Go-to-Market

Go-to-Market is goClaw's flagship use case. An outbound GTM agent handles the full prospecting cycle autonomously: identify targets, research them, write personalized outreach, manage multi-channel follow-up, and keep the CRM updated — without you building a workflow or writing a prompt library.

What the agent does

A GTM agent configured for outbound typically:

1. Researches prospects — web search on company, role, recent news, product signals
2. Files knowledge — saves research to the knowledge base for future reference
3. Drafts personalized outreach — first-touch email or LinkedIn DM grounded in research
4. Executes outreach — sends via email or SMS
5. Schedules follow-up — 3 business days default, configurable
6. Manages responses — classifies replies (interested / not interested / out of office / objection)
7. Handles objections — uses objection knowledge to respond
8. Qualifies leads — identifies signals of fit and flags hot leads in the CRM
9. Books meetings — if calendar integration is configured, proposes and books times

Configuration for GTM

Agent persona

agent:
  name: "Jordan"
  persona: |
    You are Jordan, an AI SDR for Acme Corp. You handle outbound prospecting
    for the Acme platform — a B2B SaaS tool for engineering teams.

    You are direct and technical. You assume the person you're contacting is
    busy and skeptical. Your goal is to get them to a 20-minute discovery call,
    not to close a deal in one email.

    You are transparent that you are an AI. If asked directly, say so.
    You do not use corporate jargon or generic value propositions.

  goals:
    - Research each prospect before contacting them
    - Write a first-touch email that demonstrates you've done your homework
    - Follow up twice if there's no response (3 days, then 7 days)
    - Qualify leads: are they the right company, right role, right timing?
    - Book a discovery call for qualified leads
    - Keep the CRM up to date after every interaction

ICP knowledge file

The ICP (Ideal Customer Profile) file tells the agent what "qualified" looks like:

# Ideal Customer Profile

## Target companies
- B2B SaaS companies, Series A through Series C
- 50–500 employees
- Engineering-led growth or strong technical culture
- Pain: manual customer success workflows, scaling support without headcount

## Target roles
- CTO, VP Engineering, Director of Engineering
- Head of Customer Success, VP Product
- Avoid: non-technical founders, BD titles, junior individual contributors

## Positive signals
- Recently raised a funding round
- Hiring multiple SDRs or CSMs (scaling GTM)
- Company mentions "doing things manually" in content
- Job postings for operations roles

## Negative signals
- Less than 20 employees (too early)
- Competitor tooling already adopted (Clay + Outreach already in stack)
- Company in a declining market segment

Example agent execution

Here's a trace of the agent processing an inbound lead from a conference:

[2026-03-01 09:12] Task: outreach_new_prospect
  Contact: maya.chen@acme.io (CTO, AcmeCo)
  Source: Conference badge scan

[09:12] crm_create_contact: Created contact #4421
[09:12] knowledge_search: "AcmeCo company research" → no results
[09:13] web_search: "AcmeCo SaaS 2026 product news funding"
[09:13] knowledge_file_curiosity: "What is AcmeCo's product and customer base?"
[09:14] web_search: "Maya Chen CTO AcmeCo LinkedIn background"
[09:15] knowledge_search: "email outreach CTO engineering" → skill: cto_outreach_v3
[09:15] Drafting email using cto_outreach_v3 skill + AcmeCo research
[09:16] send_email: To maya.chen@acme.io, Subject: "Re: AcmeCo's engineering workflow"
[09:16] crm_add_note: "Sent intro email. Referenced recent Series B and hiring surge."
[09:16] crm_schedule_followup: +3 days, channel: email
[09:16] Task complete. Next: follow-up on 2026-03-04

Multi-channel sequence

A typical outbound sequence across channels:

Day	Channel	Action
0	Email	Personalized intro email
3	Email	Follow-up #1: add a new angle (customer story, data point)
7	Email	Follow-up #2: explicit ask for call or alternate contact
10	SMS	Brief check-in (if phone number available)
No response	—	Archived. Re-engage after 60 days if new signal detected

The agent manages this sequence autonomously. It doesn't just schedule messages — it writes each one based on the accumulated context from previous interactions and new research.

Lead qualification signals

Configure the agent to watch for qualification signals in responses:

qualification:
  positive_signals:
    - "interested"
    - "let's talk"
    - "tell me more"
    - "how does it work"
  negative_signals:
    - "not interested"
    - "we're good"
    - "already using"
    - "unsubscribe"
  escalation:
    trigger: "hot_lead"
    action: notify_human
    channel: slack
    recipient: "#gtm-hot-leads"

When a prospect responds with positive signals, the agent flags them as a hot lead and notifies the human team via Slack (if configured). The human takes over for booking and closing.

Reporting

The admin dashboard's GTM view shows:

• Prospects contacted this week / month
• Reply rates by channel
• Qualified leads in pipeline
• Open follow-up tasks
• Knowledge files created from prospect research