Start free

Overview

The Client Path mental model in five minutes.

Client Path is a research-driven CRM for B2B sales teams. You define your offering, import your prospects, and the platform researches each one against your criteria, surfacing the signals that matter and scoring how likely each prospect is to buy right now.

The core loop

  1. Define: you create an offering that describes what you sell and a methodology that tells the agent what to look for.
  2. Import: you add clients (the companies you are prospecting).
  3. Research: you start a research run. The agent reads public sources, extracts findings, and produces a propensity score for each client.
  4. Act: you read the findings, draft outreach, and move clients through your pipeline. Automations can do this automatically.

What the agent does

The agent is not a chatbot. It is a pipeline worker. It:

  • Researches prospects using web search and document analysis.
  • Extracts structured findings organised by topic (technology, buying team, budget signals, competitive landscape).
  • Scores propensity on a 0-100 rubric anchored to your methodology.
  • Drafts outreach messages grounded in the findings.
  • Monitors signals (external events like company news) and fires automations when conditions are met.

When an automation's autonomy requires human approval, the task surfaces in your inbox. You approve, reject, or snooze it.

Key entities

EntityDescription
OfferingWhat you sell: name, description, and rubric criteria
MethodologyHow to research: a scoring rubric mapped to finding topics
ClientA company you are prospecting
Research runA batch job researching many clients against one offering
Client researchOne client's research within a run: findings, score, report
FindingA discrete piece of intelligence about a client
Propensity score0-100 likelihood to buy
AutomationA trigger-step pair the agent evaluates automatically
Agent taskA pending action waiting for your approval
SignalAn external event (company news, stage dwell, etc.)

Next steps