Five tracks, reverse-engineered from real companies’ pricing, business models, and delivery mechanics — not generic theory. Use this as the structure to react to, not build from scratch: what’s missing, wrong, or dated?
Everyone reaches working level across all five tracks in weeks 1–6, then specialises in one track and deploys on a real client in weeks 7–12 (the pod model). Track E (AI Agents) is foundational and cross-cutting — it’s the AI-native delivery layer under everything else.
Weeks 1–3
Foundations, shared: AI-agent/automation literacy (E), data & Clay basics (C), offer + copy fundamentals (D).
Weeks 4–6
Breadth pass: working level in Get Found (A), Convert (B), and the full Outreach motion (D).
Weeks 7–12
Specialise + deploy: pick one track, go deep, ship the track’s capstone for a real client.
Track A
Take a brand from “invisible” to found on Google and cited by ChatGPT/Perplexity/Gemini, and prove it with the industry-standard metrics.
| Module | Topics | Hrs | Why it matters |
|---|---|---|---|
| A1 · SEO foundations | Search intent, keyword research, on-page, technical crawlability, Core Web Vitals, AI-crawler log analysis | 22 | AI search still pulls from the ranked, crawlable web — no base, no citations. |
| A2 · Content that ranks & gets cited | Answer-first writing, chunking, E-E-A-T, gap-analysis against live LLM outputs | 15 | The daily deliverable + #1 lever on both rankings and AI citations. |
| A3 · AEO/GEO technical | Schema, entity building, G2/Capterra optimisation, prompt-panel construction, build-your-own citation-tracker lab | 20 | Makes content machine-extractable and citable. |
| A4 · Authority — mentions > links | Digital PR, brand mentions, review-platform optimisation | 10 | Unlinked brand mentions now out-predict backlinks for AI citations. |
| A5 · Measurement & reporting | Citation rate, sentiment, prominence, share of voice; AI-referral segmentation | 10 | Proof of ROI is what gets you renewed. |
| A6 · Delivery & packaging | The audit + roadmap as a sellable service; packaging AI-visibility as a retainer add-on | 15 | Turns skill into a priced offer that matches how the market actually buys. |
Tools & lean stack
Ahrefs or Semrush, GSC + GA4, Screaming Frog, an AI chatbot, a citation tracker. Lean ≈ $180–250/mo.
Capstone
“AI-Visibility Turnaround” for a real business — baseline report vs 2 competitors → SEO+GEO roadmap → 3 optimised pages live → 30-day before/after report + retainer pitch.
Best guides
Aleyda Solis learningaisearch.com · iPullRank AI Search Manual · Ahrefs GEO guide.
Track B
Build a fast, on-brand landing page/funnel and run a statistically valid experiment on it — sell “more revenue per visitor,” not “a website.”
| Module | Topics | Hrs | Why it matters |
|---|---|---|---|
| B1 · Conversion foundations | Buyer psychology, conversion math, sample-size/statistical-power calculation, minimum-traffic thresholds for valid testing | 12 | The single most-cited differentiator between amateur and CXL-grade CRO work. |
| B2 · Conversion copywriting | Hero/value-prop/objection-handling, the LIFT Model, tied to VoC research | 12 | Copy, not design, is the #1 conversion lever. |
| B3 · Build fast (no-code + AI) | AI-native prompt-to-code (v0, Lovable) vs AI-assisted no-code (Framer, Webflow) — taught as two different jobs | 15 | The tangible thing clients pay for, at 5× speed. |
| B4 · Instrumentation & analytics | GA4, Microsoft Clarity, dedicated tracking-QA unit | 14 | Broken instrumentation quietly invalidates everything downstream. |
| B5 · Research & experimentation | Heatmaps, VoC, A/B & sequential testing, the “peeking problem,” realistic win-rate expectations | 22 | The actual moat in this track — not page-building speed. |
| B6 · Packaging & reporting | Scope, price, retainer, lift-in-dollars reporting; building a wins-database from day one | 15 | Converts skill into income without selling invalid tests to under-scale clients. |
Tools & lean stack
Figma, Framer or Webflow, v0 or Lovable, GA4, Microsoft Clarity. Lean ≈ $45–55/mo build-phase; +$300–600/mo once formal A/B testing is warranted.
Capstone
Take a real page from baseline to proven lift — research brief with a justified sample size → rebuilt instrumented page (QA'd) → one live A/B test (2–4 wks) → 1-page report with $ lift + next 3 hypotheses.
Best guides
Julian Shapiro Landing Page Guide · Demand Curve · CXL Conversion Optimization · GoodUI · WiderFunnel LIFT Model.
Track C
Build an enriched, verified, signal-tagged lead system in Clay and hand it to outreach/CRM — fastest track to deliverable-ready.
| Module | Topics | Hrs | Why it matters |
|---|---|---|---|
| C1 · GTM data foundations & ICP | Firmographics/technographics, persona/title mapping, provider price-ladder exercise | 8 | A perfect list of the wrong people is worthless. |
| C2 · Clay fundamentals & unit economics | Tables, enrichments, HTTP/API columns, Clay's dual-credit model, cost-per-verified-record modelling | 10 | Failed lookups still burn credits — unit economics is a real margin question. |
| C3 · Sourcing & list building | Apollo, Sales Nav scraping, dedup + normalisation, India company-registry & workforce data (MCA/GST/director data) | 10 | Everything downstream is only as good as the raw, clean list. |
| C4 · Waterfall enrichment & deliverability | Multi-provider waterfalls, spend caps + early-abort logic, verification as a mandatory final step | 14 | The signature, hardest-to-fake skill. |
| C5 · AI enrichment & scoring | Claygent/LLM columns, fit-scoring, spot-check QA/hallucination-checking methodology | 12 | Data-quality trust is the difference between keeping a retainer and churning in month two. |
| C6 · Signals & activation | RB2B, Trigify, UserGems, 6sense as a real pricing ladder; a tiered offer mapped to it; CRM sync | 12 | Turns a static list into prioritised, timed pipeline. |
| C7 · Service delivery & partner certification | Retainer/SOW structuring, account management, Clay's Solutions Partner ladder | 8 | Service-delivery discipline, not more tool depth, separates a $3k/mo vendor from a $15k/mo one. |
Tools & lean stack
Clay, Apollo, a verifier (DeBounce/NeverBounce), Sales Nav; signals tier RB2B/Trigify. Lean ≈ $400–700/mo including real credit spend at volume.
