Research Strategy

Inputs required

Before starting, confirm these inputs with the user:

Step-by-step workflow

Step 1: Read relevant context files

Always read before starting:

• brands/{brand}/brand.md — Mission, voice, positioning statement, approved phrases
• brands/{brand}/audience.md — Target persona details, pain points, objections, buying triggers
• brands/{brand}/product.md — Features, pricing, differentiators relevant to this segment
• brands/{brand}/competitors.md — Competitive landscape, messaging gaps, comparison tables

Step 1b: Supplement with live research via WebSearch

After reading context files, use the WebSearch tool to fill gaps the static files can't cover:

• Current competitor pricing or positioning changes (e.g., "SEMrush pricing 2026", "Apollo.io new features")
• Market trends relevant to this segment (e.g., "AI search optimization trends 2026", "B2B SaaS lead generation benchmarks")
• Geographic market intelligence for ID/SG/AU campaigns (e.g., "B2B SaaS market Indonesia 2026")
• Recent industry news that could inform campaign timing or messaging

Only query WebSearch when context files lack the information needed. Cite web findings separately from context file data in your output.

Step 1c: Keyword research via DataforSEO (when task involves paid ads or SEO)

When the task involves paid ads strategy, keyword research, or SEO analysis, use the DataforSEO API to pull real search volume, CPC, and competition data.

Search Volume for specific keywords:

Use gateway MCP tool `dataforseo_search_volume`:
- fiveagents_api_key: ${FIVEAGENTS_API_KEY}
- keywords: ["keyword1", "keyword2", "keyword3"]
- location_code: 2702
- language_code: "en"
- date_from: "2025-04-01"
- date_to: "2026-03-31"

Keyword suggestions (expand from seeds):

Use gateway MCP tool `dataforseo_keyword_suggestions`:
- fiveagents_api_key: ${FIVEAGENTS_API_KEY}
- keywords: ["seed keyword 1", "seed keyword 2"]
- location_code: 2702
- language_code: "en"
- date_from: "2025-04-01"
- date_to: "2026-03-31"

Location codes: See DataforSEO docs for full list. Examples: Singapore=2702, Malaysia=2458, Indonesia=2360, Australia=2036, United States=2840.

Response parsing: Results are in tasks[0].result. Each keyword has: keyword, search_volume, cpc (USD), competition, competition_index, monthly_searches[]. Convert CPC from USD to the brand's local currency using the exchange rate from brands/{brand}/brand.md Locale section.

Competition labels: competition_index 0–100. <30 = Low, 30–70 = Medium, >70 = High.

When to use DataforSEO:

• Paid ads keyword research (search volume, CPC, competition)
• SEO keyword opportunity analysis
• Comparing search demand across markets (SG vs MY vs ID)
• Validating keyword assumptions with real data

When NOT to use: General market research, competitor analysis, content strategy (use WebSearch instead).

Step 2: Define the ICP

Using the audience persona as your base, build a specific ICP for this campaign:

• Role, seniority, company size, industry
• Primary goal and biggest blocker
• Current tools they use (from audience.md — for context only, never for competitive positioning)
• Pick the 2-3 most relevant buying triggers from brands/{brand}/audience.md → Buying triggers, and build the campaign hook around them
• Language and tone preferences (EN or ID)

IMPORTANT: "Current tools" listed under each persona in audience.md are tools the persona already uses — they are NOT competitors. Never position against them unless they are explicitly listed in brands/{brand}/competitors.md.

Step 3: Define positioning for this campaign

Read brands/{brand}/competitors.md to identify the brand's messaging gaps vs. competitors — these are the positioning angles available for this campaign. Choose 1-2 that are most relevant to the target persona and campaign goal.

IMPORTANT — competitors.md is the only source for competitive positioning. Only name companies listed there. Do not reference tools mentioned in audience.md (e.g., a persona's current stack) as competitors unless they appear in competitors.md.

Write a single positioning statement (1-2 sentences) using approved phrases from brands/{brand}/brand.md.

