PreviewYou're previewing CRO AI Momentum. Sign in to use it.Sign in
← All lessons
L0 · AI basics for revenue leaders6 min read

Assistants, copilots and agents: who does the work?

The words vendors use — assistant, copilot, agent, "digital worker" — blur together. A useful way to see them is by who decides and who acts.

Who decidesWho actsRevenue example
AssistantHumanHumanA seller asks AI to research an account and draft an email, then edits and sends it.
Copilot / embedded AIHuman, with AI suggestionsHumanThe CRM suggests next steps or flags a deal risk; the seller chooses.
Agent with approvalAI proposesHuman approves, AI executesAn agent drafts follow-ups to 300 inbound leads; a rep approves the batch.
Autonomous agentAIAIAn agent emails low-score leads, answers replies and books meetings directly.

Moving right adds scale and risk

Each step to the right takes a human out of a loop. That is where the big capacity gains come from — and where brand, legal and data problems appear. Most revenue teams end up with a mix: autonomous agents for narrow, low-risk tasks; approval steps where a wrong message is costly; and human-only work for complex, high-value deals.

The hybrid job

When agents do the repetitive part of a role, the human role changes. Salesforce describes a hybrid set-up in which a human SDR is paired with an SDR agent: instead of overseeing 500 leads a month, the person oversees 15,000, because the agent does most of the outreach. That person is no longer mainly a writer of emails. They are a supervisor, a quality checker and the one who handles the conversations that matter.

What stays human

  • Judgment on strategic accounts and complex deals
  • Negotiation and executive relationships
  • Deciding what the agent is allowed to say and to whom
  • Owning the outcome when something goes wrong

What this means for the CRO

For a CRO, the practical decision is how far right to go, task by task — not "AI or no AI".

Set a rule of thumb for your organization: autonomous only where the downside of a mistake is small and reversible; approval steps where a message could damage a key account or break a rule; human-led where the deal is complex or strategic.

Then plan for the people side. Hybrid roles need new skills (supervising agents, reviewing output, handling escalations), new metrics and, eventually, new comp. Those changes take longer than the software rollout.

Real example

Salesforce's senior director of sales development, Vanessa Tabbert, described a hybrid role in which a human SDR is paired with an SDR agent: "Instead of overseeing 500 leads a month, they'll oversee 15,000, because the agent will do most of the outreach." Source: Salesforce — Agentic sales: how Salesforce found a better way to sell (Oct 9, 2025) ↗

Use it

Quick check · 3 questions

  1. 1.An AI drafts follow-up emails for 300 leads and a rep approves them before they are sent. What is this?

  2. 2.Where is a fully autonomous agent most appropriate?

  3. 3.What changes most for a human SDR in a hybrid model?

0 of 3 answered

Sources

Last reviewed 2026-10-01.