell is an internal ranking
signal only. [linkedin.com/help/linkedin/answer/a516971]
- **Carve-outs (so the claim isn't falsifiable):** (1) **video** posts expose **Watch
time / Average watch time** (UI + member video API) — a real per-post depth metric, just
not for non-video; (2) `averageDwellTime` exists in the **ads** `adAnalytics` API (paid
campaigns only, methodology improved ~+25% in a 2026 change).
**Resolution:** "for organic posts, dwell is internal-only — no creator number; video
Watch time is the closest creator-visible depth proxy; a true dwell field exists but only
for paid ads." Do not say "LinkedIn has no dwell metric anywhere."
## External Knowledge
### Best Practice (official / primary)
Microsoft Learn (li-lms-2026-05) is authoritative. The self-serve set for a solo dev is
exactly three Open Permissions: `profile`, `email`, `w_member_social`. **Everything
analytics** is Community-Management-vetted. This single fact drives every honest boundary
statement.
### Known issues
The surface moves fast and docs lag the API (saves went invisible→UI→API in ~12 months;
`r_member_social` went from program to closed). Any boundary statement must be **dated**.
Practitioner reality adds friction (token fragility, the `/me` 403 trap) that makes
"feasible" ≠ "frictionless."
## Synthesis
1. **The audit's "honest scheduling boundary" finding (§5) was itself half-wrong.** It
assumed the plugin "stops at the clipboard because it can't auto-publish." The truth:
it *can* (self-serve), and the honest disclosure is about the **design + ToS choice**,
not an impossibility. Phase 1's boundary prose must encode the *choice*, or it ships a
brand-new false claim — exactly the failure mode this whole remediation exists to kill.
2. **The saves honesty-fix is a wording fix, not a "we can't" fix.** Saves are now
visible in the UI. The accurate downgrade points the user to LinkedIn's own analytics
for the number while being honest that the tool can't pull it. This keeps the operator's
"no manual-entry" decision *and* avoids a stale "saves aren't trackable" claim.
3. **There is a latent, deliberately-deferred capability here.** Auto-publish is buildable
self-serve. It is explicitly a **Non-Goal** of this remediation (operator: disclose
boundaries, don't engineer them away). Worth a one-line "possible but intentionally not
built (ToS + token overhead)" note so a future maintainer doesn't rediscover it as a
"missing feature."
## Open Questions
- **Exact ToS language on automated/scheduled personal posting** — the load-bearing reason
for the clipboard-stop. *Carry as: verify current Platform/Marketing API Terms before
finalizing the boundary wording.* (Not a blocker for the plan; a finalize-time check.)
- **Member document/PDF + `/rest/posts` member-author path** — doc-ambiguous. Irrelevant
unless a future version builds publishing; flag, don't resolve.
## Recommendation
For Phase 1 (honest boundaries) and Phase 2 (saves honesty-fix), write these **dated**
statements:
1. **Analytics API:** "A post-level analytics API for personal posts exists, but it's
behind LinkedIn's vetted Community Management API (verified organization + LinkedIn
Page + use-case review) — not self-serve for an individual. For a solo creator, CSV
export is the practical path. (As of 2026-05.)"
2. **Saves:** "Saves are visible in your native LinkedIn post analytics (since ~Sept 2025,
count-only). There's no self-serve API to pull them, so this tool doesn't track saves
automatically — read them in LinkedIn directly." Remove any "saves (10x weight) are the
highest-impact signal" claim presented *inside a report the tool populates*; reframe as
strategy guidance pointing to the native number.
3. **Auto-publish:** "Auto-publishing to a personal profile is technically possible
(self-serve `w_member_social`); this tool deliberately stops at the clipboard — OAuth +
token-refresh overhead and LinkedIn's Terms on automated posting. A choice, not a
limitation. (As of 2026-05.)" Reconcile the calendar/queue/"publish action" wording so
it never implies the tool auto-posts.
