A major headline grabbed the attention of everyone in Salesforce AI news circles recently. On 13th October 2025, the customer relationship management (CRM) giant launched Agentforce 360 as the latest version of its AI agent platform: Salesforce Agentforce.
Salesforce has now added voice to Agentforce 360, which will allow clients to go beyond text while using AI agents to answer customer queries. This is a very promising development in the Salesforce Agentforce platform that will converge human teams, AI agents, and enterprise data for a 360-customer experience.
Agentforce can help businesses triple their agentic AI investments with proper implementations. However, almost a year after its release, the adoption rate of Salesforce Agentforce AI isn’t what the company expected. There is still some skepticism as businesses aren’t sure whether Salesforce Agentforce will get them a better return on investment (ROI) or not.
In this article, we’ll help businesses get a clear picture by showing what Agentforce delivers in real terms based on its features and use cases. In the end, it will explore Salesforce Agentforce pricing models and how they relate to your agentic AI ROI.
What is Salesforce Agentforce?
TL;DR
Salesforce Agentforce overview
- Agentforce Salesforce is an agentic AI platform that lets organizations create autonomous, task-performing AI agents within their CRM.
- It uses real-time, trusted business data from Salesforce and connected systems so agents can make context-aware decisions safely.
- Agentforce integrates fully with the Salesforce platform and its products, like Marketing Cloud, Service Cloud, and Sales Cloud.
Salesforce Agentforce AI is a platform that allows users to build and deploy autonomous AI agents in Salesforce CRM to automate tasks, improve customer service, and support sales and marketing initiatives.
You just describe to the platform what you want your agent to do, set up the guardrails, and the agent will get the job done in no time. The Salesforce Agentforce platform was released in December 2024. Built on the Einstein 1 Platform, it combines Salesforce Einstein GPT and Data Cloud to retrieve data from company resources and customer databases for 24/7 customer support.
The primary tasks performed by a Salesforce Agentforce service agent include:
- Qualifying sales leads based on CRM data
- Resolving customer cases and escalating complex cases to human agents
- Segmenting marketing audience using Data Cloud
- Triggering business processes autonomously, like sending notifications
There is also a fair bit of confusion between Salesforce Agentforce and Einstein AI. Salesforce Einstein AI is a suite of AI capabilities embedded within Salesforce, which enhances core CRM functionalities. However, it is not an autonomous agent itself.
As opposed to Salesforce Agentforce, which creates agents that can operate independently on behalf of users.
Top Salesforce Agentforce features
When Agentforce was launched, Salesforce’s CEO Marc Benioff said that Agentforce is what AI was “meant to be.” Salesforce Agentforce has some advanced capabilities that make it the third wave of AI beyond conventional copilots.
Here are some of the top Salesforce Agentforce features:
1. Agent Builder
The Agent Builder is a key component of Agentforce’s architecture. In simple terms, it is a tool that allows users to create, edit, and manage AI agents inside the Salesforce CRM. Agent Builder is a no-code/low-code interface within Salesforce Agentforce AI, which allows admins and salesforce developers to design and deploy AI agents that can perform tasks with minimal human intervention.
2. Einstein Trust Layer
Einstein Trust Layer on the platform ensures that AI agents operate within established guardrails. It is an agentic AI security mechanism in Salesforce Agentforce for the secure operations of agents.
This layer helps prevent unauthorized data access while respecting privacy and compliance policies.
3. Agent Script
Agent Script is the latest feature addition to the new Agentforce 360 upgrade. It is a high-level, human-readable scripting language designed for controlling agents. You can control your AI agents with simple if/then rules using the scripting language in Agentforce.
4. Atlas Reasoning Engine
Atlas Reasoning Engine is the “brain” behind Salesforce Agentforce. It enables AI agents to reason, plan, and execute tasks autonomously within business workflows. And in Agentforce 360, it is now configurable, so teams can decide how much freedom or structure the LLMs of AI agents should have.
