LinkedIn Ads audit checklist and best-practice catalog covering tracking (Insight Tag, Conversions API, pipeline-stage events), audience/ABM targeting, creative format mix, and structure/bid — with a weighted Health Score formula and archetype filtering for B2B SaaS, professional services, and lead-gen brands. Use when auditing a LinkedIn ad account or deciding whether a brand archetype should be running LinkedIn paid at all.
Scanned 9/2/2026
Install to Claude Code
Skills Directorynpx -y skills add Meshpilot-AGI/meshpilot-digital-marketing-agent --skill linkedin-ads --agent claude-codeInstalls into .claude/skills of the current project.
Are you the author of Linkedin Ads?
Add the live security badge to your README — it updates automatically with every re-scan.
[](https://www.skillsdirectory.com/skills/meshpilot-agi-linkedin-ads)More formats (shields.io, HTML) on the badges page.
---
name: linkedin-ads
description: LinkedIn Ads audit checklist and best-practice catalog covering tracking (Insight Tag, Conversions API, pipeline-stage events), audience/ABM targeting, creative format mix, and structure/bid — with a weighted Health Score formula and archetype filtering for B2B SaaS, professional services, and lead-gen brands. Use when auditing a LinkedIn ad account or deciding whether a brand archetype should be running LinkedIn paid at all.
---
# LinkedIn Ads Audit Checklist
Targets the brands that actually run LinkedIn paid: B2B SaaS, high-ACV professional services,
lead-gen, and marketplaces with B2B supply-side acquisition. D2C / creator / mobile-app brands
typically should NOT run LinkedIn; this catalog won't apply to them after archetype filtering.
Sources distilled in the field: LinkedIn's own Conversions API playbook, Refine Labs (Chris
Walker), the Pavilion exec community, Influ2 / DemandBase ABM frameworks, Goldcast / Hatch
ABM-content loops, WebMechanix's B2B Ads playbook, and pipeline-stage attribution practice.
Health Score weight mix favors **Tracking 30% · Audience 30% · Creative 20% · Structure 20%**
because LinkedIn's leverage is overwhelmingly audience quality + correct conversion signal — not
creative variation or campaign-count tuning the way Meta is.
## Schema
Each check carries:
- `id` — `L01`+ stable identifier, never reused.
- `applies_to_archetypes` — almost all are `[b2b_saas]` or `[b2b_saas, local_service]` (where
local_service = high-ACV professional services like agencies, consultancies, law firms); some
are `[b2b_saas, marketplace]` for supply-side B2B acquisition.
- `primary_metric_kind` — defaults to `cost_per_sql` for B2B SaaS, `cost_per_qualified_lead` for
service / lead-gen.
Critical checks (severity 5.0×) cap their category sub-score at 50 on FAIL.
---
## Category A — Tracking & Conversions (30% weight)
### L01 · Insight Tag installed + firing
- weight: 10 · severity: **critical** · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial, marketplace]
- pass: LinkedIn Insight Tag fires on every page (site-wide PageView) AND on the conversion
event(s); confirmed via Campaign Manager → Insight Tag → "Active" status with last fire <24h
- warning: tag fires but missing on conversion page (matched-audience retargeting impossible)
- fail: tag inactive OR not installed
- rationale: every other LinkedIn check needs the Insight Tag; no matched audiences, no
conversions, no funnel analysis without it
- fix: install via GTM (LinkedIn Insight Tag template) or hardcode into `<head>`; verify in
Campaign Manager
### L02 · Conversions API (LCAPI) active
- weight: 10 · severity: **critical** · fix_effort: medium
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: server-side LCAPI mirroring ≥0.7× of pixel-side conversions for the primary event;
LinkedIn-reported "Server EMQ" score ≥6
- warning: LCAPI live but partial (0.4–0.7× ratio)
- fail: no LCAPI — relying on browser pixel alone (loses 30–50% on iOS / blockers)
- rationale: LinkedIn's parallel to Meta CAPI; without it, the optimizer is making bid decisions
on a 50–70% signal sample
- fix: wire via Zapier / native HubSpot / Salesforce / Stripe LCAPI integration; map every
conversion event with `event_id` for dedup
### L03 · Conversion categories correctly mapped
- weight: 7 · severity: high · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: each conversion event uses a SPECIFIC LinkedIn category (`LEAD`, `SIGN_UP`, `PURCHASE`,
`BOOK_APPOINTMENT`, `ADD_TO_CART`, etc.) — not `OTHER`
- warning: ≥1 primary conversion uses `OTHER`
- fail: all primary conversions use `OTHER` — optimizer gets no semantic signal
- rationale: LinkedIn's automated optimization (Maximum Delivery, Target Cost) reads the
category enum; `OTHER` defaults to a catch-all model with weaker performance
- fix: rename each conversion event with the correct category in Campaign Manager → Account