4. **Dwell:** "Dwell time is an internal ranking signal — no creator-facing number for
organic posts (video Watch time is the closest proxy)."
## Sources
| # | Source | Type | Quality | Used in |
|---|--------|------|---------|---------|
| 1 | [Member Post Statistics API (li-lms-2026-05)](https://learn.microsoft.com/en-us/linkedin/marketing/community-management/members/post-statistics?view=li-lms-2026-05) | official | high | D1, D2 |
| 2 | [Increasing Access — permissions table](https://learn.microsoft.com/en-us/linkedin/marketing/increasing-access?view=li-lms-2026-05) | official | high | D1 |
| 3 | [Community Management App Review](https://learn.microsoft.com/en-us/linkedin/marketing/community-management-app-review?view=li-lms-2026-05) | official | high | D1 |
| 4 | [Community Management Overview / FAQ (r_member_social closed)](https://learn.microsoft.com/en-us/linkedin/marketing/community-management/community-management-overview?view=li-lms-2026-05) | official | high | D1 |
| 5 | [Share on LinkedIn (self-serve)](https://learn.microsoft.com/en-us/linkedin/consumer/integrations/self-serve/share-on-linkedin) | official | high | D3 |
| 6 | [Getting Access — Open Permissions](https://learn.microsoft.com/en-us/linkedin/shared/authentication/getting-access) | official | high | D3 |
| 7 | [Post analytics for your content — LinkedIn Help (Saves listed)](https://www.linkedin.com/help/linkedin/answer/a516971) | official | high | D2, D4 |
| 8 | [Social Media Today — LinkedIn adds Save/Send data (Sept 2025)](https://www.socialmediatoday.com/news/linkedin-adds-save-and-send-data-to-content-insights/759828/) | community | medium-high | D2 |
| 9 | [Refresh Tokens with OAuth 2.0](https://learn.microsoft.com/en-us/linkedin/shared/authentication/programmatic-refresh-tokens) | official | high | D3 |
| 10 | [Marcus Noble — Posting to LinkedIn via the API (first-hand)](https://marcusnoble.co.uk/2025-02-02-posting-to-linkedin-via-the-api/) | community | medium-high | D3 |
| 11 | [GitHub — linkedin-api-js-client #35 (403 /me trap)](https://github.com/linkedin-developers/linkedin-api-js-client/issues/35) | community | medium | D3 |
| 12 | [Recent Marketing API Changes (ads averageDwellTime)](https://learn.microsoft.com/en-us/linkedin/marketing/integrations/recent-changes?view=li-lms-2026-01) | official | high | D4 |
| 13 | [ConnectSafely — LinkedIn API guide 2026 (approval reality)](https://connectsafely.ai/articles/linkedin-api-complete-guide-2026) | community | low-medium | D1 |

225
plugins/linkedin-studio/docs/remediation/research/03-coverage-gap-specs.md Normal file
View file

@ -0,0 +1,225 @@
---
type: trekresearch-brief
created: 2026-05-29
question: "For LinkedIn in 2026: hard short-form video requirements the algorithm rewards (aspect ratio, resolution, hook timing, captions); what triggers the templated-AI/engagement-bait down-rank; and newsletter-distribution mechanics (notification leverage, cadence, realistic cold-start numbers)?"
confidence: 0.80
dimensions: 5
mcp_servers_used: [tavily]
local_agents_used: []
external_agents_used: [docs-researcher, community-researcher, contrarian-researcher]
---
# Coverage-Gap Feature Specs — Video · De-AI · Newsletter Distribution (2026)
> Generated by trekresearch (standard external swarm: docs + community + contrarian)
> on 2026-05-29. Topic 3 of 3 for the linkedin-studio remediation. Feeds the Phase-2
> video gate, short-form de-AI gate, and newsletter-distribution surface.
