Gbp Suite

Research source (Exa)

Default: Exa, per .claude/rules/exa-protocol.md (auto-loaded for research, audit, competitor, ICP, AEO, content sourcing, sales prospecting work).

Primary Exa tools for this skill: web_search_exa.

Use case: competitor review pattern research.

Tool surface during the migration window:

• New plugin (preferred): mcp__plugin_exa_exa__web_search_exa (after claude plugin i exa@claude-plugins-official).
• Legacy MCP (still mounted): mcp__exa__web_search_exa.
• Both backends route to the same Exa API — they don't double-bill.

Citation: every Exa-derived claim uses [VERIFIED: exa_search, {url}, accessed {YYYY-MM-DD}] per .claude/rules/ontology.md.

Quality gate (research outputs): ≥3 sources per major claim, ≥50% [VERIFIED] confidence, date filter for any "recent / latest" claim, no fallback to WebSearch without flagging the data gap.

Worked examples + tool catalog: .claude/skills/meta-skills/exa/.

GBP review strategy

Analyzes competitor review velocity, keyword mentions in reviews, neighborhood mentions, and recurring complaints. Then generates review response templates (5-star, 4-star, 3-star, 1-2 star) with 3 variations each that naturally incorporate service + location keywords. Covers article prompts #3 (competitor review teardown) and #4 (review response strategy) from the local SEO playbook.

Doctrine inherited (Step 7 — 0626 rollout, locked 2026-06-04)

Output complies with:

• output-tenets.md, output-simplicity.md
• Step 6 calibration: see [[feedback_execution_doctrine_refinements_step6]]

Refinements applied to this skill:

Process Flowchart

┌──────────────────────────────────────────────────────────────┐
│ GBP REVIEW STRATEGY PROCESS │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌──────────────────────────────────────────────────────────────┐
│ INPUT VALIDATION │
│ Required: │
│ □ Client GBP URL │
│ □ 2-3 competitor GBP URLs │
│ □ Target keywords (3+) │
│ □ Service areas (neighborhoods/cities) │
│ Optional: Current review count, response rate baseline │
│ → If missing: Ask for GBP URLs and target keywords │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌──────────────────────────────────────────────────────────────┐
│ PHASE 1: REVIEW DATA EXTRACTION │
│ □ Scrape last 50 reviews per listing (client + competitors) │
│ □ Extract: total count, avg rating, 30/60/90 day velocity │
│ □ Extract: mentioned services, neighborhoods, complaints │
│ ✓ Checkpoint: Review data captured for all listings │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌──────────────────────────────────────────────────────────────┐
│ PHASE 2: VELOCITY ANALYSIS │
│ □ Calculate reviews/month for each listing (30/60/90 day) │
│ □ Identify top competitor by velocity │
│ □ Calculate reviews/month needed to catch top competitor │
│ □ Estimate time-to-parity at target velocity │
│ ✓ Checkpoint: Velocity gap quantified with catch-up target │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌──────────────────────────────────────────────────────────────┐
│ PHASE 3: RESPONSE AUDIT │
│ □ Analyze owner responses: response rate, avg response time │
│ □ Check keyword usage in existing responses │
│ □ Evaluate tone and negative review handling │
│ □ Compare response quality across all listings │
│ ✓ Checkpoint: Response gaps identified vs. competitors │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌──────────────────────────────────────────────────────────────┐
│ PHASE 4: TEMPLATE GENERATION │
│ □ Create 5-star templates (3 variations) │
│ □ Create 4-star templates (3 variations) │
│ □ Create 3-star templates (3 variations) │
│ □ Create 1-2 star templates (3 variations) │
│ □ Each template includes service keywords + location mention │
│ ✓ Checkpoint: 12 templates ready, keywords naturally placed │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌──────────────────────────────────────────────────────────────┐
│ PHASE 5: STRATEGY DOCUMENT │
│ □ Set monthly review velocity target │
│ □ Identify where/when to ask for reviews │
│ □ Define what to ask customers to mention (service + area) │
│ □ Create 90-day review growth roadmap │
│ ✓ Checkpoint: Actionable strategy with measurable targets │
└──────────────────────────────────────────────────────────────┘
                              │
                              ▼
┌────────────────────────────────────┬─────────────────────────┐
│ REVIEW GATE: Level 1 (Quick) │ CHAIN SUGGESTIONS │
├────────────────────────────────────┼─────────────────────────┤
│ Present: Velocity analysis, │ → gbp-content-engine │
│ response audit, 12 templates, │ → gbp-listing-opt │
│ strategy doc │ → content-strategy │
│ Actions: [Approve] [Adjust] │ → Export to Google Docs │
│ [Add competitors] │ │
└────────────────────────────────────┴─────────────────────────┘

