Skip to main content
A lead lands from a website form at 11pm. By the time you open your laptop the next morning, it has been qualified, given an owner, and scheduled for its first follow-up — and the timeline shows exactly which agent did it and when. None of that happens without structure underneath. That structure is the pipeline: the path a lead travels from first contact to closed, and the record of everything that happened along the way. This page covers how Apollo Space models that funnel — pipelines, stages, leads, and activities — so your team, human and agent alike, always knows where every deal stands and who has the ball.

​
Simple hierarchy

Each org can have multiple independent pipelines. Each pipeline has its own stages. A lead lives in ONE pipeline + ONE stage at a time (but can be moved between pipelines).

​
Pipeline

A pipeline represents a specific flow in your funnel — the path a lead travels from start to finish. Common examples: Keeping pipelines separate means each one tracks its own metrics and cadence. Outbound conversion rates should never be mixed with inbound.

​
Stages

Stages are the steps a lead moves through inside a pipeline. Sensible defaults for B2B outbound:
You can customize freely — add / remove / rename / reorder stages. Stages can be archived (hidden from the UI, but historical leads retain whichever stage they were in at the time).

​
”Closed” stages

Stages can be marked as closed-won or closed-lost. This:
  • Moves the lead to the “history” view automatically
  • Counts toward conversion rate metrics
  • Signals to agents that “no further action is needed on this lead”

​
Lead

A lead is a card that lives in a pipeline stage. Each lead carries:
  • Contact data (name, email, phone, website)
  • Company data (name, location, industry)
  • Owner — the person responsible (human or agent)
  • Potential value — what closing this lead is worth (in cents of the configured currency)
  • Tags — customizable labels
  • Custom fields — fields defined by your org

​
Owner

The lead’s “owner” determines who the system notifies on events (reply received, time for a follow-up, etc.). The owner can be:
  • A human (org member)
  • An agent (Marcus, Athena, or a custom agent)
When an agent is the owner, the agent operates autonomously according to its configuration — and routes events to the human supervisor when appropriate.

​
Activities

Every meaningful lead event generates an activity in the history log. This is the CRM’s audit log — append-only, nothing is ever deleted. Automatic activity types: Each activity records who did it (human X or agent Y), when, and the change payload. Full auditability — you can reconstruct any lead’s complete history in seconds.

​
Human vs. agent actions

When Marcus moves a lead between stages (e.g., “In conversation” → “Proposal sent” after firing off a proposal), the activity carries actor_type = 'agent' + the agent’s ID. When you move it manually, it’s actor_type = 'user' + your user ID. This shows up in the lead’s timeline — instead of “Stage changed to X”, you see “Marcus moved to X” or “João Silva moved to X”. No ambiguity.

​
Filters and segmentation

Filters apply to the visible set of leads. You can combine:
  • Pipeline + stage
  • Owner (specific human, any human, specific agent)
  • Tags
  • Custom field values
  • Date (creation, last activity, next action)
  • Potential value (range)
Filters can be saved for reuse. They are per-user — each person has their own.

​
Next steps

CRM (feature)

How to work day-to-day: kanban, list view, custom fields, and import.

Outbound

How leads become campaigns managed by Marcus.

Agents

When an agent is a lead’s owner, it operates autonomously.