Apollo Icp Engine

Teqfocus Apollo ICP Engine — Master Skill

This skill governs all Apollo data operations for Teqfocus. It ensures every contact pull, scoring run, and segmentation output is aligned to the defined ICP — with zero manual Excel work.

How to Use This Skill

1. Read this SKILL.md fully to understand the ICP and workflow
2. Identify which operation is needed → load the relevant reference file
3. Execute the operation following the reference file instructions
4. Always present a summary of what was pulled, scored, or segmented

Sub-Skill Reference Files

Section 1: The Teqfocus ICP — Source of Truth

Never deviate from this definition. Every Apollo pull must filter against these.

Target Personas

C-Suite (Highest Priority — +3 pts)

• CIO / Chief Information Officer
• CDO / Chief Data Officer
• Chief Digital Officer
• Chief Digital Transformation Officer
• CRO / Chief Revenue Officer
• Chief Customer Officer / Chief Experience Officer

VP Level (+2 pts)

• VP of Technology / Business Technology / Information Technology
• VP of Go-To-Market
• VP of Digital Transformation
• VP of Customer Experience
• VP of Enterprise Technology
• VP of Patient Experience / Patient Support
• VP of Data / VP of Analytics / VP of Engineering

SVP Level (+2 pts)

• SVP of Technology / Business Technology
• SVP of Go-To-Market
• SVP of Digital Transformation
• SVP of Customer Experience / Enterprise Technology
• SVP of Patient Experience / Patient Support
• SVP of Data / SVP of Analytics

Apollo Seniority Mapping

• C-Suite personas → person_seniorities: ["c_suite"]
• VP personas → person_seniorities: ["vp"]
• SVP personas → include in person_titles explicitly (Apollo has no SVP bucket)
• Always run separate searches for C-Suite and VP/SVP to maximize coverage

Target Industries

CRITICAL RULE: Always run one search per industry vertical — never combine all four into a single search. Tag each batch at source so you always know which vertical a contact came from.

Company Size Bands

Apollo employee range strings:

• Mid-Market: "200,1000"
• Upper Mid-Market: "1001,5000"
• Enterprise: "5001,10000", "10001,1000000"

Target Geographies

• United States, Canada, United Kingdom, Australia

Excluded Industries (never pull from these)

• Staffing & Recruiting
• Non-Profit / Government / Public Sector
• Education
• Real Estate

Section 2: Standard Workflow

Every time a user asks to "pull a list" or "find contacts", run this sequence:

STEP 1 → SEARCH    (apollo-search.md)    Pull per-industry, per-seniority
STEP 2 → SCORE     (icp-scoring.md)      Score each contact 0–9 pts
STEP 3 → CLEAN     (data-cleaner.md)     Dedupe, standardize, flag gaps
STEP 4 → SEGMENT   (segmentation.md)     Split by vertical + persona + tier
STEP 5 → SUMMARIZE                       Present counts table to user

Always present a final summary table in this format:

PULL SUMMARY — [Date]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Total contacts pulled:     [N]
After deduplication:       [N]
After ICP scoring:
  Tier 1 (7–9 pts):        [N]  ← Priority ABM targets
  Tier 2 (5–6 pts):        [N]  ← Standard outbound
  Tier 3 (3–4 pts):        [N]  ← Nurture only
  Excluded (<3 pts):       [N]

BY INDUSTRY:
  HLS:      [N] contacts  (T1: X | T2: X | T3: X)
  FinServ:  [N] contacts  (T1: X | T2: X | T3: X)
  SaaS:     [N] contacts  (T1: X | T2: X | T3: X)
  HiTech:   [N] contacts  (T1: X | T2: X | T3: X)

BY PERSONA:
  C-Suite:  [N]
  VP:       [N]
  SVP:      [N]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

Section 3: Credit Conservation Rules

Apollo enrichment (revealing full names + emails) costs credits. Never enrich all contacts. Always score first, enrich after.

Rule: Only enrich Tier 1 contacts (score 7–9) unless user explicitly requests more. Always tell the user how many enrichments will be used before calling apollo_people_match.

Section 4: Known Apollo API Constraints

1. Last names are obfuscated in search results — enrichment required for full name
2. SVP has no dedicated seniority filter — must use explicit title strings
3. Industry not returned in people search results — tag industry at search time
4. Max 100 results per page — paginate for larger pulls
5. 50,000 record display limit — add filters if results exceed this
6. include_similar_titles: false — set this for strict ICP title matching

Section 5: Full Apollo Tool Inventory

These are the Apollo tools available in this project. Know which to use for each task.

When to Use Each Search Tool

Use apollo_mixed_people_api_search when:

• Prospecting for new ICP-fit contacts not yet in your CRM
• Running the standard 8-search ICP pull matrix
• Building a net-new list for outbound

Use apollo_contacts_search when:

• User says "find contacts we already have" or "search our database"
• Looking up a specific person by name, title, or company already in Apollo
• Verifying if a contact exists before enriching or enrolling
• Supports: q_keywords, sort_by_field, sort_ascending, page, per_page

Sequence Enrollment — Critical Warning

apollo_emailer_campaigns_add_contact_ids sends real emails from a real mailbox. This action is irreversible once emails are dispatched. Always load references/sequence-enrollment.md before calling this tool. Never enroll contacts without explicit user confirmation — even if user says "just do it".

Section 6: Extended Workflow (Search → Score → Enrich → Enroll)

When user wants to go from pull all the way to sequence enrollment:

STEP 1 → SEARCH      apollo_mixed_people_api_search   Net-new ICP prospects
STEP 2 → SCORE       icp-scoring.md                   Score 0–9, assign tier
STEP 3 → CLEAN       data-cleaner.md                  Dedupe, standardize
STEP 4 → SEGMENT     segmentation.md                  Split by tier + vertical
STEP 5 → ENRICH      apollo_people_match               Tier 1 only — confirm credits first
STEP 6 → ENROLL      sequence-enrollment.md            Confirm sequence + sender + count

Never skip the confirmation gates at Steps 5 and 6. Both cost something — credits (Step 5) or real email sends (Step 6).