Agentforce 2.0 arrived in December 2024 with more skills, Slack integration and an expanded path for building agents. As capability grew, the planning unit had to move beyond the individual use case.
A company can easily create agents that duplicate one another, compete for the same action or move a bottleneck downstream. Digital labor design considers the complete system of work.
Cloud Group point of view
Build a capability portfolio, not an agent zoo. Shared context, actions, supervision and outcome measures should increase reuse as the number of agents grows.
A practical playbook
The strongest next step is narrow enough to govern and useful enough to produce evidence. We would structure the work around these moves:
- Map value streams and locate work constrained by context switching or queues.
- Define reusable capabilities before packaging them into role-based agents.
- Decide which agent owns orchestration when multiple roles touch one outcome.
- Create a common supervision and escalation model.
- Fund the portfolio based on business throughput, not number of agents launched.
The architecture and operating implication
Use a shared catalog for data products, topics, actions and policy. Keep agent-specific instructions thin and role-oriented. Central telemetry should let leaders compare performance without forcing every team onto an identical operating cadence.
Measure what changes
Model activity is not a business result. Track a small set of indicators that connect behavior to accountable work:
- Capability reuse across agents
- End-to-end cycle time across human and agent steps
- Duplicate actions or conflicting ownership
- Portfolio value relative to consumption
Digital labor becomes strategic when it improves the system, not simply a task. Agentforce 2.0 made that wider design responsibility difficult to ignore.
Primary sources
This field note is grounded in the product and market context available at the time of publication.