Claude Code Triggers

Invoke this skill when user says:

• "Review strategy for [business]"
• "Review teardown"
• "Competitor review analysis"
• "Review response templates"
• "How do I get more reviews?"
• "Review velocity analysis"
• "Help me respond to reviews"
• "Google review strategy"
• "Review gap analysis"

Do NOT invoke when:

• User wants full local SEO audit (use /local-seo-audit orchestrator)
• User wants GBP category or attribute analysis (use /gbp-category-audit)
• User wants GBP posts or photo strategy (use /gbp-content-engine)
• User wants services section or description optimization (use /gbp-listing-optimization)

Input Requirements

Required Inputs

Optional Inputs (improve quality)

Input Validation Checklist

Before proceeding, verify:

• Client GBP URL is accessible and has reviews
• At least 2 competitor GBP URLs provided
• At least 3 target keywords specified
• Service areas defined (neighborhoods or cities)

If inputs are missing: Ask for client GBP URL first. Offer to research competitors via Exa/Firecrawl if user doesn't have competitor URLs.

Process (Step-by-Step)

Phase 1: Review data extraction

Purpose: Scrape and structure review data from client and competitor GBP listings.

Steps:

1. Step 1.1: Scrape client reviews

   • Pull last 50 reviews from client GBP URL
   • Extract: reviewer name, rating, date, review text, owner response (if any)
   • Output: Client review dataset

2. Step 1.2: Scrape competitor reviews

   • Pull last 50 reviews per competitor GBP URL (2-3 competitors)
   • Extract same fields as client
   • Output: Competitor review datasets

3. Step 1.3: Parse review content

   • For each listing, extract:
     • Total review count and average rating
     • Services mentioned (map to target keywords)
     • Neighborhoods/locations mentioned
     • Recurring complaints (grouped by theme)
     • Recurring praise (grouped by theme)
   • Output: Parsed review content analysis per listing

Phase 1 Checkpoint:

• Review data captured for client + all competitors
• Services, neighborhoods, and complaints extracted
• Data is sourced — no invented review counts or ratings

Phase 2: Velocity analysis

Purpose: Quantify review generation speed and calculate the gap to close.

Steps:

1. Step 2.1: Calculate review velocity per listing

   • Count reviews in last 30, 60, and 90 days for each listing
   • Calculate reviews/month average for each window
   • Output: Velocity table (listing x time window)

2. Step 2.2: Identify velocity gap

   • Rank all listings by 90-day velocity
   • Calculate gap between client and top competitor
   • Output: Gap measurement (reviews/month behind)

3. Step 2.3: Calculate catch-up projections

   • Reviews/month needed to reach parity with top competitor in 6, 9, and 12 months
   • Factor in competitor's ongoing velocity (they don't stop)
   • Output: Time-to-parity projections at different velocity targets

Phase 2 Checkpoint:

• Velocity calculated for all listings across 30/60/90 day windows
• Gap quantified with specific catch-up numbers
• Projections account for competitor's ongoing velocity

Phase 3: Response audit

Purpose: Evaluate how well each business responds to reviews and identify gaps.

Steps:

1. Step 3.1: Calculate response metrics

   • Response rate (% of reviews with owner reply)
   • Average response time (if timestamps available, otherwise note as [UNAVAILABLE])
   • Output: Response rate comparison table

2. Step 3.2: Analyze response quality

   • Check for target keyword usage in responses
   • Check for location/neighborhood mentions in responses
   • Evaluate tone: professional, personal, template-feeling, defensive
   • Assess negative review handling: apologetic, defensive, solution-oriented, ignored
   • Output: Response quality assessment per listing

3. Step 3.3: Identify response patterns

   • Does competitor use templates? (look for repeated phrases)
   • Do responses mention specific services or staff?
   • Do responses include calls-to-action (come back, try X)?
   • Output: Response pattern analysis

Phase 3 Checkpoint:

• Response rate and quality compared across all listings
• Keyword usage in responses quantified
• Best practices identified from top-performing competitor

Phase 4: Template generation

Purpose: Create 12 review response templates (4 tiers x 3 variations) with embedded keywords.

