Thought Leader Waterfall

Why it works right now

Campaign architecture

• Stage 1 — Trust (50–60% of budget): TLAs from the founder/SME's profile (2–3 posts live, image + short video), engagement/video-view objective, full ICP audience 10K–30K (company list or industry + function + seniority). No CTA beyond the content.
• Stage 2 — Educate (25–30%): Document Ads and webinar/event ads to Stage-1 engagers, video viewers (25%+), and 90-day site visitors; Lead Gen Forms for content offers.
• Stage 3 — Convert (15–20%): demo/consult ask to warm-only: Stage-2 engagers, form-openers, pricing-page visitors. Landing page (not native form) for demo requests — practitioners report 20–40% better SQL rates from the added friction^L19. Exclude each stage's audience from the one above it so prospects graduate.
• Bidding: manual CPC at Stage 1 (cap the TLA arbitrage), max delivery at Stages 2–3. Audience floor 300; ABM sweet spot 1K–30K^L10,L30.

Creative & offer requirements

• 2–3 executive/SME posts per month written for the feed (opinionated, first-person, no gating) — the client must commit a named person, not a logo.
• 1 substantial mid-funnel asset per quarter (report, teardown, benchmark) cut into a 6–10 page Document Ad; 1 webinar or event per quarter.
• Refresh Stage-1 creative every 4–6 weeks; Document Ads run ~11 weeks before fatigue^L17. Trim single-image to 40–50% of impressions — the 61.87% spend share it gets by default is over-allocation^L4.

Setup checklist

• Insight Tag + Website Actions enabled; CAPI/server-side conversions (advertisers with CAPI see ~20% lower CPA)^L5.
• Matched audiences: company list upload (expect B2B match rates of only 10–20% on personal-identifier platforms — LinkedIn's company match is the exception and the point)^X30; retargeting audiences per stage with exclusions.
• CRM connected for company-level Revenue Attribution^L11; lead-scoring stage defined pre-launch.
• TLA permissions requested from each poster; verify legacy geo targeting is off (unsupported after Aug 31, 2026)^L13.

Expected performance ranges (benchmarks, not guarantees)

Timeline

1. 01

   Week 1–2: setup, TLA permissions, launch Stage 1; don't touch for 5–10 days^L16,L32.

2. 02

   Week 3–4: Stage 2 live once engager pool passes ~1K. First read on CTR/engagement quality.

3. 03

   Month 2: Stage 3 live; first MQLs/demos. Judge creative and audience signal, not pipeline.

4. 04

   Month 3+: first pipeline proof; report on 6-month cohorts, never calendar-month last-click^L16. Success gate at day 90: engaged-audience growth, form completion ≥10%^L4, demo CPL trending toward target.

Scaling & optimization playbook

• When it works: add posters (2–3 SMEs); widen the ICP list before raising frequency; Conversation Ads only for high-value plays against 15K+ audiences^L14; LinkedIn CTV at Scale.
• When it doesn't: kill formats, not stages — video for clicks is the worst performer (0.24% CTR, $15.61 CPC)^L4. If Stage-3 CPL is fine but SQL rate isn't, the fix is targeting seniority and the demo page, not the ads.
• Key levers: poster quality, audience tightness, stage-exclusion hygiene, offer strength at Stage 2.

Risks & failure modes

• 1 · The client won't feed the person-led machine. No executive posts, no TLAs, no play. Prevent: get 6 posts drafted and approved before signing the scope.

• 2 · Judged on 30-day last-click. LinkedIn always loses that report (0.3–0.5x at 30 days)^L16. Prevent: set 6-month cohort reporting in the contract; show leading indicators monthly.

• 3 · Budget below frequency threshold. $1,500/month against a 50K audience is ~0.6 monthly frequency — invisible^L7. Prevent: shrink the audience to match the budget, or don't run.

• 4 · Engagement-quality mirage. TLA clicks skew to profile/post engagement^L4. Prevent: never judge TLAs on CPL; judge the waterfall end-to-end.

Minimum viable version