Executing Skill
Initialising context layer...
⚡ Programme Overview

Blue Array — From SEO Consulgency®
to AI-Native Operation

A 16-week phased transformation covering SEO, GEO, Digital PR, Technical SEO, Content, CRO, ASO and all client delivery workflows. Built on Claude CoWork Pro + n8n + Google Workspace — the stack Blue Array has already sanctioned.

Duration
16
weeks + ongoing
Phase 0
Discovery & Audit
Weeks 1–2
Phase 1
Foundation & Knowledge Layer
Weeks 3–4
Phase 2
12–20 Skills Built & Tested
Weeks 5–8
Phase 3
Company-Wide Rollout
Weeks 9–12
Phase 4
Integration & Automation
Weeks 13–16+
Technology Stack — What Blue Array Has Sanctioned
AI Execution — Team
Claude CoWork Pro
Primary AI for all SEO managers, analysts, PR team, customer success, sales. Already rolled out. Skills live here.
AI Execution — Technical
Claude Code
Technical SEO specialists + dev-adjacent tasks. Screaming Frog scripting, log file parsing, schema generation.
Primary LLM
Claude (Anthropic)
Main execution model. Best wrapper tooling for agency workflows. AUP v2.0 compliant.
Advanced Automation
n8n
Complex multi-step workflows. API integrations. Scheduled tasks. Tier 1–2 automation orchestration in Phase 4.
Knowledge & Orchestration
Notion (TBD)
System of record. Skills Library. Client Hub. Service Playbooks. Platform confirmed collaboratively in Phase 0.
SEO Data Layer
Ahrefs + Semrush + GSC
API exports feeding into Skills. Screaming Frog + Sitebulb for crawl data. GA4 + Looker Studio for reporting.
Communication
Slack
Team + client comms. Integrated with n8n for automated alerts, Tier 1–2 routing, and Skills Leaderboard.
File Storage
Google Drive
Client deliverables. Auto-save protocol for CoWork outputs. Integrated via Google SSO (Blue Array account).
Reporting
Looker Studio
Client-facing dashboards. Automated data feeds from GSC/GA4. Skill outputs formatted for direct Looker import.
Core Framework
Inspiration (Human)
What consultants focus on
Client strategy sessions · Campaign concepts · North Star Goals® setting · Pitch ideation · Relationship building · Quality assurance · Skill improvements
Cannot be automated
Perspiration (AI)
What Skills handle
Crawl data processing · Keyword research compilation · Report formatting · Backlink analysis · Status updates · Meeting notes · Content brief structuring
Captured in a Skill
The AUP Rule that sits above all others: A consultant carefully reviews, refines, and signs off on any AI-assisted output before it reaches a client. No exceptions. Every Skill in this programme produces a first draft — never a final deliverable.
Phase 0

Discovery & Audit

Before building anything: interview every team member, audit existing AI work, map friction points, and score every task against the perspiration matrix. In person at espacio_RES with Simon and the team.

Duration
1–2
weeks
Starting point, not zero: Blue Array already has internal Skills and tools in development, and the automation survey is in progress. Phase 0 audits what exists first — we build on it, not over it.
Structured Interview Programme — by Department
🔧 Technical SEO
How long does a full crawl analysis take? What tool — Screaming Frog / Sitebulb?
How are issues categorised and prioritised after the crawl?
How is the dashboard built? Looker Studio template or manual?
How long from GSC export to client-ready migration pre-check?
🤖 GEO Team
How do you currently track brand citations across ChatGPT / Gemini / Perplexity?
What existing internal tools or third-party tools are already in use?
How frequently are LLM checks run? What triggers a manual check?
How is competitive AI share of voice tracked today?
✍️ Content Team
How long from keyword research export (Ahrefs / Semrush) to published brief?
How many revision cycles does a brief typically go through?
How are content calendars populated and maintained?
What does a Blue Array A-grade brief look like vs a weak one?
📢 Digital PR
How are prospect lists built? Which tools — Cision / MuckRack / manual?
How long does campaign ideation take per client?
How is outreach personalised at scale today?
How is backlink profile analysis structured in reports?
👥 Customer Success
How much time goes into a monthly report — data pull to PDF?
How is client meeting prep done? What gets assembled and when?
How are action items tracked post-meeting?
Where does time get lost in the client comms cycle?
💼 Sales & Marketing
How long from inbound enquiry to pitch deck delivery?
What research goes into a prospect before a proposal is written?
How is ABM structured today — who manages the list?
How is the weekly newsletter currently curated and written?
Audit Scoring Matrix
FieldDescriptionExample
Task NameSpecific, measurable task"Monthly technical audit report — RAC"
Service LineWhich department owns itTechnical SEO
FrequencyHow often it recursMonthly × 12 clients
Time / OccurrenceHours end-to-end3.5 hours
Monthly Hours (All Staff)Freq × Time × Headcount3.5h × 12 × 2 analysts = 84h/month
Perspiration %Rule-based vs strategic85% perspiration
Tools InvolvedCurrent data sourcesScreaming Frog, Ahrefs, GSC, GA4, Looker Studio
Automation Score1–5 (5 = fully automatable)4 — structured data, repeatable output
Priority ScoreMonthly Hours × Auto Score × Impact84 × 4 × 3 = 1,008
Phase 0 Deliverables
Deliverable 1
Friction Map
Every team member interviewed. All tasks scored with Priority Scores. Sorted by highest-impact automation candidates.
Deliverable 2
Inspiration / Perspiration Baseline
Actual ratio measured per department. Target ratios set collaboratively with Simon. Business case built on real figures only.
Deliverable 3
Prioritised Skills Backlog (Top 20)
Ranked list validated with leadership. Existing AI work catalogued. Quick wins confirmed for Phase 2 start.
Phase 1