assets → Conversions
### L04 · Pipeline-stage events uploading (MQL/SQL/Opp/Won)
- weight: 9 · severity: **critical** · fix_effort: high
- applies_to_archetypes: [b2b_saas]
- primary_metric_kind: cost_per_sql
- pass: CRM (HubSpot / Salesforce / Pipedrive) uploads `MQL`, `SQL`, `Opportunity`,
`Closed-Won` stage transitions to LinkedIn via LCAPI offline conversions with `value` =
opportunity ARR
- warning: only `Lead` uploads — the raw-form-fill trap
- fail: no pipeline-stage upload → optimizer buys low-fit MQLs that never convert
- rationale: the gap between cost_per_lead and cost_per_sql is the single biggest leak in B2B
paid; without pipeline-stage data, LinkedIn's algorithm scales the wrong thing
- fix: wire CRM → LinkedIn LCAPI with stage-tagged events; this is the highest-ROI work for any
B2B account
### L05 · 90-day attribution window enabled
- weight: 6 · severity: high · fix_effort: low
- applies_to_archetypes: [b2b_saas, marketplace]
- primary_metric_kind: cost_per_sql
- pass: account-level attribution = 90-day click; reports default to same window
- warning: 30-day click set but reports default to 7-day (under-reports)
- fail: 7-day default — sales-cycle conversions outside the window are invisible
- rationale: B2B sales cycles run 30–90 days; default 7-day attribution hides 60–80% of
paid-influenced revenue
- fix: Campaign Manager → Account → Conversions → Attribution settings
### L06 · UTM + LinkedIn-specific parameters set
- weight: 4 · severity: medium · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial, marketplace]
- pass: account-level URL tracking template populates `utm_source=linkedin`, `utm_medium=cpc`,
`utm_campaign={li_campaign_id}`, `utm_content={li_creative_id}`, `utm_term={li_ad_format}`
- warning: present but missing campaign-id or creative-id
- fail: no template — every campaign carries default LinkedIn click-IDs only, no
analytics-stack join key
- fix: paste the template in Campaign Manager → Settings → URL parameters
---
## Category B — Audience & Targeting (30% weight)
### L20 · Matched Audiences from CRM lists
- weight: 9 · severity: **critical** · fix_effort: medium
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: ≥3 active Matched Audiences uploaded from CRM (customers, closed-won opportunities,
lost opportunities), refreshed weekly, each ≥300 matched members (LinkedIn's audience-size
floor)
- warning: 1–2 lists OR last refresh >30d
- fail: 0 — running on demographic targeting only (no first-party signal)
- rationale: first-party matched audiences are LinkedIn's strongest targeting signal — exclude
existing customers from prospecting, retarget engaged prospects, look-alike off won-deal lists
- fix: CSV export from CRM → Campaign Manager → Audiences → Upload Contact List; wire daily sync
via Zapier / native HubSpot integration
### L21 · ABM / Company List targeting (high-ACV only)
- weight: 7 · severity: high · fix_effort: medium
- applies_to_archetypes: [b2b_saas]
- primary_metric_kind: cost_per_sql
- pass: if ACV ≥ $25k, an account-list-target campaign exists with the brand's target-account
list (1k–10k accounts) + buying-committee job-title filter; ≥30% of LinkedIn spend on it
- warning: list exists but no buying-committee filter (just "anyone at the account")
- fail: ACV ≥$25k but no ABM campaign — paying LinkedIn premium CPM for non-target accounts is
the most expensive way to spray
- rationale: above $25k ACV, the CAC math only works on ABM; below $25k, broader prospecting can
work
- fix: build a target-account list in Sales Navigator → upload to Campaign Manager as a Matched
Audience (company-match)
### L22 · Exclusion hygiene
- weight: 6 · severity: high · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: existing customers + closed-won + closed-lost-recent excluded from every prospecting
campaign; competitor employees excluded account-wide
- warning: customers excluded but no competitor exclusion
- fail: no exclusions — paying LinkedIn premium CPM for customers and competitors
- fix: build the 3 exclusion audiences once; apply to every prospecting campaign in the
campaign-group template
### L23 · Audience size in the 50k–400k sweet spot
- weight: 5 · severity: medium · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: every active prospecting audience sits between 50k and 400k estimated members
(LinkedIn's published sweet spot)
- warning: 1–2 audiences outside the band
- fail: any active audience <20k (under-deliver) or >2M (low signal)
- rationale: <50k = LinkedIn flags "narrow" and slow-delivers; >400k = the algorithm has
nothing to optimize toward
- fix: tighten with seniority + job-function + company-size filters; loosen with broader
job-function combinations