Capstone
Live Clay workspace + delivered list — ICP spec → 500–1,000 sourced/normalised → multi-layer waterfall with spend caps → verification (<3% bounce) → AI fit-score with QA spot-check → one live signal workflow → CRM sync → 1-page readout framed as a retainer pitch.
Best guides
Clay University Clay 101 · Clay Waterfall guide + Templates · Clay Solutions Partner directory.
Track D
Stand up cold-email infra that lands in the inbox, run AI-personalised multichannel sequences, and book meetings — with realistic timelines.
| Module | Topics | Hrs | Why it matters |
|---|---|---|---|
| D1 · Foundations & sender rulebook | SPF + DKIM + DMARC, <0.3% spam, one-click unsub | 5 | Ignore it → lose 30–50% deliverability overnight. |
| D2 · Domains & architecture + infra sizing | Secondary domains, a numeric infra-sizing lab (domains needed for a target send volume) | 10 | Turns “unglamorous infra” into a gradeable artifact instead of folklore. |
| D3 · DNS authentication | SPF, DKIM signing, DMARC, PTR, custom tracking domain | 10 | No aligned auth = straight to spam. |
| D4 · Mailboxes, warm-up & true cost modelling | Provider cost spread, warm-up ramp, true all-in cost vs sticker price | 10 | Sticker prices in this category are consistently misleading. |
| D5 · Scale, rotation, monitoring & incident response | Per-mailbox caps, blacklist detection-and-remediation drill | 14 | A blacklisted domain is dead for a full warm-up cycle again. |
| D6 · ICP, offer & outbound math | ICP + no-go filters, quantified value prop, true cost-stack modelling | 8 | No copy or tool saves a wrong list or weak offer. |
| D7 · Cold-email copywriting | Josh Braun 4-T + 15 principles, one-idea/one-ask, binary CTAs | 15 | Message quality turns a delivered email into a reply. |
| D8 · AI personalisation at scale | Claygent/LLM columns, tiering, quality-gating, consumption-based pricing literacy | 17 | Cost discipline separates a profitable retainer from a margin-losing one. |
| D9 · Multichannel, reply handling & the AI SDR trust gap | Email+LinkedIn sequences, human-in-the-loop design, platform-ToS/compliance | 22 | Operators who understand the trust gap can sell against black-box competitors credibly. |
Tools & lean stack
Clay, Instantly or Smartlead (all-in cost, not sticker price), domains + inboxes, MailReach/Warmy, Lavender, a LinkedIn tool. Lean ≈ $300–500/mo.
Capstone
Live campaign for one narrow ICP — infra-sizing calc → 250–500 enriched/validated leads → ≥2 warmed domains → live 5–7 touch sequence → results dashboard with ≥1–3 booked meetings + true cost-per-meeting.
Best guides
ColdIQ Scalable Outbound System · Instantly/Smartlead deliverability guides · Josh Braun 15-principles.
Track E
Wire LLMs + tools into a reliable, monitored automation for a marketing use case — the AI-native delivery layer under every other track.
| Module | Topics | Hrs | Why it matters |
|---|---|---|---|
| E1 · How LLMs & automation think | Tokens/context, structured JSON output, AI vs plain logic | 10 | Can't debug an agent you don't understand. |
| E2 · Prompt engineering for reliability | System vs user prompts, output schemas, guardrails, escalate-to-human | 12 | The prompt is the “code” of your agent. |
| E3 · No-code orchestration (n8n-primary) | n8n as the primary agent-building platform; Make as a single client-adapter lesson | 20 | What you'll ship 80% of the time. |
| E4 · APIs, webhooks & data plumbing | REST/JSON, webhooks, HTTP nodes, MCP intro | 15 | Agents are only as good as the tools they can reach. |
| E4.5 · Deliverability & compliance infrastructure | Domain warming, SPF/DKIM/DMARC, CAN-SPAM/GDPR basics | 8 | Load-bearing plumbing for anything the agents in E3–E5 actually send. |
| E5 · Enrichment & agentic patterns | Clay waterfalls as the mandatory core; CrewAI/LangGraph as the premium-tier path | 15 | Honestly weighted toward what the market pays for today. |
| E6 · Testing, evals & monitoring | Operational reliability for GTM pipelines + LLM evals/observability (Langfuse) | 16 | Reliability is the product — demoware doesn't earn retainers. |
Tools & lean stack
n8n, OpenAI/Anthropic API, Clay, a sequencer + LinkedIn tool, Sheets/Slack, Langfuse. Lean ≈ $250–450/mo. No-code first.
Capstone
“Inbound Lead Qualifier & Router” — form/CRM trigger → Clay/Apollo enrich → LLM scores fit → hot→Slack + auto-draft reply; cold→nurture; 20-record test set, human-review fallback, Langfuse run log + before/after writeup.
Best guides
Nate Herk n8n masterclass · Clay Guide to GTM Engineering · Anthropic MCP docs · Langfuse docs.