~/case-study · greenfield build no. 3

Zero to $145K ARR in nine months. One engineer. One AI system.

Client: Omnibound AI · AI search marketing platform · B2B scale-up · Role: AI GTM Engineer (Oct 2025 - present)

< 3 daysEngine deploy time
$145KARR closed (5 logos)
$897K+Qualified + active pipeline
1.5%+Positive reply rate
4Positioning shifts shipped

The situation

Omnibound sells AI search marketing - helping brands win visibility inside AI answers, a category that barely existed two years ago. New category, no inherited GTM infrastructure, no playbook to copy, and scale-up expectations on pipeline. The classic answer is hire a marketer, two SDRs, a RevOps contractor, and wait two quarters for the machine to assemble itself.

That answer costs roughly $400K+ a year in loaded headcount before the first meeting is booked. The mandate here was the opposite: one GTM engineer, an AI system, and revenue accountability from day one.

The build

PHASE 01
Strategy

Strategy compiles first

Before any tooling: TAM sized, ICP defined, personas mapped, and 1P/2P/3P signal sources selected - which job-posting patterns, stack signatures, website visitors, and LinkedIn behaviors actually indicate buying intent for AI search marketing. The engine was designed on paper before a single API key existed.

PHASE 02
Engine

Full stack live in under 3 days

The end-to-end chain below - previously a 2-week deploy - was engineered, tested, and sending in under 3 days. 150+ domains and 300+ mailboxes warmed and orchestrated, deliverability engineered before volume.

PHASE 03
Workflows

Unattended workflows layered on top

The production workflows: a 7-stage LLM copy engine (zero human review), agentic research (6-hour batches to under 30 minutes), 500+ personalized ABM landing pages tracking Share of Answer and AI Share of Voice, RB2B visitor de-anonymization, and the full event + webinar engine - every one wired into CRM instrumentation.

PHASE 04
Revenue

The loop closes on ARR, not activity

Every play attributes to closed-won through a self-built HubSpot layer. BANT qualification, deal-stage hygiene, and Slack alerts keep human attention only where judgment matters: the conversations and the closes. Buyer language from those calls feeds back into positioning - 4 narrative shifts validated in-market so far.

signals trigify + phantombuster + firecrawl intel claude code (analysis · scoring · copy) orchestrate n8n -> supabase enrich clay (5-source waterfall) activate smartlead + heyreach + ABM pages revenue hubspot -> slack (attributed to the dollar)

The results (9 months in)

MetricOutcome
Customers closed5 logos · $145K ARR
BANT-qualified SQLs8 · $264K ARR pipeline
Deals in active follow-up28 · $633K+ ARR pipeline
Positive reply rate1.5%+ across 150+ domains
Engine deploy time2 weeks → under 3 days
Research throughput6+ hrs/batch → under 30 min
Positioning iterations4 shifts, market-validated

"Every workflow either attributes to revenue or gets killed. That discipline is the whole system." - the operating principle behind the build

Why this matters if you're venture-backed

Seed-to-Series-B math is unforgiving: the plan says 3x pipeline, the budget says 1.2x headcount. Traditional GTM closes that gap with more people and more handoffs - each handoff leaking context and pipeline. A system-led build closes it with compounding infrastructure: the engine that took 2 weeks to deploy now takes 3 days, the copy that needed review now ships itself, and the research that took an analyst-day runs in 30 unattended minutes.

This was greenfield build number three - after an IT services firm ($180K+ ARR growth, 40% pipeline-efficiency gain) and a MENA cybersecurity startup (zero-to-one ABM engine). The pattern holds across markets: strategy first, system second, revenue attribution always.