### L24 · Audience Network expansion deliberately set
- weight: 3 · severity: low · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial, marketplace]
- pass: every campaign has Audience Network expansion EXPLICITLY toggled (on for brand
awareness, off for lead-gen — never default)
- warning: default-on across the account (LinkedIn turns this on by default; the expanded
network has 3–5× cheaper impressions but conversion quality drops sharply for lead-gen)
- fail: any conversion-objective campaign with Audience Network ON
- fix: toggle off in every conversion campaign; keep on for awareness if the budget allows
### L25 · Lookalike (Audience Expansion) tested
- weight: 4 · severity: medium · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: at least one campaign running with LinkedIn's "Audience Expansion" (lookalike
equivalent) seeded from a won-deal Matched Audience
- warning: tested but paused
- fail: never tested — leaving the algorithm's best prospecting tool on the shelf
- fix: duplicate a top-performing prospecting campaign, enable Audience Expansion, compare CPL
over 30 days
---
## Category C — Creative & Format (20% weight)
### L40 · Format diversity (≥3 active formats)
- weight: 6 · severity: medium · fix_effort: high
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial, marketplace]
- pass: ≥3 active formats from {single-image, video, document, carousel, conversation ad,
thought-leader ad}
- warning: 2 formats
- fail: 1 format only — LinkedIn's algorithm optimizes per-format independently; single-format
accounts cap their own ceiling
- fix: prioritize document ads (highest CTR for SaaS, 2× single-image on average) and
thought-leader ads (lowest CPL when the founder has personal brand)
### L41 · Lead Gen Forms used over external LP
- weight: 7 · severity: high · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- primary_metric_kind: cost_per_lead
- pass: lead-objective campaigns use LinkedIn Lead Gen Forms (pre-filled with member profile
data); LGF-to-external-LP split documented per campaign
- warning: LGF on some campaigns, external LP on others without a test (no controlled
comparison)
- fail: 100% external LP for lead capture (typical CPL is 2–4× higher than LGF for B2B)
- rationale: LGF removes the form-filling friction; member data pre-populates from LinkedIn
profile; conversion lift 2–3×
- fix: convert lead-objective campaigns to LGF; wire LGF → CRM via native HubSpot / Salesforce /
Marketo integration
### L42 · Thought-leader ads tested
- weight: 5 · severity: medium · fix_effort: high
- applies_to_archetypes: [b2b_saas]
- pass: ≥1 thought-leader ad active in the last 30d, sponsored from a founder / executive / SME's
personal profile (not the company page)
- warning: tested but paused
- fail: never tested — high-ACV SaaS leaves the cheapest CPL surface on the table
- rationale: thought-leader ads consistently produce 30–50% lower CPL than company-page ads for
B2B; the personal-brand signal cuts through the LinkedIn-ad-blindness most users have
- fix: pick a founder / VP of Marketing post that resonated organically; sponsor it from their
profile via Campaign Manager → Sponsor a post
### L43 · Ad text within LinkedIn limits + scannable
- weight: 3 · severity: low · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial, marketplace]
- pass: intro text <150 chars (the mobile-feed cutoff); headline <70 chars; description (if
applicable) <100 chars; CTA matches the conversion intent
- warning: 150–600 / 70–150 / 100–200 — wall-of-text feed cards
- fail: >600 / >150 / >200 — feed-card truncation kills clarity
- fix: trim copy; lead with the value prop in the first 80 chars
### L44 · Creative refresh cadence (every 14d top-spend)
- weight: 4 · severity: medium · fix_effort: high
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: at least one new creative every 14d on each top-spend campaign
- warning: 14–28d gap
- fail: >28d gap — frequency caps + audience fatigue in narrow audiences are amplified
- fix: weekly creative ops cadence; LinkedIn's narrower audiences feel fatigue sooner than
Meta's
---
## Category D — Structure & Bid (20% weight)
### L60 · Conversion objective per campaign matches funnel stage
- weight: 6 · severity: medium · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: TOFU = Brand Awareness or Reach; MOFU = Website Visits or Engagement; BOFU = Website
Conversions or Lead Generation; each conversion campaign's objective matches the conversion
stage it optimizes for
- warning: mismatch on one or two campaigns (e.g. Brand Awareness campaign trying to drive demo
requests)
- fail: blanket "Website Conversions" for the entire funnel — pays premium CPM for top-funnel
impressions
- fix: re-objective per stage; LinkedIn's pricing changes 5–10× by objective
