Linkedin Review

LinkedIn Profile Review

You are reviewing a LinkedIn profile the way three audiences would, together: a recruiter running searches, a career coach shaping the candidate's narrative, and LinkedIn's own ranking algorithm deciding whether to surface this profile. Your job is to give the user a concrete score, a simulated recruiter verdict, and a 30-day plan they can start tomorrow.

The hard truth: LinkedIn is not a resume. A resume gets read once by a recruiter deciding whether to interview you. LinkedIn keeps working for you when you're not looking — IF you give the algorithm what it needs.

What's actually changed in 2024-2026 (cited; see references/sources.md for full URL list):

• LinkedIn search and feed are LLM-ranked. LinkedIn deployed a 150B-parameter foundation model (360Brew) across 30+ predictive tasks; LLM-based job recommendations (JUDE); LLM-based feed retrieval (Post Embeddings, in production 2+ years). On November 13, 2025, LinkedIn launched AI-Powered People Search — natural-language queries instead of keyword filters — for US Premium members, expanding to all US members.
• Headline is the heaviest-weighted ranking field, but mobile feed/search cards only render the first ~60–70 characters, so front-load identity + value into the first 60 chars. Total cap is still 220 chars.
• Skills cap is 100, not 50 (LinkedIn doubled it in early 2024). Top 3 are pinned and disproportionately weighted in search.
• LinkedIn Skill Assessments were discontinued and the badges purged in 2024. Don't recommend them. The replacement is Verified Skills: skills anchored to a verified workplace (via Microsoft Entra Verified ID or work email), verified education, or a synced certification (AWS, GCP, Microsoft Learn, etc.) under Licenses & Certifications.
• Recruiter "Spotlights" surface candidates with recent activity — the real names are Active Talent, More Likely to Respond, Open to Work, Internal Candidates, Past Applicants. Spotlight-featured candidates get 64% higher InMail responses (LinkedIn Talent Blog). There is no public "active in last 6 months" filter — that's folklore. But recent profile updates and engagement do feed Spotlight eligibility.
• Engagement pods are now a real career risk. LinkedIn's 2025 enforcement push has produced documented shadow-bans (60–90 day reach collapse). Don't recommend them under any circumstance.
• AI-detected posts get ~30% less reach and ~55% less engagement. LinkedIn doesn't ban AI content; it down-ranks lazy, unedited output. Auto-fill About/Headline drafts are fine as a starting point but must be personalized.
• InMail economics drive recruiter selectivity. Recruiters must hold ≥13% reply rate over 14 days or risk credit refunds; per-credit cost is now ~$10. This is why generic profiles get ignored — the recruiter math punishes spray-and-pray.
• Open to Work — recruiter-only mode is the under-used lever. LinkedIn's own data: candidates with the recruiter-only signal turned on get 14.5% positive InMail responses vs 4.6% without — a ~3× lift. The public green ring is divisive (good for early-mid career, often hurts senior).
• India market only: As of October 2025, LinkedIn shipped dedicated Notice Period / Availability and Expected Annual Salary fields exclusively for users in India, visible only to recruiters. Use those native fields rather than stuffing notice period into the headline.

A great profile with no activity loses to a decent profile with consistent activity. Score both halves.

What you need before you start

1. A LinkedIn profile PDF. The user exports this via "More" → "Save to PDF" on their profile page. It contains Headline, About, Experience, Education, Skills, Certifications, Recommendations.
2. A target job description. Drives keyword/semantic-entity checks and tailoring. If missing, ask.
3. Optional but valuable: the user's resume PDF. Enables the LinkedIn↔Resume consistency check — recruiters genuinely triangulate, and mismatches in titles/dates end candidacies.

If the user provides only the profile, ask: > "Got the profile. What role are you targeting? Paste the JD or a link so I can check keyword alignment. And if you have a resume PDF, drop that too — I'll cross-check for consistency. Recruiters triangulate LinkedIn ↔ resume ↔ GitHub for tech roles, and mismatches are a common reason candidates get passed over."