> **The de-AI signal is covered in depth in Topic 1 (D8); this brief carries the
> video + newsletter specs and a short de-AI cross-reference.**
## Research Question
For LinkedIn in 2026: hard short-form video requirements the algorithm rewards (aspect
ratio, resolution, hook timing, captions); what triggers the templated-AI / engagement-bait
down-rank; and newsletter-distribution mechanics (notification leverage, cadence, realistic
cold-start numbers)?
## Executive Summary
**Three of the audit's proposed coverage-gap features rest on wrong premises, and building
them naively would bake new false claims into the plugin.** (1) A **hard 9:16 video gate is
wrong** — LinkedIn's audience is desktop-heavy; 9:16 is mobile-only and **crops on desktop**;
the broad-distribution picks are **4:5 / 1:1**, with 9:16 a mobile-only opt-in. The plugin's
internal contradiction ("4:5 preferred" vs "deprioritized") resolves toward **"4:5 preferred."**
The "3-second hook" is **imported TikTok folklore**, not a LinkedIn rule. The one video spec
**safe to enforce is captions** (80–85% watch muted; +12% watch-time per LinkedIn; caption
text is indexed) — but framed as best-practice, not "LinkedIn requires SRT." (2) **Native
video reach is DOWN ~36% YoY** (per-video views, Socialinsider 1.3M); the "video is king
2026" narrative is hype from a benchmark misread — so the video gate must be a **quality
gate for users who choose video**, never copy that positions video as top-reach. (3) The
newsletter **"triple-notification" framing is oversold** — LinkedIn's own FAQ says the three
channels are **deduplicated** (one notification per event); the real benefit is
"**bypasses organic feed ranking**," not three guaranteed touchpoints; and the **one-time
launch invite** makes a sub-~1–2K-follower newsletter premature. **Key caveat:** aspect-ratio
performance has **no rigorous comparative study** — encode as heuristic, never a hard gate;
date every claim ("as of 2026-05"). (Confidence 0.80 — high on captions / video-decline /
newsletter mechanics; medium on aspect-ratio + cold-start ranges.)
## Dimensions
### D1. Video upload specs (official, enforceable) — Confidence: high
**External findings (LinkedIn Help, primary):**
- Aspect ratio accepted **1:2.4–2.4:1**; resolution **256×144–4096×2304**; duration **min
3s desktop / 2s mobile, max 15min desktop / 10min mobile**; file **75KB–5GB**; **10–60fps**;
**192Kbps–30Mbps**; **MP4 the safe default**. [linkedin.com/help/linkedin/answer/a548372; .../a1311816]
- **Official-source conflict:** the member troubleshooting page lists MOV/AVI as *supported*;
the Pages spec says LinkedIn "no longer supports AVI, QuickTime, or MOV." **Resolution for
a gate:** enforce MP4 as the safe default; **warn-only** (not block) on MOV/AVI.
### D2. Aspect ratio — 9:16 is NOT a clean win — Confidence: medium (no rigorous data)
**External findings:**
- LinkedIn runs an official **full-screen vertical video experience** (video tab + carousel)
that **vertical (≈9:16) fills uncropped**; landscape is letterboxed. BUT you **cannot
directly post into** that surface — placement is algorithmic. [linkedin.com/help/linkedin/answer/a6290168]
- The contrarian + spec-guide consensus: for the **main feed** on a **desktop-heavy
professional audience**, **4:5 and 1:1** are the broad-distribution picks; **9:16 is
delivered mobile-only and crops to 1:1 on desktop** — a real degradation. The only official
numeric 9:16 target (720×1280 rec / 1080×1920 max) is scoped to **ads**, not organic.
[aspectratiocalculator.com/linkedin-aspect-ratios/; LinkedIn recommendation post]
- **No peer-reviewed or official study compares 4:5 vs 9:16 vs 1:1 engagement.** The
"vertical takes more feed real estate → more engagement" claim is uncorroborated heuristic.