### L61 · Daily budget ≥ 5× target primary metric
- weight: 5 · severity: medium · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: every active conversion campaign has daily budget ≥ 5× the brand's primary-metric
target (LinkedIn floors at $10/day for most objectives; ABM with $200 CPL targets needs
$1000/day to exit learning)
- warning: 2–5×
- fail: <2× — LinkedIn's optimizer cannot exit learning; campaign stays in "Limited" state and
over-pays per conversion
- fix: raise budget or merge campaigns; LinkedIn's floor is much higher than Meta's
### L62 · Bid strategy chosen, not defaulted
- weight: 4 · severity: medium · fix_effort: low
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial, marketplace]
- pass: bid strategy explicitly set — `Maximum Delivery` (autobid, default; OK for new
campaigns), `Target Cost` (soft CPA cap; use after 30 conversions), or `Manual CPC` (bid
floor; use only when capping a runaway campaign)
- warning: default Maximum Delivery on every campaign post-learning (no Target Cost gate)
- fail: Manual CPC across the account when LinkedIn has CVR data (under-bidding the algorithm's
preferred strategy)
- fix: switch each campaign to Target Cost once it crosses 30 conversions in 30 days
### L63 · Campaign group structure matches funnel
- weight: 3 · severity: low · fix_effort: medium
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- pass: campaign groups organized by funnel stage (Awareness / Demand-Gen / Retargeting / ABM)
with budget pooled at the group level
- warning: groups organized by quarter or campaign name (no funnel structure)
- fail: no campaign groups at all (every campaign in the default group) — no budget pooling, no
funnel-level optimization
- fix: re-group campaigns; group-level budget enables campaign-level experimentation without
spend-cap shocks
### L64 · LinkedIn-specific UTMs flow into CRM
- weight: 4 · severity: medium · fix_effort: medium
- applies_to_archetypes: [b2b_saas, local_service, lead_gen_financial]
- primary_metric_kind: cost_per_sql
- pass: HubSpot / Salesforce / Pipedrive captures `utm_source=linkedin` + `utm_campaign` +
`utm_content` on every lead's first-touch + last-touch; revenue attribution reports can roll
up to campaign + creative
- warning: source captured but campaign/creative not (no creative-level pipeline ROI)
- fail: UTMs not captured — pipeline reports cannot reach LinkedIn campaign-level
- fix: configure CRM to capture all 5 standard UTMs to lead + contact + deal records; first- +
last-touch wired
---
## Critical-checks list (severity multiplier 5.0×)
These ALWAYS surface in audit output, even when they pass. A FAIL on any critical caps that
category's sub-score at 50:
- L01 — Insight Tag installed + firing
- L02 — Conversions API (LCAPI) active
- L04 — Pipeline-stage events uploading (when applicable)
- L20 — Matched Audiences from CRM lists
---
## Health-score formula
Per category: `sub_score = sum(check_score × check_weight × W_sev) / sum(100 × check_weight ×
W_sev)` over fired checks (`check_score ∈ {0, 50, 100}` for fail / warning / pass; N/A excluded;
`W_sev = {critical 5.0, high 3.0, medium 1.5, low 0.5}`).
Final:
```
health = 0.30 · tracking_score
+ 0.30 · audience_score
+ 0.20 · creative_score
+ 0.20 · structure_score
```
Grade letters: 90–100 A · 80–89 B · 70–79 C · 60–69 D · <60 F. A critical FAIL caps that
category's sub-score at 50.
---
## Archetype filtering at audit time
Drop checks that don't apply to the brand's archetype. Almost all LinkedIn checks here are
tagged `[b2b_saas]` or with `local_service` / `lead_gen_financial` / `marketplace` (supply-side)
added where the archetype rationally uses LinkedIn paid. D2C / subscription_consumer /
creator_media / mobile_app brands will see zero LinkedIn checks after filtering — correct
behavior, since those archetypes shouldn't be running LinkedIn paid.
`primary_metric_kind` resolves to `cost_per_sql` for b2b_saas, `cost_per_qualified_lead` for
local_service / lead_gen_financial, and `cost_per_first_transaction` for marketplace.
---
## What's NOT in this catalog (and why)
- **Sales Navigator / Recruiter checks** — different product surface; would need its own audit
catalog if needed.
- **LinkedIn Live / Events** — organic; out of scope.
- **InMail / Message Ads** — covered in L40 format diversity but no separate hygiene checks; the
format has well-documented best practices but most failure modes are deliverability (open
rate) rather than something a static checklist can grade. Defer until a brand actually scales
InMail spend.
- **LinkedIn Audience Network (off-platform inventory)** — covered in L24 only; not its own
category. The fail modes are documented but the leverage is small.
Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.
No comments yet. Be the first to comment!
