Base Cloud
Sales Cloud, Service Cloud or an equivalent bundle. If you already use Salesforce, this cost is usually sunk.
FAQ
Agentforce is licensed on top of your existing Salesforce edition. You need the right base Cloud, the Agentforce add-on for the channel you want, and enough user seats to cover the people or conversations the agent will handle.
Salesforce AI agents are not a standalone product you buy instead of Sales Cloud or Service Cloud. Agentforce is an add-on that sits on top of an existing org, and it is priced per user or per conversation depending on the product.
For most Nordic companies the first decision is which cloud the agent will live in: Service Cloud for support use cases, or Sales Cloud for sellers and revenue operations. Each has its own Agentforce SKU and packaging.
Agentforce for Service gives you AI agents that answer customer questions over chat, messaging and email, resolve routine cases and hand over to a human with full context.
Licensing is conversation-based. You purchase a bundle of autonomous conversations — typically billed monthly — and allocate them to the channels where the agent is active. A conversation is counted when the agent handles a customer interaction from start to finish without a human taking over.
Agentforce for Sales helps sellers qualify leads, research accounts, draft follow-up emails and update opportunities. It acts as an assistant inside Sales Cloud rather than a replacement for the seller.
This SKU is user-based. Each seller or revenue operations user who gets AI assistance needs an Agentforce for Sales license. It is usually sold in monthly per-user packs, with volume tiers that make it cheaper per seat at scale.
A realistic first budget has three parts: the base cloud licenses you already have, the Agentforce add-on, and the implementation work to make the agent trustworthy.
Sales Cloud, Service Cloud or an equivalent bundle. If you already use Salesforce, this cost is usually sunk.
Conversation packs for Service, or per-user seats for Sales. Start with a small pilot group and scale from real usage.
Readiness review, data clean-up, topics, actions, guardrails and testing. This is where most first projects spend the majority of the initial budget.
Salesforce's AI trust layer keeps your customer data inside Salesforce's own security boundary. It is never used to train external AI models. Read how it works and why it matters for GDPR.
Read the answerTell us what you are trying to automate. We will map it to the right Salesforce base edition, add-on and pilot scope — with an honest view of what can wait.
Ask about licensing