**Resolution:** the plugin's contradiction resolves toward **"4:5 / 1:1 preferred for broad
distribution; 9:16 mobile-only opt-in / for the video tab."** Fix the file that calls 4:5
"deprioritized." **Do not build a hard 9:16 gate** — make aspect ratio guidance, not enforcement.
### D3. Hook timing + captions — Confidence: high (captions) / low (3-sec hook)
**External findings:**
- The **"3-second hook" is cross-platform folklore** (TikTok/Reels), absent from
LinkedIn-specific algorithm analyses; LinkedIn's only official "3 seconds" is the *minimum
video length*. The real LinkedIn-native reason the opening matters is **muted autoplay**.
[dataslayer.ai/blog/linkedin-algorithm-february-2026-whats-working-now; sproutsocial.com/insights/linkedin-video/]
- **Captions are the one spec safe to enforce:** ~80–85% watch muted; LinkedIn's own data
~**+12% watch-time** with captions; **caption text is indexed for search/discovery** and
factored into distribution. Both **SRT upload and native auto-captions** are first-party
(auto-captions in 10 languages, opt-in, reviewable). [opus.pro/blog/linkedin-video-caption-subtitle-best-practices; linkedin.com/help/linkedin/answer/a1327025; .../a552177]
**Resolution:** enforce a **captions** quality gate (BLOCK is defensible at 80% muted),
accept SRT OR auto-captions, label it "best-practice / algorithmic signal" (not "required").
Replace any "3-second hook" rule with **"front-load value for muted autoplay."**
### D4. De-AI / engagement-bait down-rank — Confidence: high (cross-ref Topic 1 D8)
**External findings (see Topic 1 D8 for full sourcing):**
- **Officially confirmed:** LinkedIn VP Laura Lorenzetti (2026-05-19) — active program
suppressing generic AI posts/comments + automation + attention-bait video; mechanism is
**reach-suppression (down to first-degree), not deletion**, via ML trained on human-
annotated "original thinking vs lacking substance." Engagement-pod crackdown officially
confirmed too (Sachdeva, 2026-02-16). [entrepreneur.com/business-news/linkedin-is-fighting-back-against-ai-slop-and-ai-comments]
- **Engagement bait** ("Comment YES", "Like for Part 2") → post-level throttle; genuine
open questions are **not** penalized. The line is *real answer* vs *reflexive token*.
**Resolution:** **build the short-form de-AI gate** targeting the signals LinkedIn *named*
(personal substance, original thinking, concrete specifics, genuine voice) — not an
unverified SEO "tell-list." Add a soft engagement-bait check (block mechanical-response CTAs,
allow genuine questions).
### D5. Newsletter distribution mechanics — Confidence: high (mechanics) / medium (cold-start)
**External findings:**
- **Official, solid:** all members can create newsletters (**max 5, 2-week cooldown**); on
first edition LinkedIn **auto-invites all connections/followers to subscribe** (and on each
new follow thereafter); editions are **also posted to the feed** + resurface via engagement
+ appear in interest/trending sections; **lowest-friction subscribe** (no typing — LinkedIn
has the email). [linkedin.com/help/linkedin/answer/a517925; .../a522525; .../a517914]
- **"Triple-notification" is OVERSOLD / contested:** LinkedIn's FAQ states the in-app / push /
email channels are **deduplicated** — "if you receive an in-app or push notification, you
should NOT expect to also receive an email for the same." So it is **one notification per
event via the subscriber's preferred channel**, not three guaranteed touchpoints. Delivery
failures are reported; one creator measured **~2–3% click vs 8–10% on their own email list.**
The defensible benefit is **"bypasses organic feed ranking,"** not "triple notification."
- **One-time launch invite → follower floor:** the mass "invite all followers" fires **once,
at any size** — launching sub-~1–2K followers permanently spends the blast on a tiny base.