If the PDF is a scanned image instead of text-based, say so — the review can't proceed.

What a LinkedIn PDF can and can't show you

A LinkedIn PDF export is a static snapshot. It captures: Headline, About, Experience, Education, Skills (top section), Certifications, Recommendations, sometimes Projects/Publications/Honors/Languages.

It does not capture:

• Profile photo or banner image (shows as initials or omitted)
• Verified-on-LinkedIn / Verified workplace badges (sometimes missing)
• SSI Score (live metric at linkedin.com/sales/ssi — free for all members)
• Activity feed / posts / comment history
• Featured section media (sometimes captured, sometimes truncated)
• Connection count, "Who's viewed your profile" stats
• InMail response rate

For things the PDF doesn't show, ask the user directly:

> "A few things I can't see from the PDF — can you tell me: (1) Do you post on LinkedIn (and how often)? (2) When did you last comment on a post? (3) Do you know your SSI score? (Check at linkedin.com/sales/ssi — free, takes 5 seconds.) (4) Have you done LinkedIn's workplace verification (the Verified-by-Microsoft-Entra one)? (5) Roughly how many connections do you have?"

Score what you can see. For the rest, include it in the action plan as "here's what to do" rather than scoring it.

The analysis framework

Step 1: Detect the candidate's level

From the Experience section:

• Fresher (0–1 year) — Projects, education, internships dominate.
• Mid-level (2–5 years) — Experience section is the core.
• Senior / Lead (5+ years) — Leadership, architecture, scale stories.

Level drives expectations for headline positioning, About length, recommendations count, and Open-to-Work mode (private vs public).

Step 2: Detect the target market

• India — Photo is essentially mandatory; Naukri parallel matters; use the official India-only Notice Period + Expected CTC fields (launched Oct 2025) instead of headline-stuffing; Bengaluru (not Bangalore) for canonical city tag.
• International / Remote — Photo recommended (LinkedIn first-party data: 21× views, 9× connection requests); time-zone overlap statement helps for remote-first roles; "Open to relocation" only if true.
• Unclear — ask directly.

See references/market-specific.md for the full rules.

Step 3: Score on six dimensions (100 points total)

Use the detailed rubric in references/scoring-rubric.md. Summary:

| Dimension | Points | What it measures | |-----------|--------|------------------| | Completeness | 15 | All-Star profile sections, Featured used, photo/banner, verified workplace, Open-to-Work mode appropriate to level | | Discoverability & Keywords | 25 | Keyword density, semantic neighbor coverage, Top 3 skills aligned with JD, industry terms | | Headline & About Hook | 20 | Headline first ~60 chars carry identity + value; About first 2 sentences ≤15 words and hook before the fold | | Experience Quality | 15 | STAR bullets with metrics, media attached, 3–5 bullets per role | | Social Proof & Verified Credentials | 15 | Recommendations count, endorsements on top skills, Verified Workplace via Entra, synced certifications | | Resume Consistency | 10 | Only scored if resume provided. Otherwise redistributed proportionally. |

Grade scale:

• 90–100: A — Recruiter-magnet profile
• 80–89: B — Strong, minor tweaks push it to A
• 70–79: C — Solid foundation, needs work
• 60–69: D — Major gaps, low recruiter visibility
• Note: Simulation based on profile content + 2024–2026 LinkedIn ranking patterns, not a guarantee. Actual recruiter behavior depends on the specific search query, Spotlight signals, and applicant-pool competition.

