PreviewYou're previewing CRO AI Momentum. Sign in to use it.Sign in
← All lessons
L5 · Expert: the AI-era revenue organization9 min read

Designing the AI-era revenue organization

When agents do part of the work, the shape of the revenue organization changes: fewer people doing repetitive tasks, more people supervising, specialising and building relationships. The change is uneven across roles, and the best designs are deliberate.

What is already happening

  • Salesforce describes a hybrid SDR role where a person paired with an agent oversees 15,000 leads a month instead of 500.
  • In September 2025 Salesforce's CEO said support headcount had fallen from about 9,000 to about 5,000 as AI agents took on work, while the company said hundreds of employees were redeployed into professional services, sales and customer success.
  • The Bridge Group's 2025 SDR report found 2025 was the first year "AI SDRs" appeared as a distinct category (1% of respondents) — early days for most organizations.
  • Gartner predicts AI agents will outnumber sellers 10 to 1 by 2028, and warns of "agent sprawl" if data, workflow and seller experience are not redesigned.

Design questions

  1. Which tasks move to agents? Map tasks, not jobs. Most roles keep much of their work but change how it is done; a few roles change shape entirely.
  2. What new roles are needed? Agent supervisors, RevOps AI owners (configuration, data, guardrails), conversation and content designers, revenue analysts.
  3. Where do humans concentrate? Complex deals, executive relationships, expansion in key accounts, handling escalations.
  4. What are the spans of control? A manager of humans plus agents needs different metrics and dashboards.
  5. How do people move? Reskilling paths from SDR to agent supervisor, from support to customer success.

Capacity planning

Plan capacity in work units (accounts covered, leads worked, meetings held, renewals managed), then decide which units humans, agents or both deliver. Use pilots to measure agent throughput and quality rather than vendor claims.

Avoid two traps

  • Cutting first, learning later: Klarna's public step back on AI-only support shows the cost of optimising for headcount before quality.
  • Adding agents without redesign: the "agent sprawl" Gartner warns about — more activity, same results.

What this means for the CRO

Org design is where AI becomes a CRO decision rather than a tooling decision. Build a 12–24 month view of roles, spans and capacity with HR and finance, based on measured pilot results, and update it each half-year.

Be explicit with your teams about the direction: which roles change, what new roles open, and how people can move into them. Silence breeds attrition among the people you most want to keep.

Revisit the plan as measured results come in; the right ratio of people to agents will differ by segment and will keep moving as tools improve.

Real example

Fortune reported on September 2, 2025 that Salesforce CEO Marc Benioff said the company had reduced its support headcount 'from 9,000 heads to about 5,000' with AI agents; Salesforce said it had redeployed hundreds of employees into professional services, sales and customer success. Source: Fortune — Salesforce CEO Marc Benioff says his company has cut 4,000 customer service jobs as AI steps in (Sep 2, 2025) ↗

Use it

Quick check · 3 questions

  1. 1.What is the best unit for planning capacity with agents?

  2. 2.What is 'agent sprawl'?

  3. 3.Which trap does the Klarna story illustrate?

0 of 3 answered

Sources

Last reviewed 2026-10-01.