Steps:

1. Step 4.1: Define template structure

   • Each template must naturally include:
     • At least 1 target service keyword
     • At least 1 location/neighborhood mention
     • Personal touch (reference specifics from review)
     • Forward-looking statement (invitation to return, try another service)
   • Output: Template structure guidelines

2. Step 4.2: Create 5-star response templates (3 variations)

   • Variation A: Service-focused (highlights the specific service praised)
   • Variation B: Team-focused (credits staff, builds personal connection)
   • Variation C: Community-focused (emphasizes neighborhood/local pride)
   • Output: 3 x 5-star templates

3. Step 4.3: Create 4-star response templates (3 variations)

   • Variation A: Grateful + improvement-curious (asks what would make it 5 stars)
   • Variation B: Service expansion (mentions related services they might enjoy)
   • Variation C: Loyalty-building (offers reason to return)
   • Output: 3 x 4-star templates

4. Step 4.4: Create 3-star response templates (3 variations)

   • Variation A: Empathetic + action-oriented (acknowledge gap, state fix)
   • Variation B: Dialogue-opening (invite offline conversation)
   • Variation C: Improvement commitment (specific steps being taken)
   • Output: 3 x 3-star templates

5. Step 4.5: Create 1-2 star response templates (3 variations)

   • Variation A: Empathetic + escalation (apologize, provide direct contact)
   • Variation B: Fact-based + resolution (address specific issue, state resolution)
   • Variation C: Service recovery (offer to make it right, specific next step)
   • Never: defensive, dismissive, or argumentative tone
   • Output: 3 x 1-2 star templates

Phase 4 Checkpoint:

• 12 templates total (4 tiers x 3 variations)
• Every template includes at least 1 service keyword naturally
• Every template includes at least 1 location mention naturally
• Negative review templates are empathetic, never defensive
• Templates have [BRACKET] placeholders for personalization

Phase 5: Strategy document

Purpose: Create an actionable review growth plan with measurable targets.

Steps:

1. Step 5.1: Set monthly velocity target

   • Based on Phase 2 catch-up projections
   • Recommend realistic velocity (with reasoning)
   • Output: Monthly review target with justification

2. Step 5.2: Map review solicitation touchpoints

   • In-person: after service completion, at checkout
   • Digital: follow-up email/SMS, thank-you page, QR codes
   • Timing: optimal ask window (24-48 hours post-service)
   • Output: Touchpoint map with timing

3. Step 5.3: Define review content guidance

   • What to ask customers to mention: specific service, neighborhood, staff name
   • How to frame the ask (natural, not scripted)
   • Example scripts for staff to use when asking
   • Output: Content guidance with example ask scripts

4. Step 5.4: Create 90-day roadmap

   • Month 1: Set up systems (templates, ask scripts, QR codes)
   • Month 2: Launch review solicitation at all touchpoints
   • Month 3: Measure velocity, adjust approach, respond to all new reviews
   • Output: 90-day action plan

Phase 5 Checkpoint:

• Velocity target set with time-to-parity math
• Touchpoints identified with timing guidance
• Customer ask scripts provided
• 90-day roadmap with measurable milestones

GBP Review Strategy: [Business Name]

GBP URL: [URL] Date assessed: [Date] Competitors analyzed: [Competitor 1], [Competitor 2], [Competitor 3] Assessor: Genesys Growth

Review Velocity Comparison

Gap to #1: [X] reviews/month behind [Competitor Name] Time to parity: ~[X] months at [Y] reviews/month target

Review Content Analysis

Services mentioned in reviews

Neighborhoods mentioned in reviews

Recurring complaints (by theme)

Response Audit

Key gaps:

• [Gap 1]
• [Gap 2]
• [Gap 3]

Review Response Templates

5-star responses

Variation A — Service-focused:

[Template with [SERVICE KEYWORD], [LOCATION], [REVIEWER NAME], [SPECIFIC DETAIL] placeholders]

Variation B — Team-focused:

[Template]

Variation C — Community-focused:

[Template]

4-star responses

Variation A — Grateful + improvement-curious:

[Template]

Variation B — Service expansion:

[Template]

Variation C — Loyalty-building:

[Template]

3-star responses

Variation A — Empathetic + action-oriented:

[Template]

Variation B — Dialogue-opening:

[Template]

Variation C — Improvement commitment:

[Template]

1-2 star responses

Variation A — Empathetic + escalation:

[Template]

Variation B — Fact-based + resolution:

[Template]

Variation C — Service recovery:

[Template]

Review Growth Strategy

Monthly velocity target

• Target: [X] reviews/month
• Current: [X] reviews/month
• Gap: [X] reviews/month
• Time to parity with [Competitor]: ~[X] months

Where to ask for reviews

What to ask customers to mention

• Service: [Specific service they received]
• Location: [Neighborhood or area name]
• Experience: [Specific aspect of service]

Staff ask scripts

Script 1 (in-person, post-service):

"[Natural ask script with specific service + location mention guidance]"

Script 2 (follow-up text/email):

"[Natural digital follow-up script]"

90-day roadmap

Iteration Prompts

1. "Want me to add more competitors to the velocity analysis?"
2. "Should I create review solicitation email/SMS sequences?"
3. "Want me to run the GBP content engine to complement this review strategy?"
4. "Should I export this to Google Docs for the client?"

