Answer capsule
Salesforce says its Hunter outbound sales agent can work from research through outreach over weeks or months and that a new runtime can preserve memory, continue execution, and adjust plans as information changes. A revenue leader should not govern that work as a longer chat. Every material change to target, contact, message, opportunity, forecast, commitment, or next step needs a dated reason, evidence link, authority state, seller approval where required, and recovery path.
What the source establishes
- Salesforce says Hunter is an outbound sales agent that can work a pipeline from research to outreach with sellers over weeks and months; it was in pilot on September 11 with general availability stated for November 2026.
- The announced long-horizon runtime uses memory, durable execution, and dynamic steering to preserve progress, continue work, and adapt plans across days and weeks.
- Salesforce's at-risk-deal example says an agent can turn an objective into a measurable goal and plan, with guardrails determining when it may act and when seller approval is required.
- The page reports that 60% of Perk's sales pipeline is built by Hunter, but it does not disclose the pipeline definition, denominator, time window, quality, conversion, incremental contribution, or independent validation.
Define the revenue state the agent is allowed to change
List every object and field the agent can read, propose, create, update, send, schedule, or close across account, contact, lead, opportunity, activity, sequence, quote, contract, forecast, and customer records. For each state transition, name the required source evidence, freshness limit, business rule, owner, approval, and reversal. Separate research and drafting from external outreach, CRM writes, stage changes, amount and probability changes, discount or term commitments, and forecast submission. Preserve channel consent, suppression, territory, account ownership, contact identity, relationship history, and legal or contractual restrictions. A broad objective such as rescuing at-risk deals cannot itself authorize every action that could pursue it.
Log how the plan changes over time
For every consequential plan revision, retain prior and new goal, trigger, new evidence, model and agent version, memory used, rules evaluated, tool call, proposed action, approval state, execution result, downstream record, exception, and responsible seller. Make the timeline readable from the CRM rather than hiding it in an opaque session trace. Test stale contact data, conflicting stakeholder signals, ownership transfer, account suppression, changed pricing, a closed-lost reason, late-stage procurement, renewal overlap, duplicated tasks, and a seller who revokes direction. Memory can improve continuity while also preserving an error; durable execution can continue value-creating work or continue the wrong work. Both require explicit expiry, correction, and stop controls.
Keep pipeline creation separate from revenue contribution
Define sourced pipeline, influenced pipeline, accepted opportunity, qualified stage, closed-won amount, recognized revenue, retained revenue, and incremental contribution before reporting performance. Reconcile every counted opportunity to creation time, prior human activity, contact source, opportunity history, seller acceptance, stage evidence, duplicates, splits, reversals, and outcome. Compare a governed pilot with a credible baseline while controlling for segment, territory, seasonality, rep tenure, marketing, pricing, and product change. Include software, data, enablement, review, remediation, compliance, and seller time in cost. A provider customer percentage can motivate diligence; it cannot establish the agent's causal contribution or economics in another revenue system.
Expand only when sellers can inspect, stop, and recover
Require sellers and RevOps to see active goals, next planned actions, evidence, memory age, approval requirements, and all executed changes. Test global and account-level pause, approval timeout, access revocation, rollback, reassignment, duplicate prevention, complaint handling, suppression propagation, and incident reconstruction. Review message accuracy, relevance, consent, brand and relationship harm, opportunity quality, seller workload, forecast accuracy, customer outcome, conversion, and total cost. Expand only when the ledger reconciles to CRM and communication systems, exceptions stay within threshold, and humans can intervene before a material commitment. Keep pilot status distinct from announced general availability and revalidate the current product and terms before purchase.
Turn this source into a reviewable decision
For AI for Chief Revenue Officers, use this briefing as a dated decision record rather than a substitute for the source. Preserve Salesforce Expands Agentforce With a New Portfolio of AI Agents Built for High-Value Work, the exact URL, the September 12, 2026 review date, the supported facts above, the editorial interpretation, the limitations, and any buyer-specific evidence. Link that record to the decisions most directly affected: Seller outreach assistance; Pipeline inspection and deal risk; Revenue forecasting; Revenue operations and data quality. State whether the source changes the scope, evidence requirement, control, sequence, or only the language used to describe the decision.
Before action, name the accountable owner, affected population and workflow, exact offering or configuration, source data and rights, human decision point, exception and appeal path, complete cost, expected benefit, failure and stop conditions, retained evidence, and next review date. Keep official facts, provider statements, buyer observations, representative tests, measured outcomes, editorial inferences, and unknowns visibly separate. Reopen the record when the source, offer, model, integration, data, policy, population, responsible person, or measured result changes.
Limitations and unknowns
Salesforce is the provider, and its feature, availability, and customer-result descriptions are provider claims. The September 11, 2026 publication is a verified post-cutoff material source; Hunter was described as a pilot with November general availability, while other capabilities have different statuses and the page warns that some references may be unreleased. It does not establish a buyer's permissions, data quality, consent, pipeline definition, opportunity quality, conversion, revenue attribution, relationship effect, forecast accuracy, control operation, or ROI. Current product documentation, contracts, CRM and channel records, traces, pilot evidence, and qualified revenue, sales, RevOps, marketing, customer, finance, legal, compliance, privacy, security, data, and procurement review control.
Decision test
Ask whether the source changes the decision itself, the evidence required, the implementation sequence, or only the language used to describe an existing capability. Record which claims are directly supported, which are provider statements, which require an independent test, and which remain unknown. A source-linked review should make uncertainty easier to see, not bury it inside a blended score.
Questions to take into review
- Why is this contact appropriate now?
- Which claim and source support each sentence?
- What evidence defines each stage?
- Which risk factors are causal, correlated, or heuristic?
- How is error measured across horizons and segments?
- What happens when market conditions shift?
- Which fields may change automatically?
- How are false merges detected and reversed?
The publication supports research and executive decision preparation. It does not provide legal, financial, accounting, employment, clinical, cybersecurity, investment, procurement, or implementation advice.