Foundation — Data & Knowledge Layer

AI has nothing to work with if Blue Array's methodology lives in people's heads and scattered Google Drives. This phase creates the structured context layer that makes Blue Array's AI different from a generic ChatGPT prompt.

Duration
3–4
weeks
Single System of Record — Platform TBD (Notion / Asana)
LayerContentsAccess
Company Core Blue Array methodology, Consulgency® approach, brand voice guide, B Corp policies, AUP v2.0 All staff
Client Hub Per-client spaces: briefs, North Star Goals®, history, key contacts, performance data Assigned teams
Service Playbooks Technical SEO checklist, GEO methodology, Content audit framework, Digital PR process Dept + cross-team
Skills Library All AI Skills with instructions, owners, changelog, categorised by service line All read · Trio edit
Quality Standards Examples of A-grade deliverables, common mistakes, review criteria per service line All staff
Governance Audit logs, Skill change requests, Skills Leaderboard, version history Leadership + Trio
Context Layer — What Makes Blue Array's AI Different
SEO Methodology
Blue Array Technical Framework
How Blue Array runs technical audits. The exact crawl analysis framework. The prioritisation scoring system. Fed into every Technical SEO Skill.
GEO Framework
LLM Citation Methodology
AI citation tracking methodology. RAG optimisation approach. LLM visibility scoring. Integrated with existing GEO internal tools.
Client Context
North Star Goals® per Client
Current campaigns, historical performance, stakeholder preferences. Loaded per-client into Skills — no cross-contamination between accounts.
Voice Guide
How Blue Array Communicates
Confident but not arrogant. Specialist language. No jargon in client-facing materials. Encoded in every Skill's output format rules.
Tool Integration Notes
Data Flow Map
How data moves between Screaming Frog → Ahrefs → GSC → GA4 → Looker Studio. Skills know which export to expect and how to parse it.
Quality Benchmarks
A-Grade Deliverable Examples
Anonymised examples of top-tier work from past client engagements across each service line. Skill output is scored against these.
Context Window Strategy: Claude CoWork loads only Skill names and short descriptions (lightweight index). When given a task, it identifies the relevant Skill(s), then loads only those full instructions. Blue Array can have 100+ Skills with zero performance degradation.
Phase 2

12–20 Skills Built & Tested on Real Data

Each Skill is reviewed and approved by the Skills Owner. Output tested on real client data before sign-off. No speculative ROI — time savings measured against Phase 0 baseline.

Duration
5–8
weeks
Live Skill Demos — Click a Skill to Run It
Skills Library
Content Brief Generator
CONTENT
✓ Live
GEO Citation Tracker
GEO
✓ Live
Technical SEO Audit
TECH SEO
✓ Live
Monthly Client Report
CUSTOMER SUCCESS
✓ Live
Digital PR Prospect Research
DIGITAL PR
✓ Live
Proposal / Pitch Builder
SALES
In Review
Meeting Notes → Actions
CROSS-SERVICE
✓ Live
Content Brief Generator
CONTENT Owner: Head of Content
Skills Approval Status Board
All 13 Skills — Review & Sign-off Tracking
9 Live · 3 In Review · 1 Queued
Phase 3

Training & Company-Wide Rollout

All staff trained across 3–4 cohorts. Trio governance active. Skills improving from real-world feedback. Target: >75% weekly active users by Week 12.

