A productized SEO/GEO unit, built to scale past a two‑person team.

Operating premise

The SEO/GEO unit sells one bundled outcome — rank in search, get cited by AI answer engines, and convert the traffic — delivered by an AI-led production system with senior human curation on top. This page is how the unit runs today, how it scales to many clients without diluting quality, and the funnel + lifecycle engine that keeps high-quality clients flowing in.

What every client buys

One bundled offer, not à-la-carte tasks: a quarterly technical + GEO audit that produces a 30-60-90 roadmap, a steady cadence of articles from the factory (~30/mo), and a standing optimization layer (weekly health checks + a monthly compliance sweep). The audit finds the gaps, the factory closes them, the optimization layer keeps them closed.

Three load-bearing assumptions

Two assumptions this blueprint adds

Scaling past the founding pair surfaces two more, both load-bearing for what follows:

The problem at scale

Today the unit is two people: Rafid — initial audit, roadmaps, strategy, article production, and final approval — and Ian — publishing, technical-SEO fixes, and first-pass QA. On article volume that stretches to roughly 10–20 clients. Past that, three things break, and they break in a specific order.

The single-approver bottleneck

Every article ships through Rafid as the one accountable approver — "per-article human approval is on… Rafid signs off on every article." That is the right call for quality and the wrong thing to scale: it makes the most senior person the rate limiter the moment client count climbs. The fix isn't to remove the gate — it's to staff it.

The three break-signals, in order

Each red line maps to a named hire in §07. The first to fire — the editorial QA queue — is why the first scaling hire is a dedicated content editor, not another generalist.

What we sell

A productized organic-growth engine, not a task list. Four capabilities — all shipped and running today — bundled into scope-based tiers.

The four capabilities

The tier ladder

Tiers are defined by scope and cadence, not price (pricing is open — see §12). Each tier contains the one below it; the ladder doubles as the offer ladder in the funnel (§08).

ICP & positioning

Who the unit sells to, why they buy now, and how to present the package so it lands. Grounded in marketerhire's own ICP data; the SEO/GEO-specific cut is flagged where it's a refinement.

The SEO/GEO buyer

Company size

10–50 employees (sweet spot); paid targeting widens to 10–200.

Revenue

$2–20M.

Marketing budget

$7–10.5K/mo — median won deal ≈ $8,533/mo.

Best-fit verticals

B2B SaaS / Tech — lowest budget-rejection (38.8%), highest LTV, multi-deal expansion. E-commerce / DTC — 25.5% of all contacts, clear organic + content need.

Decision maker

44% founder/CEO, 30% marketing leader (VP / Director / Head of Growth). Signs fast — 88% of won deals close within 30 days.

Why they buy now

The SEO/GEO-specific cut

On top of those, the buyer who needs this unit says: "we publish but nothing ranks," "we're invisible in AI answers," "our content doesn't convert." Those are the exact lines the funnel and lifecycle copy lead with (§08, §09).

Who to walk away from

The qualification gate exists because most inbound isn't a fit. Disqualify: budget under $5K/mo (60% of all leads fail the budget gate), "just exploring / looking for info" (#2 closed-lost reason), no-shows (#1 reason — 62% of losses), and anyone who wants a "CMO" title (0% close rate).

How to present it

Five differentiators

1. SEO + GEO in one engine. Optimized for Google and the answer engines (ChatGPT / Perplexity / Gemini), not just blue links.
2. AI factory + senior human gate. Scorecard-gated, brand-voice-matched, human-approved. Nothing ships without a human checking it.
3. Audit-to-action, not audit-and-leave. The roadmap is executed by a named owner, not handed over as a PDF.
4. Conversion-aware. CRO advisory at the top tier turns traffic into customers — the lesson MH learned when its own session→form rate halved.
5. Compounding intelligence. The factory gets cheaper and better every quarter; every client makes the engine stronger.

The delivery engine

One production system, two lines running through it — content (the article factory) and technical (audit → fixes) — with a human QA layer on top and a conversion layer at the end. This is the real five-stage pipeline the unit runs, generalized from the framework already live on a regulated client (Meribel Health).

The five stages

Inside Stage 2 — the factory

Keyword scoring is a weighted composite — business value 30%, opportunity 25%, difficulty 20%, relevance 15%, content gap 10%. Internal links are verified against the client config, never fabricated. AI-tell removal is mandatory. Each article ships as publication-ready HTML + JSON-LD schema + a scorecard.

The parallel technical line

Alongside content, the audit engine runs a second production line: crawl (up to 500 pages) → a 0–100 health score across 7 weighted categories → a severity-ranked roadmap — Critical: fix now · High: 1 week · Medium: 1 month · Low: backlog. Ian ships the fixes today; at scale a Technical SEO Specialist owns the roadmap queue (§07).