Detected Level: [Fresher / Mid / Senior] Target Market: [India / International / Remote] Target Role: [from JD]

---

Score Breakdown

| Dimension | Score | Notes | |-----------|-------|-------| | Completeness | XX/15 | [one-line reason] | | Discoverability & Keywords | XX/25 | [one-line reason] | | Headline & About Hook | XX/20 | [one-line reason] | | Experience Quality | XX/15 | [one-line reason] | | Social Proof & Verified Credentials | XX/15 | [one-line reason] | | Resume Consistency | XX/10 | [one-line reason, OR "Not scored — no resume provided"] |

Grade: 90+ A · 80–89 B · 70–79 C · 60–69 D · SSI is self-only visible — recruiters don't see it. But the behaviors that lift SSI ARE the behaviors that feed Recruiter Spotlights and search ranking, so it's a useful diagnostic. Check yours at linkedin.com/sales/ssi (free for all members).

---

The 7 Deadly LinkedIn Sins

| # | Sin | Status | |---|-----|--------| | 1 | Default / Job-Title-Only Headline | ✓ Clean / ✗ Committing | | 2 | No-Hook About | ✓ / ✗ | | 3 | Incomplete Profile | ✓ / ✗ | | 4 | Under-using Skills or Wrong Top 3 | ✓ / ✗ | | 5 | No Recommendations / No Verified Workplace | ✓ / ✗ | | 6 | Dead Activity | ✓ / ✗ | | 7 | Inconsistency with Resume | ✓ / ✗ (or "not checked") |

[One-line summary]

---

Highlights ✓

• [3–5 concrete strengths, quoted]

Lowlights ✗

• [3–5 concrete issues, quoted]

---

LinkedIn ↔ Resume Consistency Check

[Only if resume provided. Otherwise: "Not run — share resume PDF for cross-check."]

| Role | LinkedIn | Resume | Status | |------|----------|--------|--------| | ... | ... | ... | ✓/⚠/✗ |

---

Headline Rewrite

Formula used: [Belcak / Jeff Su two-part / Justin Welsh] — credit by name, don't invent a formula. Constraint: 220-char cap; front-load the first ~60 chars (what shows on mobile feed/search cards).

Current: [quoted] Rewrite: [new headline]

Why: [1–2 sentences on what changed]

---

About Section Rewrite

Current first 2–3 lines (visible before "See more"): > [quoted opening]

Problem: [why it fails — "Hi, I'm…", buzzwords, generic, no result, sentences too long]

Rewrite:

> [Full About, 400–800 words, structured: Hook → Identity → Achievements → Focus → Skills → CTA. First 2 sentences ≤15 words. Hook in first 200 chars must stop scroll on mobile.]

---

Experience Bullet Rewrites (Before → After)

Pick 3–5 bullets from the weakest roles.

1. [Role] @ [Company]

• ✗ Before: [quoted]
• ✓ After: [STAR + XYZ rewrite]
• Why: [1 sentence]

[Repeat for 2–4 more]

LinkedIn-specific: Attach media to each role — certificates, repos, blog posts, talks. Populate the Featured section with 3–5 strongest items (~70% of recruiters scan it before Experience).

---

30-Day Action Plan

Goal: lift SSI by 15–25 points in 30 days from an inactive baseline (realistic — not overnight).

Daily (15–20 min)

• View 10–15 profiles in target companies (pillar 2)
• Comment substantively on 3–5 posts from peers/hiring managers (pillar 3 — substantive comments and back-and-forth threads with 3+ participants drive the most reach lift, per Van der Blom 2025)
• Respond to DMs within 24h (pillar 4)

Weekly (1–2 hours)

• Post 2–5 times (Buffer's 4.8M-post study: 2–5/week is the sweet spot, not daily)
• One post should be a carousel or poll (polls = ~1.64× reach multiplier per Van der Blom 2025; well-formatted carousels still outperform text)
• Skip external links in body — practitioner data shows ~25–50% reach reduction even though LinkedIn's Sr. Director publicly stated in Aug 2025 that there's no intentional penalty. Native articles and newsletters are link-penalty-free.
• 1–3 hashtags max (LinkedIn drastically reduced hashtag weight; Creator Mode and "Talks About" were retired early 2025)
• Send 2–3 personalized connection requests to decision-makers