Practitioner floor: **500+ min, 3,000+ ideal.** Defensible plugin gate: ~**1–2K floor**
(aligns with the existing ~1K `/monetize`/`/outreach` unlock).
- **Realistic cold-start (don't inflate):** true zero-audience start ≈ **0–100 subs months
1–3**; viral "0→9K/7-days" and "0→10K" case studies **were NOT cold starts** (leveraged
existing audiences / 12-month grinds). Cadence: **weekly common among top performers;
biweekly a safe default for original analysis.**
- **Honest downsides to disclose:** subscribers are **non-exportable** (platform lock-in — lose
them all if LinkedIn kills the feature); LinkedIn **outranks your own site** for the same
article (no canonical) — harmful if building an owned property; **no read/open/unsubscribe
analytics**; per-subscriber reach **decays** as the list grows.
**Resolution:** the newsletter-distribution surface must teach the **honest** version:
notification-bypass-of-feed (with the dedup caveat), the one-time-launch-blast + follower
floor, realistic cold-start floors, and the lock-in/no-canonical/no-analytics downsides — NOT
the audit's "triple-notification leverage" framing.
## External Knowledge
### Best Practice (official)
Enforceable/teachable from LinkedIn's own docs: video upload limits; captions (SRT + auto);
the full-screen vertical experience exists but isn't directly postable; newsletter mechanics
(5-max/2-week cooldown, auto-invite, feed resurfacing, dedup notifications). Everything about
*reach magnitude* (video, aspect ratio, newsletter click rates) is practitioner-sourced.
### Alternatives / contrarian
The "video is king 2026" and "triple-notification leverage" narratives are **both refuted**
— the first by a benchmark misread (views −36% YoY per video; +36% is the platform aggregate),
the second by LinkedIn's own dedup FAQ. Documents/carousels lead video on engagement (~7% vs
~6%). Build features on the corrected premises.
### Known issues
Aspect-ratio guidance has no rigorous data — heuristic only. Newsletter data is non-portable.
The MOV/AVI support conflict is unresolved in LinkedIn's own docs. Date every claim.
## Synthesis
1. **"Close the coverage gaps" ≠ "build what the audit sketched."** The audit named the
gaps correctly (no video enforcement, no de-AI gate, thin newsletter distribution) but
sketched two of the fixes on hype: a 9:16 gate and "triple-notification leverage." The
research says build a **captions/aspect-guidance** video gate (not 9:16-mandatory) and an
**honest** newsletter surface (bypass-feed + caveats), not the sketched versions. This is
the same disease the whole remediation treats — features must rest on verified premises.
2. **The de-AI gate is the highest-confidence Phase-2 build** (D4 + Topic 1 D8): officially
confirmed, named-executive, with a stated mechanism. It is also the **single most robustly
triangulated 2026 down-rank signal** the audit flagged as unguarded on short-form. Prioritize it.
3. **The newsletter follower-floor reconciles two findings cleanly:** the one-time launch
blast + the existing ~1K `/monetize`/`/outreach` unlock → gate the newsletter behind a
~1–2K floor framed as "wait until you can spend the launch blast well." Below it, the
plugin should steer the user to short-form/document content to *build* the base.
## Open Questions
- **Newsletter email-delivery behavior** — genuinely contested (FAQ dedup vs third-party
"always emailed"). *Carry as: state the dedup behavior from the FAQ; note delivery is not
guaranteed; a one-edition live test would resolve it.* Not a blocker.
- **Aspect-ratio engagement** — no rigorous data exists. *Carry as heuristic; never a hard gate.*
- **Dedicated vertical video feed as a primary surface** — if LinkedIn promotes it, the 9:16
calculus could flip. *Re-check before finalizing the video gate copy.*
## Recommendation
For Phase 2:
1. **Video gate = quality gate, not reach-push.** Enforce MP4 + within-limits (warn on
MOV/AVI); **enforce/strongly-recommend captions** (SRT or auto); make aspect ratio
**guidance — 4:5 / 1:1 preferred, 9:16 mobile-only opt-in**; fix the "4:5 deprioritized"
contradiction toward "4:5 preferred"; drop any "3-second hook" rule and any "video
maximizes reach" copy. Add a one-line note that per-video reach is declining and
documents out-engage video.