Duration
9–12
weeks
120-Minute Training Session Plan
TimeBlockActivityOutcome
0–20 minVisionThe Consulgency® AI advantage. Inspiration vs Perspiration. Why Blue Array's context layer beats generic ChatGPT.Shared mental model
20–50 minTechnicalClaude CoWork walkthrough. Notion Skills Library. Context window strategy. AUP v2.0 rules built in.Understanding the full stack
50–70 minLive BuildEach person builds a real Skill for a task they do every week. Starts simple (e.g. client email draft), then a team Skill.First working Skill in their hands
70–90 minGroup WorkTeams identify their top 3 repetitive tasks and build the Skill together.Service-line-specific Skill, owned by the team
90–110 minConfigPersonal instructions, CoWork preferences, Google Drive auto-save setup.Personalised Claude environment
110–120 minSafetyPII rules, B Corp data standards, AUP v2.0 review. What never goes into a Skill input.Safe, compliant usage habits
Training by Department — Focus Skills
Technical SEO
Technical Audit Report · Migration Pre-Check · Core Web Vitals
Every technical SEO analyst can run audit Skills independently from Day 1. Screaming Frog exports become Skill inputs.
GEO Team
GEO Citation Tracker · AI Visibility Report · LLM Monitoring
GEO team has an automated citation monitoring pipeline. Manual LLM checks replaced with structured Skill output.
Content
Content Brief Generator · Keyword Clustering · Content Calendar
Brief-to-calendar pipeline fully operational. Ahrefs keyword exports become structured content briefs in minutes.
Digital PR
Prospect Research · Pitch Drafting · Backlink Profile Analysis
Outreach prep time cut significantly. Personalised pitch angles generated from topic + industry inputs.
Customer Success
Monthly Client Report · Meeting Prep Pack · Action Item Extraction
Report generation time cut dramatically. GSC/GA4 data pulls formatted directly into North Star Goals® templates.
Sales & Marketing
Proposal Builder · Newsletter Curation · Lead Research
Faster proposal turnaround. Consistent newsletter voice. ABM opportunities flagged automatically from meeting transcripts.
Trio Governance Structure
Trio 1
SEO Delivery
Domain Expert: Senior SEO Manager / Technical SEO Lead
Scope: Technical SEO, Content, International SEO, ASO Skills
Tools: Screaming Frog, Ahrefs, GSC, Claude CoWork
Trio 2
GEO + Digital PR
Domain Expert: Head of GEO / Head of Digital PR
Scope: GEO, Digital PR, ORM, Short-Form Video Skills
Tools: Perplexity, ChatGPT, Gemini monitoring, Semrush
Trio 3
Client Ops & Internal
Domain Expert: Jenny Ovens / Head of Customer Success
Scope: Client reporting, sales, HR, internal admin Skills
Tools: Looker Studio, Google Drive, Slack, Notion
Change Management: Early adopters identified in Phase 0. Champions demonstrate real time savings in team meetings — peer proof over management mandates. Skill Change Requests logged via CoWork. All updates versioned. Skills not updated in 30+ days trigger an alert to the Skill Owner.
Phase 4

Integration, Automation & Scaling

Work flows through the system with minimal manual routing. Tier 1–2 automations live via n8n. Tier 3 Skills producing client-ready output queued for human review. Blue Array operates as an AI-native agency.