The GEO / citation line

A third workstream runs beside content and technical: audit brand visibility across the major AI engines → build the trust nodes models rely on → rewrite into self-contained, ~134–167-word passages (the citation sweet spot) → monitor citations over time. It rides the same Stage 1–5 flow, owned by Rafid today and folded into the pod at scale. It's grounded in one hard finding — brand mentions predict AI visibility ~3× more than backlinks.

The new conversion layer

Team & org design

Roles, not headcount for its own sake. Each role owns specific stages of the delivery engine and a clear success bar. Two are live today; two more — the editor and the CRO advisor — are what this blueprint adds; the rest fill in as pods multiply.

Live today

What this blueprint adds

As pods multiply

Scaling & capacity

The unit scales by replicating a known unit — a pod — not by piling clients onto the founders. Here's the pod, the per-pod ceiling, and the trigger-driven path from two people at 10–20 clients to many pods at many clients.

The pod

A pod is Managing Editor + SEO Copilot/Pub Ops + Technical SEO Specialist, with a freelance Content Editor on the QA gate. It serves 6–10 clients end-to-end. Two cross-pod specialists — AI Engineer, Ops Manager — compound efficiency across every pod.

The phased path

* Stretched two-person reality. 6–10 per pod is the sustainable, no-dilution number — confirm the live count in §12.

The hire order, and why

1. Editor first. The QA queue saturates first; the editor breaks the single-approver bottleneck and protects quality at the top of the current range — before adding clients.
2. Then the Technical SEO Specialist. Offloads Ian's fix backlog so a full pod reaches the 6–10 client / 240-article steady state.
3. Then CRO + the AI Engineer. CRO makes the top tier convert; the engineer keeps cost/article falling so each new pod starts more efficient than the last.
4. Then a second Managing Editor — the hardest hire. Get it right and pod two takes; Rafid steps off direct pod ops.

The lead funnel

A consistent inflow of high-quality clients comes from one structure, adapted from Matthew Larsen's 1000x Leads Funnel: Traffic → Lead Magnets → Conversion Mechanisms → Offers. It exists to solve the three problems every service business has — not enough leads, can't convert cold traffic, and 99% of leads aren't a fit for the flagship.

Layer 1 — Traffic

The point is compounding: paid warms cold outreach, content makes outbound land, and the unit's own rankings + citations prove the product.

Layer 2 — Lead magnets

Layer 3 — Conversion mechanisms

The magnet's natural next step is an audit-results call — here's your gap, here's the roadmap to close it. Then a qualification gate kills the "99% unqualified" problem (mirroring MH's budget gate: 60% of leads fail it). Survivors go to a strategy call and a tiered proposal. Non-bookers route into the lifecycle campaign (§09).

Layer 4 — The offer ladder

Same ladder as §03, expressed as Matthew's offer tiers — so the funnel captures the whole market, not just the flagship buyer:

The lifecycle campaign

The fastest source of high-quality clients isn't cold — it's marketerhire's own database. This is the campaign to sell the SEO/GEO package into it, cloning the multi-segment template MH already ran (4 upsell + 2 re-engagement segments, all Day 0 / 3 / 7) plus an active-customer expansion sequence, re-pointed at the new offer. Generated to the real send rules; nothing here ships to HubSpot without sign-off.

Segments (real counts, re-cut for SEO/GEO fit)

The six sequence types

Sample copy · Sequence A — Hot, exec plain-text

To past customers $10K+ (333), from Raaja. ≤150 words, lowercase subject, one CTA, no fabricated results.

Sample copy · Sequence B — Warm nurture, branded

To past QTBs $5K+ (3,508), from "MarketerHire Team". VALUE → PROOF → ASK.

Sample copy · Expansion — active customers

Send guardrails (non-negotiable)

Roadmap & phasing

Crawl, walk, run — each phase gated by a capacity trigger from §07 and paired with the GTM move that fills the next pod.

Metrics & KPIs

What we watch, by layer. Quality and throughput KPIs are the gameplan's, verbatim; the funnel and lifecycle targets are new and flagged.

Capacity triggers

When any one fires, hire or spin a pod: client count ≥ 8 · roadmap lead time > 2 wks · editorial QA backlog > 2 days · technical backlog > 30 items · Rafid IC work > 50% (target < 20% steady at scale).

Funnel & lifecycle · new

* Real prior-campaign benchmark (Proactive Match, 2023): 12.1K sends · 48% open · 65 QTBs · 7 contracts. SEO/GEO-campaign targets to be set.

Proposed vs. confirmed

Everything tagged proposed or confirm on this page, in one place. Grounded facts trace to cited sources; these are the decisions to lock before this goes external or drives hiring.

Decisions to lock

Source-data notes

Two figures conflict across MH's own sources — flagged so no one treats a single number as settled: single-deal company count (3,661 in revenue-model vs 4,547 in lever-priorities) and the expansion opportunity ($5.5M at 10% conversion vs $3–5M at 5%). Use the lever-priorities figures (more recent live pull) where one is needed.