LinkedIn Insight Tag and conversion tracking checklist

fact_check Audit scorecard

1. 1

   Insight Tag present and firing on every paid path

   Open a paid landing URL, thank-you page, and pricing page in an incognito session with the LinkedIn helper or your tag debugger. Confirm the Insight Tag loads once per page, not twice from GTM plus hardcode. Good: partner ID matches the ad account, no duplicate hits, no 404 on the tag endpoint. Fail costs you retargeting pools, conversion attribution, and any bid strategy that needs post-click events. US B2B stacks often break when IT ships a new SPA route and nobody re-fires the tag on client-side navigation.

2. 2

   Conversions defined as pipeline events, not raw form spam In Campaign Ma

   Conversions defined as pipeline events, not raw form spam In Campaign Manager, list every conversion action tied to active campaigns. Score pass only if primary optimization points at something sales would defend: qualified demo request, meeting booked, opportunity created, or a gated asset that requires work-email validation. Weak: “any form submit” or “page view of /contact.” Fail: optimizing solely to Lead Gen Form opens or button clicks. LinkedIn will happily fill Instant Forms when the signal is shallow. You then pay for CPL while SDR calendars stay empty.

3. 3

   Event firing rules match the real thank-you moment For website conversio

   Event firing rules match the real thank-you moment For website conversions, confirm the event fires once on successful submit, not on page load of a multipurpose URL. For Lead Gen Forms, confirm the form completion event is the one attached to the campaign objective. Check GTM triggers for double-fire on validation errors. Good: one conversion per successful lead with a stable event name. Fail: inflated conversion counts that make ROAS look healthy while CRM shows 40% junk. Privacy loss already knocks 20%+ of web conversion data in many stacks; duplicate events make the remaining signal worse, not better.

4. 4

   UTM and click ID hygiene into the CRM Spot-check 20 recent LinkedIn lead

   UTM and click ID hygiene into the CRM Spot-check 20 recent LinkedIn leads in the CRM. Every row should carry campaign, creative or ad ID where available, and a consistent source/medium. LinkedIn click identifiers should survive the hop from ad to form to Salesforce, HubSpot, or Marketo. Good: marketing can filter “LinkedIn paid, last 30 days, meeting held” without a spreadsheet merge. Fail: source = “website” or blank UTM on Instant Form sync. Without that spine you cannot send offline wins back or kill wasteful ads with evidence.

5. 5

   Lead Gen Forms vs landing pages: intentional split Document when you use

   Lead Gen Forms vs landing pages: intentional split Document when you use Instant Forms versus a full page. Instant Forms win on friction for top-of-funnel content. Dedicated pages win when you need proof, pricing context, or multi-step qualification. Good: form fields map to ICP (company size, role, use case) and sync fields match CRM required properties. Fail: Instant Forms with three vanity fields, no enrichment, and no sales SLA. Mid-market teams in markets like Chicago and Atlanta often see form volume spike while held-meeting rate collapses because the form never asked the one disqualifier sales actually uses.

6. 6

   Landing page message match and post-click proof For traffic ads, open th

   Landing page message match and post-click proof For traffic ads, open the ad creative next to the destination. Headline promise, offer, and CTA must match above the fold. Proof (logos, short case metrics, security notes for enterprise) must sit before the form, not after. Run a Core Web Vitals pass on the landing URL; only about 55.9% of origins pass all three in recent CrUX data, and B2B pages with heavy chat widgets often fail INP on mobile. When traffic does not convert, session-level behaviour matters. HeyLead Insights helps you see scroll depth, rage clicks, and form abandon on those LinkedIn landers without guessing. Fail cost: you scale spend into a page that never earned the click.

7. 7

   Matched audiences and retargeting pools are real

   Check Insight Tag-based website retargeting segments and engagement retargeting. Confirm exclusions for customers, recent opportunity owners, and job applicants. Good: pools large enough to serve without constant “audience too small” warnings, refreshed on a known cadence. Fail: tag never collected, or you retarget everyone who hit the homepage including job seekers. Wasted retargeting CPMs are quiet. They still eat budget that should train conversion campaigns.

8. 8

   Offline or CRM-stage conversions wired back If your sales cycle is weeks

   Offline or CRM-stage conversions wired back If your sales cycle is weeks, website pixel events alone will under-teach LinkedIn. Inspect whether qualified stages (SQL, opportunity, closed-won) can be uploaded or streamed as conversions with timestamps and matching keys. Good: at least one downstream stage feeds optimization or reporting weekly. Weak: quarterly CSV experiments nobody trusts. Fail: Campaign Manager only knows form fills while finance knows CAC from closed revenue. Long B2B cycles in SaaS and professional services make this the difference between “LinkedIn does not work” and “LinkedIn only saw the wrong event.”

9. 9

   Conversion window and attribution settings fit the cycle Read the click

   Conversion window and attribution settings fit the cycle Read the click and view windows on each conversion. A seven-day click window on a 45-day enterprise motion will starve learning. An aggressive view-through window will credit noise. Good: windows negotiated with sales cycle reality and documented so finance does not fight marketing dashboards every month. Fail: default windows nobody reviewed since the account was opened. You do not need perfect multi-touch theatre. You need windows that do not lie about which ads deserve budget.

10. 10

    Campaign objectives align with the conversion you score Website conversi

    Campaign objectives align with the conversion you score Website conversions campaigns should optimize to the qualified event, not a proxy click. Lead gen campaigns should optimize to completed forms that pass your quality bar, with creative that sets the same expectation as sales outreach. Good: one primary conversion per campaign cluster, secondary conversions for diagnostics only. Fail: awareness objective with conversion reporting bolted on, or three competing conversion goals in one campaign so the algorithm never settles.

11. 11

    Consent, ad blockers, and first-party fallbacks US state privacy rules a

    Consent, ad blockers, and first-party fallbacks US state privacy rules and browser restrictions already thin client-side tags. Confirm consent mode or CMP behaviour does not silently block the Insight Tag for the majority of paid sessions. Prefer server-side or CRM-backed signals where you can. Good: you know roughly what share of conversions never hits the browser tag and you compensate with offline imports. Fail: you treat Campaign Manager totals as complete truth while 20%+ of conversion paths never report. Smart delivery then optimizes on a biased sample.

12. 12

    Reporting joins spend to meetings, not only CPL Your weekly view should

    Reporting joins spend to meetings, not only CPL Your weekly view should show spend, qualified leads, meetings held, and pipeline created by campaign. CPL alone is theatre when AEs reject half the queue. Good: a simple join key from ad account to CRM stages updated on a fixed cadence. Fail: marketing celebrates LinkedIn CPL while sales ops quietly marks the source as low intent. If you cannot answer “which LinkedIn conversion definition correlates with held meetings,” stop scaling until you can.