Duration
13–16+
weeks
Four-Tier Task Architecture
T1
Admin
Basic admin tasks — fully automated, no human required. Triggered by events in Slack, Notion, or email.
⚙️ n8n
📋 Notion
💬 Slack
📅 Google Calendar
Update client status on task board
Duplicate monthly report template
Create new client Slack channel
Archive completed campaign folder
Send automated weekly digest
T2
Integrations
External API integrations — n8n workers fetch and push data across Blue Array's SEO tool stack. Runs on schedule or trigger.
⚙️ n8n
📊 Ahrefs API
🔍 GSC API
📈 Semrush API
📉 Looker Studio
📁 Google Drive
Pull monthly GSC performance data → format for report Skill
Fetch Ahrefs backlink changes → flag new links to account manager
Sync keyword rankings to Looker Studio dashboard
Pull Screaming Frog crawl export → queue for Audit Skill
Auto-send calendar invite + agenda draft post-meeting booking
T3
Skills
Complex knowledge work — Claude CoWork + Blue Array Skill. Produces client-ready first drafts. Queued for human review and sign-off. Never delivered without consultant approval.
✦ Claude CoWork Pro
🤖 Claude API
🕷️ Screaming Frog
📊 Ahrefs
🔍 GSC
📋 Notion (Skills Library)
Generate monthly client report → queued for AM review
Run competitive SEO analysis → queued for strategist
Draft content briefs from keyword export → queued for Head of Content
Run GEO citation report → queued for GEO lead
Build proposal sections → queued for sales team
T4
Strategic
Strategic & novel work — Claude CoWork + Human collaboration. The consultant drives. AI is the thinking partner and research engine. Highest value output.
✦ Claude CoWork Pro
🤖 Claude (Anthropic)
Build new client North Star Goals® strategy
Design site migration architecture
Creative Digital PR campaign ideation
GEO methodology development for new vertical
New business pitch strategy for enterprise prospect
Live Automation Flows — Tier 1 & 2 (n8n)
n8n Workflow Scenarios — Running on Blue Array's Stack
Tier 1 & 2 Automations
Trigger
New client signed (Notion)
n8n
Create Slack channel #client-name
n8n
Duplicate Notion client hub template
n8n
Create Drive folder structure
Output
Kick-off pack ready in Notion
TriggerNotion database row status changes to "Signed" in CRM space
Slack (Tier 1)Channel created with AM, Head of SEO, and client team auto-invited
Notion (Tier 1)Client Hub page duplicated from master template, client name auto-populated
Google Drive (Tier 2)Folder hierarchy created: /Clients/[Name]/Briefs, /Reports, /Assets, /Strategy
Result — Zero manual setup
New client is fully onboarded in under 2 minutes from Notion status update. Account Manager's first task is the strategy call — not folder creation. All triggered by a single status change.
Trigger
1st of month (cron)
n8n + GSC API
Pull 28-day performance data
n8n + Ahrefs API
Fetch ranking changes + new links
Claude CoWork (T3)
Monthly Report Skill
Output
Draft report → AM review queue
GSC API (Tier 2)Clicks, impressions, CTR, avg. position pulled for all tracked URLs per client
Ahrefs API (Tier 2)New backlinks, lost links, DR changes, keyword ranking movements
Semrush API (Tier 2)Competitor visibility trends, SERP feature ownership changes
CoWork Skill (Tier 3)Data formatted into North Star Goals® template, narrative insights generated — queued for AM
Result — Report arrives in AM's queue by 9am on the 1st
AM reviews, adjusts narrative, adds strategic commentary, and signs off. Data pull + first-draft narrative: fully automated. Strategic interpretation and client relationship: fully human.
Trigger
Meeting ends (Calendar)
n8n + Transcription
Auto-transcribe recording
Claude CoWork (T3)
Meeting Notes → Actions Skill
n8n
Create tasks in Notion + Slack summary
Output
AM reviews action items
Transcription (Tier 2)Google Meet / Zoom recording auto-ingested. Transcript sent to CoWork Skill.
CoWork Skill (Tier 3)Extracts: action items with owners + deadlines, decisions made, upselling opportunities flagged, ABM signals
Notion (Tier 1)Action items created as tasks on client board with assignees auto-populated
Slack (Tier 1)Meeting summary posted to #client-name with action items and owners listed
Result — Meeting notes done before the AM closes their laptop
Skill also flags missed upselling opportunities and ABM signals from the conversation. AM reviews, adjusts owners if needed, hits send. Meeting admin: zero manual effort.
Trigger
Ahrefs / GSC webhook
n8n
Detect >5 position drop on tracked KW
Claude CoWork (T3)
Competitive SEO Analysis Skill
n8n
Post alert to Slack #client-name
Output
SEO Manager notified with context
Detection (Tier 2)Ahrefs API polled daily. GSC Search Analytics compared week-on-week. Threshold: >5 position drop on top-20 keywords.