Agentforce 360 also supports Google Gemini, OpenAI, and Anthropic’s models for the Atlas Reasoning Engine, which gives users a variety of AI models for reasoning and planning.
5. Data Cloud integration
Data is the lifeblood of the Salesforce Agentforce platform. Users rely on data to train their agents to perform desired tasks. This data can be structured or unstructured, and in terms of Salesforce, structured data means an internal Salesforce record, and unstructured data mean emails or memos.
Agentforce is directly connected to Salesforce Data Cloud, which gives agents the full picture of your organization in real time. For example, if a customer asks a query that the agent doesn’t have the answer to immediately, it will use RAG models to fetch business data that is relevant to the customer’s query.
This enables agents to give relevant, smart responses based on unified business data during conversation. Moreover, this integration extends beyond Data Cloud to other Salesforce products as well.
Using Salesforce Customer 360, you can connect an existing Salesforce product with other Salesforce offerings to get a full 360 view of the customer journey.
Salesforce Agentforce use cases
There are several ways you can use Salesforce Agentforce AI to deploy agents in different business operations. Wiley, a big name in online education and publishing, is one of the long lists of Salesforce success stories.
The education platform faced a spike in service requests from students at the start of every new semester. Their support team needed a better way to handle this high volume without overloading human agents.
Wiley deployed AI agents using Agentforce for self-service, autoreplies, and resolution of common problems like login issues. Salesforce Agentforce handled routine customer queries to provide a better experience to students and made human agents more productive. Wiley also saved around $230,000 annually through automation.
Here are some ways you can also use Agentforce in your business:
1. Banking agent
The common perception is that banking roles are sinecures. You get to sit in air-conditioned rooms all day, get paid well, and just click around on your computer. But the reality is far from it. A job at a bank can be extremely hectic and high-pressure because of client demands, deadlines, and the constant push by higher-ups to hit financial targets.
For banks that use Salesforce, which are many in the United States and Canada, Agentforce can deploy agents to automate banking and investment tasks. It can help in analyzing data, managing client requests, and providing financial advice based on banking data.
2. Sales development rep
Sales representatives can use Agentforce to respond to leads quickly through agents. Such AI-powered sales development reps can capture client data and feed it to Salesforce Sales Cloud for a Customer 360 view.
For example, a conversational bot answers the questions of a potential lead. If the lead is warm, the bot can book meetings or pass the lead to human agents for more insights. This makes the lead identification and qualification process faster and more efficient.
3. E-commerce assistant
Shopping has always been a social activity. Research has shown that having a companion with you enhances the shopping experience. That is why e-commerce and online retail stores are now looking to bring this social connection to online shopping.
Online stores widely use Salesforce Commerce Cloud. Agentforce can deploy conversational agents that act as shopping partners for users. They can recommend buyers’ products, browse catalogues for them, and assist with checkouts or shipping.
4. IT helpdesk agent
IT operations and services within an organization are usually interconnected. Resolving all the knots in this tangled IT web consumes a lot of time and effort. Agentforce IT Service is one of the popular options for IT service management (ITSM) in the market.
It is facing tough competition from products like ServiceNow ITSM, but it’s still a top choice. Salesforce Agentforce can be used to handle customer or employee IT tickets with ease.
AI agents will troubleshoot software and device issues for you, which will cut down on the internal workload of human IT support agents. You can also create agents that can monitor recurring IT issues and prevent them from happening proactively.
5. Customer support agent
As in the case of Wiley, you can deploy AI agents using Salesforce Agentforce to help human agents work faster and better. Regardless of what industry your business works in, AI agents can boost agent productivity, improve consistency, and speed up resolution time.
Customer experience is one domain that your business must excel in. Happy customers are your brand ambassadors. The goodwill they create about your company through word of mouth is more powerful than many marketing campaigns or paid advertising.
Salesforce Agentforce pricing model
We now know how Salesforce Agentforce works and its use cases. But the financial aspects of using Salesforce Agentforce AI are as important. In May 2025, Salesforce made some major changes to the pricing of Agentforce.