Step 4: Build the messaging framework

For each core message, define:

• Claim: What the brand does for this persona
• Proof point: Feature or capability from brands/{brand}/product.md that backs it up
• Competitive angle: How this beats the alternative — use only companies from brands/{brand}/competitors.md
• Objection handled: Use the pre-built objection responses from brands/{brand}/audience.md → Objections as the base. Reference the exact objection label (e.g., "It's too expensive") and use the documented response angle — do not invent new ones.

Produce 3-5 core messages maximum per campaign.

Step 5: Define the offer

Based on brands/{brand}/product.md pricing and the persona's plan tier:

• Which plan fits this persona (per brands/{brand}/product.md)?
• What is the trial or entry offer?
• Are there add-ons or managed service options relevant to this persona?
• What is the primary CTA? (e.g., "Start free trial", "Book a demo", "See pricing")

Step 6: Produce recommendations

Summarize into 3-5 clear, prioritized strategic recommendations:

• Which channel(s) to focus on for this persona
• Which 1-2 messaging angles are strongest
• Which competitor to position against — only from brands/{brand}/competitors.md
• Which buying trigger from brands/{brand}/audience.md to lead with
• What content formats will resonate (landing page, email, comparison page, etc.)

Output format

Save location — local workspace:

outputs/{brand}/strategy/

Folder by research type:

Naming convention:

[ResearchType]_[DDMonYYYY].md

Examples:

• StrategyBrief_10Mar2026.md
• CompetitorAnalysis_10Mar2026.md
• ICPDefinition_10Mar2026.md
• MarketResearch_10Mar2026.md

Output metadata:

---
Date: YYYY-MM-DD
Skill Used: research-strategy
Persona: [Persona name from audience.md]
Campaign: [Campaign name]
Market: EN | ID
Status: Draft | Final
---

Output sections:

1. ICP Summary — Who we're targeting (3-5 bullet points)
2. Positioning Statement — 1-2 sentence campaign positioning
3. Core Messages — 3-5 messages with claim, proof, competitive angle, objection handled
4. Offer — Plan tier, trial, CTA
5. Recommendations — 3-5 prioritized strategic actions

Quality checklist

Before finalizing any strategy output:

• All product features cited exist in brands/{brand}/product.md
• Pricing mentioned is accurate per brands/{brand}/product.md
• Persona details match brands/{brand}/audience.md
• Buying triggers selected from brands/{brand}/audience.md → Buying triggers (not invented)
• Objection responses use the pre-built angles from brands/{brand}/audience.md → Objections
• Every competitor named in the brief exists in brands/{brand}/competitors.md — no exceptions
• No tool from a persona's "Current tools" (audience.md) is positioned against unless it also appears in competitors.md
• No invented customer logos, testimonials, or case studies
• No guaranteed promises or timeframes (unless documented in product.md)
• Positioning uses approved phrases from brands/{brand}/brand.md where applicable
• Output saved to outputs/{brand}/strategy/ with correct naming and metadata
• Agent run logged to dashboard

Final Step — Log to Dashboard

See docs/new_agent_onboarding/metrics-spec.md for the full JSONB contract.

Use gateway MCP tool `fiveagents_log_run`:
- fiveagents_api_key: ${FIVEAGENTS_API_KEY}
- skill: "research-strategy"
- brand: "<active-brand>"
- status: "<success|failed>"
- summary: "<1 line, <200 chars>"
- started_at: "<ISO timestamp>"
- completed_at: "<ISO timestamp>"
- metrics: {
    "date": "YYYY-MM-DD",
    "type": "<competitive-analysis|icp|positioning|campaign-brief>",
    "persona": "<slug>",
    "markets": ["SG", "ID", "MY"],
    "campaign": "<campaign name>",
    "content_status": "Final",
    "competitors_analyzed": 0,
    "core_messages": ["..."],
    "keywords_analyzed": 0,
    "deliverable": "<filename>",
    "output_path": "outputs/{brand}/strategy/"
  }