---

## Anti-Hallucination Guardrails

1. **Only report what was scraped.** If review data can't be extracted for a listing, mark as "[UNAVAILABLE: could not scrape reviews for [listing]]" — don't estimate.
2. **No invented review counts or ratings.** Every number in velocity tables must come from scraped data or user-provided data.
3. **Velocity math must be shown.** Show the calculation behind reviews/month and time-to-parity projections so user can verify.
4. **Response templates are templates, not real responses.** Clearly mark all placeholders with [BRACKETS] and never include real customer names or review content in templates.
5. **Complaint themes must come from actual reviews.** Don't invent complaint categories — only report themes that appear in scraped review text.
6. **Review data is point-in-time.** Note the scrape date on all data tables — review counts change daily.
7. **Time-to-parity is an estimate.** Tag all projections with [ESTIMATED: based on current velocity trends] — competitors can change their velocity too.

---

## Quality Checklist (Pre-Delivery)

### Data quality
- [ ] Review data scraped for client + all competitors (or marked [UNAVAILABLE])
- [ ] Velocity calculated from actual review dates, not estimated
- [ ] All numbers traceable to scraped data
- [ ] Scrape date noted on all data tables

### Analysis quality
- [ ] Velocity comparison table complete (30/60/90 day)
- [ ] Service keyword mentions extracted and compared
- [ ] Neighborhood mentions extracted and compared
- [ ] Recurring complaints grouped by theme with counts
- [ ] Response audit covers rate, time, keywords, tone

### Template quality
- [ ] 12 templates total (4 tiers x 3 variations)
- [ ] Every template includes at least 1 service keyword naturally
- [ ] Every template includes at least 1 location mention naturally
- [ ] Negative review templates are empathetic, never defensive
- [ ] All placeholders clearly marked with [BRACKETS]
- [ ] Templates sound human, not robotic or over-optimized

### Strategy quality
- [ ] Monthly velocity target set with catch-up math
- [ ] Touchpoints identified with timing and method
- [ ] Customer ask scripts provided (in-person + digital)
- [ ] 90-day roadmap with measurable milestones
- [ ] Strategy is realistic for business size and type

---

## Post-Output: Iteration Prompts

After delivering output, proactively offer these iteration options:

### Refinement prompts
1. "Want me to adjust the velocity target based on your capacity?"
2. "Should I customize the templates for a specific service line?"
3. "Want me to add more competitors to the analysis?"

### Expansion prompts
1. "Want me to create review solicitation email/SMS sequences?"
2. "Should I run /gbp-content-engine to build a posts strategy that reinforces review themes?"
3. "Want me to create a review monitoring dashboard spec?"

### Quality prompts
1. "Want me to test the response templates against your actual recent reviews?"
2. "Should I analyze seasonal patterns in your review velocity?"
3. "Want me to identify which competitor response patterns correlate with higher ratings?"

---

## MCP Data Integration

**Level:** 0 — Context (heavy data gathering, competitive analysis)

### Primary tools

| Source | What to pull | Tool | When |
|--------|-------------|------|------|
| **Apify** | Structured Google Maps review data | `call-actor` (Google Maps Reviews actor) | Primary extraction method |
| **Firecrawl** | GBP listing pages (fallback) | `firecrawl_scrape` | When Apify unavailable or for supplementary data |
| **Exa** | Competitor discovery if URLs not provided | `web_search_exa` | When user doesn't have competitor GBP URLs |

### Apify integration

```javascript
// Google Maps Reviews actor — extract last 50 reviews per listing
call-actor({
  actorId: "compass/google-maps-reviews-scraper",
  input: {
    startUrls: ["[GBP_URL]"],
    maxReviews: 50,
    reviewsSort: "newest"
  }
})

What Apify provides:

• Review text, rating, date, reviewer name
• Owner response text and timestamp
• Total review count and average rating
• Structured data ready for analysis

Fallback (no MCP)

• WebFetch for manual GBP page fetching (limited review data)
• User provides review data export (Google Takeout or third-party tool)
• Manual review counting from GBP screenshots
• User provides competitor review counts directly

Final ship gate

Run /premortem --output before ship. See /premortem skill for the 5 execution domains (will-it-resonate / will-it-convert / will-it-stay-on-brand / will-stakeholder-push-back / will-it-degrade-over-time) and output template.

Trivial-case escape: ## Premortem\nNo failure modes — trivial change satisfies the contract for genuinely trivial outputs.

Persuasion & stickiness pass

Output complies with persuasion-and-stickiness.md — Cialdini's 7 persuasion levers + Heath's SUCCESs. Deploy the 1-2 Cialdini levers that fit the reader's barrier (never all seven; every lever must be TRUE), run the SUCCESs diagnostic (Simple / Unexpected / Concrete / Credible / Emotional / Stories) over the near-final draft, then the rule's pre-ship gate.