2. **Build the short-form de-AI gate** (highest-confidence build) on LinkedIn's named signals
(substance / original thinking / specifics / voice) + a soft engagement-bait check.
3. **Newsletter-distribution surface = the honest version:** "bypasses organic feed ranking
(one deduplicated notification per subscriber per edition)"; one-time launch-blast +
**~1–2K follower floor**; realistic cold-start floors (0–100 subs months 1–3); disclose
non-export/no-canonical/no-analytics/per-subscriber-decay. Steer sub-floor users to build
the base first.
## Sources
| # | Source | Type | Quality | Used in |
|---|--------|------|---------|---------|
| 1 | [Video sharing troubleshooting (specs)](https://www.linkedin.com/help/linkedin/answer/a548372) | official | high | D1 |
| 2 | [Video specs for Pages (MOV/AVI conflict)](https://www.linkedin.com/help/linkedin/answer/a1311816) | official | high | D1 |
| 3 | [Full-screen vertical video](https://www.linkedin.com/help/linkedin/answer/a6290168) | official | high | D2 |
| 4 | [Aspect Ratio Calculator — LinkedIn 2026](https://www.aspectratiocalculator.com/linkedin-aspect-ratios/) | community | medium | D2 |
| 5 | [Add Closed Captions (SRT)](https://www.linkedin.com/help/linkedin/answer/a552177/add-closed-captions-to-videos-on-linkedin) | official | high | D3 |
| 6 | [Auto captions for videos](https://www.linkedin.com/help/linkedin/answer/a1327025) | official | high | D3 |
| 7 | [OpusClip — caption best practices (80% muted, +12%)](https://www.opus.pro/blog/linkedin-video-caption-subtitle-best-practices) | community | medium | D3 |
| 8 | [Socialinsider 2026 benchmarks (video −36% YoY)](https://www.socialinsider.io/social-media-benchmarks/linkedin) | community | medium-high | D2, D5 |
| 9 | [Omni Lab — video views down 36% YoY](https://www.omnilabconsulting.com/blog/linkedin-video-views-down-36-yoy) | community | medium | D2 |
| 10 | [Entrepreneur — LinkedIn fights AI slop (Lorenzetti)](https://www.entrepreneur.com/business-news/linkedin-is-fighting-back-against-ai-slop-and-ai-comments) | official (reported) | high | D4 |
| 11 | [Manage a newsletter (5-max, cooldown, auto-invite)](https://www.linkedin.com/help/linkedin/answer/a517925) | official | high | D5 |
| 12 | [Newsletters overview (triple notification wording)](https://www.linkedin.com/help/linkedin/answer/a522525) | official | high | D5 |
| 13 | [Newsletters FAQ (dedup; resurfacing)](https://www.linkedin.com/help/linkedin/answer/a517914) | official | high | D5 |
| 14 | [The Lime One — follower floor for newsletter](https://thelime.one/blog/how-many-followers-do-you-need-for-linkedin-newsletter) | community | medium | D5 |
| 15 | [The Science Marketer — newsletter pros/cons (lock-in, click rates)](https://thesciencemarketer.com/p/linkedin-newsletter-pros-cons) | community | medium | D5 |
| 16 | [InfluenceFlow — newsletter cold-start ranges 2026](https://influenceflow.io/resources/linkedin-newsletter-strategy-complete-guide-to-building-an-engaged-subscriber-base-in-2026/) | community | low-medium | D5 |
| 17 | [dataslayer — LinkedIn algorithm Feb 2026](https://www.dataslayer.ai/blog/linkedin-algorithm-february-2026-whats-working-now) | community | low-medium | D3 |