Kai Abm

Kai ABM Skill

Plan and execute account-based marketing campaigns: account selection, personalization, multi-channel touches, and measurement.

Phase 0: Load Product Context

Check if MARKETING.md exists in the project root (same directory as CLAUDE.md, README.md, package.json).

If it exists: Read it — skip product discovery questions. It has the product name, ICP, value prop, monetization, brand voice, current channels, and competitive landscape.

If it does NOT exist: Auto-explore the codebase to create it in the project root (next to CLAUDE.md). Do NOT ask the user what the product is. Read CLAUDE.md, README.md, PROJECT.md, package.json, landing pages, and any project files. Search for email/ad/analytics config. Then create MARKETING.md using the template from /kai-email-system. Present draft to user for confirmation.

Phase 1: Discovery

Read from MARKETING.md. Only ask about things not covered there:

1. ICP definition — What does the ideal target company look like? (industry, size, revenue, tech stack, signals)
2. Target account list — Named accounts or criteria for building the list?
3. Buying committee — Who are the decision-makers, influencers, and blockers? (titles, roles)
4. Product/offer — What are we selling? Price point, deal size, sales cycle length.
5. Current pipeline — Any existing relationships or warm contacts at target accounts?
6. Sales alignment — Is sales involved? What's the handoff point?
7. Channels available — Email, LinkedIn, ads, direct mail, events, phone?
8. Budget — Per-account spend range
9. Persona alignment — Which harness persona(s) map to the buying committee? Load from knowledge/personas/_persona-index.md

Phase 2: Plan

Build the ABM campaign architecture:

1. Load ABM playbook: knowledge/playbooks/account-based-marketing.md
2. Tier accounts by fit and intent:
   • Tier 1 (1:1): Top 10-25 accounts. Fully personalized. High-touch.
   • Tier 2 (1:few): 25-100 accounts. Cluster by industry/use case. Semi-personalized.
   • Tier 3 (1:many): 100-500 accounts. Programmatic with light personalization.
3. Account research template (Tier 1):
   • Company priorities (earnings calls, press releases, job postings)
   • Technology stack (what they use today)
   • Key personnel (LinkedIn profiles, content they publish, mutual connections)
   • Pain signals (hiring patterns, tech changes, complaints in reviews)
   • Competitive landscape (who else is selling to them)
4. Multi-channel touch sequence:
   • Week 1-2: Warm-up (LinkedIn engagement, content sharing, ad impressions)
   • Week 3-4: Direct outreach (personalized email, LinkedIn message)
   • Week 5-6: Value delivery (relevant content, case study, invite to event)
   • Week 7-8: Conversion push (meeting request, demo offer, executive intro)
5. Content mapping — Map existing assets to buying stages and personas
6. Ad targeting — Account-level targeting on LinkedIn, programmatic display
7. Sales/marketing SLA — Response times, follow-up rules, feedback loops

Phase 3: Produce

Create ABM campaign assets:

1. Account briefs (Tier 1) — One-page dossier per account with research, contacts, and approach
2. Personalized email sequences — 4-6 touches per tier, customized by account/cluster
3. LinkedIn outreach templates — Connection request + message sequence
4. Targeted ad copy — Account-aware messaging for LinkedIn and display ads
5. Custom content (Tier 1) — Account-specific landing pages, case studies, or presentations
6. Event invitations — Personalized invites to webinars, dinners, or roundtables
7. Sales enablement — Account cheat sheets for sales team with talking points

Apply harness writing rules:

• Conditions AFTER main clause
• Instructions start with verbs
• Short sentences, high specificity
• Bold the answer

Phase 4: Quality Gates

Validate before launch:

1. Four U's Score: python scripts/quality_gates/four_us_score.py <file>
   • Minimum: 10/16 for emails and outreach
   • Minimum: 12/16 for content and landing pages
2. Banned Word Check: python scripts/quality_gates/banned_word_check.py <file>
3. AI Slop Check — Zero filler. ABM demands precision.
4. Personalization depth check — Does each Tier 1 message reference something specific to the account?
5. Platform policy compliance — Check ad copy against platform rules:
   • LinkedIn: harness/references/linkedin-ads-rules.md
   • Google Display: harness/references/google-ads-policy-reference.md
6. CAN-SPAM / cold email compliance: harness/references/cold-email-rules.md

Max 2 auto-retry cycles on gate failures.

Phase 5: Output

Deliver the ABM campaign package:

• ABM strategy document (ICP, tiers, channels, timeline, budget)
• Account briefs (Tier 1 dossiers)
• Email sequences (per tier)
• LinkedIn outreach templates
• Ad copy (per platform)
• Content map (existing assets mapped to buying stages)
• Sales enablement materials (account cheat sheets)
• Measurement framework (engagement score, pipeline influence, deal velocity)
• Gate pass/fail summary

Write output to workspace/ with filename pattern: abm-campaign-YYYY-MM-DD.md