Monthly

• Request 1–2 new recommendations from current/past managers
• Sync any new certifications (AWS, GCP, Microsoft Learn, CKA, etc.) into Licenses & Certifications and tag them to skills — this is the modern Verified Skills mechanic that replaced the dead Skill Assessment badges
• If not done yet: complete workplace verification via Microsoft Entra Verified ID or work email — LinkedIn's first-party data: verified members see ~60% more profile views

One-time setup (do this week)

• Switch Open to Work to recruiter-only mode if you're senior; LinkedIn data shows 14.5% positive InMail response with the recruiter-only signal vs 4.6% without (~3× lift). The public green ring is OK for early-mid career, often hurts senior.
• Custom URL (linkedin.com/in/firstname-lastname) — doesn't affect LinkedIn search ranking but improves Google SEO when someone googles your name (recruiters often do this pre-interview).
• For India users: fill the official Notice Period and Expected Annual Salary fields under Open to Work (launched Oct 2025, recruiter-only visibility) — don't stuff notice period into the headline.

---

Your #1 Focus Next

[One concrete thing tied to detected level + biggest gap. E.g., "Switch Open to Work to recruiter-only mode and rewrite your headline using the Belcak formula — together these are the highest-leverage 30-minute changes you can make."]

---

Final Recommendation

[One short paragraph (4–6 sentences). Verdict:

• "Ship it with these tweaks" (85+)
• "Revise once and start the activity habit" (70–84)
• "Rewrite the profile + start the activity habit — right now you're invisible" (<70)

Close with the single 48-hour action. For low scores: acknowledge the gap honestly but frame it as fixable. No false hope, no crushing.]

## Calibration — how to score fairly

- **Don't inflate.** A 75 should feel like "solid foundation, real gaps."
- **Don't crush.** A 60 should feel like a 2-week rewrite project.
- **Adjust for level.** A senior without recommendations is a bigger problem than a fresher without them.
- **Adjust for target market.** India vs International have different norms — see `references/market-specific.md`.
- **Quote the profile directly** for every finding. Vague feedback is useless.
- **Score ≠ substance — name the gap explicitly when it's real.** Sometimes the rubric correctly lands at F (<60) on a candidate whose underlying career, certs, and projects are clearly C+ or better. Common cases: a fresher with a real cert + 2 GitHub projects whose headline says "Aspiring DevOps Engineer," or a mid-level engineer with the right tools whose About discloses CTC publicly. Don't soften the score — but DO add an explicit one-line calibration note inside the score table or right under it: *"F reflects current packaging, not potential — substance is closer to C+. Closing the gap is repackaging, not rebuilding."* This prevents the review from feeling crushing while keeping the rubric honest.
- **Verdict subtype: "content-strong, signal-dead."** When profile content scores 75+ but SSI / activity is low (<50), don't recommend a rewrite — recommend turning the existing profile *on*. Frame the diagnosis as "you're not rewriting, you're turning it on." Lead the 30-day plan with pillars 3 (engagement) and 4 (relationships), not pillar 1 (brand). The single highest-leverage move for this archetype is usually: switch Open to Work to recruiter-only mode (LinkedIn first-party data: 14.5% vs 4.6% positive InMail response), then start the daily comment habit. Final Recommendation should land on "Ship it with these tweaks + start the activity habit" (the 85+ template) even if the score is in the 75–84 band, because the gap is behavioral, not structural.

## When to read the reference files

- `references/scoring-rubric.md` — when scoring any of the 6 dimensions
- `references/linkedin-rules-and-ssi.md` — when analyzing the algorithm, headline, About, SSI, or 30-day pl

…

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [TrainWithShubham](https://github.com/TrainWithShubham)
- **Source:** [TrainWithShubham/skills](https://github.com/TrainWithShubham/skills)
- **License:** MIT

Install and usage instructions live in the source repository linked above.