Competitive Skill (Tier 3)Analyses who gained the position, their content changes, backlink changes — produces 3-bullet diagnosis
Slack Alert (Tier 1)Alert posted to client channel with: keyword, drop magnitude, competitor who gained, and Skill diagnosis
Human (Tier 4)SEO Manager reads context, decides response — brief, content update, or escalate to Head of SEO
Result — Proactive alert with context, not just a number
Instead of SEO Manager discovering a drop in Monday's manual check, they receive a Slack alert with a diagnosis before the client notices. Response time: hours not days.
Trigger
Daily cron (n8n)
n8n + API
Query ChatGPT / Gemini / Perplexity
Claude CoWork (T3)
GEO Citation Tracker Skill
n8n
Update GEO dashboard in Notion
Output
GEO lead reviews trends
Query Layer (Tier 2)n8n sends predefined query set to ChatGPT API, Gemini API, Perplexity API. Responses stored per run.
Citation Skill (Tier 3)Analyses responses for: client brand cited (Y/N), citation quality, competitor citations, LLM confidence signals
Notion Dashboard (Tier 1)Citation rate, competitor share of voice, trending queries — all updated automatically per client
Integrate Existing (Tier 2)Must integrate with Blue Array's existing internal GEO tools and third-party monitoring — not replace them
Result — Manual → Daily automated GEO monitoring at scale
Blue Array's GEO team moves from periodic manual spot-checks to daily automated coverage across all LLMs for all GEO clients. The flywheel: more data → better RAG optimisation → better client results → more GEO clients.
Full Client Delivery Pipeline — Target State
End-to-end automated client delivery flow Tiers 1–4 working together
1
Client meeting recorded & auto-transcribed
Google Meet / Zoom recording processed automatically on meeting end. Transcript sent to CoWork via n8n webhook.
Google Meet n8n webhook Tier 1
2
Transcript → structured action items + ABM flags
Meeting Notes → Actions Skill runs. Extracts: action items with owners + deadlines, decisions, upselling opportunities, ABM signals. Posted to Notion client board and Slack channel.
Claude CoWork Notion Slack Tier 3
3
Task router assigns each item to correct tier
n8n reads action items from Notion. Classifies each task by type. Routes: admin tasks → Tier 1 automation, data pulls → Tier 2 workers, knowledge tasks → Tier 3 Skill queue, strategy → Tier 4 CoWork session.
n8n router Notion Tier 1 Tier 2
4
Tier 1–2 execute automatically
Slack channels created, calendar invites sent, Google Drive folders organised, data pulled from GSC/Ahrefs/Semrush APIs, Looker dashboards updated. All without human input.
n8n Google APIs Ahrefs API GSC API Tier 1 Tier 2
5
Tier 3 Skills run — output queued for human review
Client report drafted, competitive analysis generated, content briefs built, GEO citation report compiled. Each output lands in the consultant's review queue in Notion. Nothing reaches the client without sign-off.
Claude CoWork Pro Review Queue Tier 3
6
Tier 4 — CoWork session for strategic work
Assigned strategist opens a CoWork session for complex items: new client strategy, migration architecture, GEO methodology. Claude is the thinking partner — consultant drives, signs off, delivers.
Claude CoWork Pro Tier 4 — Human led
7
All statuses auto-update — client-facing PM visible in real time
Notion board updates automatically as tasks complete. Client-facing project management reflects real progress. AM's bandwidth freed from status chasing — focused on relationship and strategy.
Notion Slack alerts Looker Studio Tier 1
Key constraint: Every Skill output is client-facing quality but never auto-delivered. The n8n router queues Tier 3 outputs for human review in Notion. The consultant reads, refines, and sends. This is how AUP v2.0 compliance is built into the architecture — not bolted on afterwards.
Measurement Framework — Post-Phase 4
Leading Indicators (Weeks 1–12)
MetricTarget
Team interviewed100%
Skills built & live12–20
Training completion100% staff
Weekly active users>75%
Skill executions / week40+
Tier 1–2 automations liveAll 5 scenarios
Lagging Indicators (Months 3–6+)
MetricMethod
Hours saved / person / monthPer-Skill time vs Phase 0 baseline
Client report turnaroundPM data — time from data pull to client send
Revenue per employeeFinancial — quarterly comparison
Client NPS / CSATSurvey — maintained or improved
Skill library growth40+ active Skills by Month 6
Staff satisfaction with AI toolsInternal survey — target >80% positive
Live Demo | Confidential — Blue Array Ltd | Powered by Geniia.ai | Built by Generative Z | April 2026