Salesforce claimed that it has introduced some major improvements in Salesforce Agentforce pricing:
- Flex credits
- Conversations
- Flex agreements
- Use licenses and add-ons
Let’s unpack everything.
1. Three buying models
First up, there are three ways you can buy Salesforce Agentforce or make a commitment that affects pricing.
- Pre-purchase: You pay upfront for a set amount of usage for the full contract term.
- Pay-as-you-go: You pay only for what you use, with no upfront commitment. Usage is metered, and you’re billed monthly.
- Pre-commit: You agree to use and pay for a minimum level of usage over the contract period, in exchange for a discount. The more you commit, the more features and benefits you get.
2. Flex Credits
The Flex Credits system bills you per “action” that the Salesforce Agentforce service agent performs. That action can be anything, like updating a record, automating a task, or summarizing a case. It is a consumption-based model, which ensures you pay for the specific actions Agentforce performs.
- Each action in Agentforce consumes 20 Flex Credits at $0.10 per action.
- Flex Credits bought in bulk come in packs of 100K Credits at $500.
3. Conversations
The Conversations pricing model in Salesforce Agentforce has a flat rate of $2 per conversation. What is a conversation in Agentforce? It is the entire interaction session between a user and an AI agent, regardless of how many actions happen within it.
A full chat session between a customer and an AI assistant will count as one conversation.
4. Flex Agreements
Flex Agreements is a new Salesforce Agentforce pricing model that allows businesses to buy and manage the platform’s AI features flexibly, based on Flex Credits. It is basically buying a bundle of Flex Credits at a discounted price that you can negotiate.
However, this agreement also allows you to use Credits to spread the AI features across multiple Salesforce AI products, like Einstein and Copilot.
5. Agentforce add-ons
An Agentforce add-on is an extra feature pack you can apply on top of an existing user license. They are usually used for internal, employee-related issues in Agentforce. Add-ons are like an extra layer of AI functionality.
- Add-ons for Sales, Service, and Field Service cost $125/user/month
- Add-ons for Salesforce Industries cost $150/user/month
- Agentforce 1 Edition offers advanced add-ons for $550/user/month
How Agentforce pays off for your business: Measuring ROI
We know that the above information was a bit too much to grasp. Salesforce Agentforce’s pricing structure is complicated and opaque. It is one of the reasons why teams have been skeptical about adopting Agentforce, particularly small and medium enterprises (SMEs). And it is perfectly understandable; no business would invest in any technology without understanding the potential return on investment (ROI).
But Salesforce has addressed many of these customer concerns in the new pricing model of Agentforce. Furthermore, calculating ROI depends on what you’re trying to measure: short-term gains or long-term value?
Long-term ROI sustainability
If it’s the latter that you want, then Agentforce can pay off more quickly than you might think. The best thing about Agentforce is its modular approach, which means you don’t have to roll out everything from the start. Yes, Agentforce offers lots of agents for many jobs, but you don’t need every agent.
Starting small by choosing just one high-volume issue to solve with agents, and then gradually automating everything else, will show tangible results. The ROI in this way will accumulate within 2-3 years to cover all the investment costs of implementing the Salesforce Agentforce platform.
This step-by-step pilot phase approach will also determine which pricing model and use cases deliver the highest ROI. You can also use the Agentforce ROI Calculator provided by Salesforce to forecast how Agentforce can benefit your business.
Chasing quick wins
Many businesses can’t wait two or three years to see paybacks. If you want short-term gains from Agentforce, there are still clear ways to achieve them. The key is to target high-impact and low-complexity use cases. Automation can quickly reduce workload and boost efficiency in these areas, such as:
- Lead qualification
- FAQ handling
- Password resets
Salesforce also offers preconfigured agents for common functions across sales, service, and marketing. Using these avoids long customization cycles, helping you deploy and realize ROI within a single quarter.
Conclusion
Salesforce is a trend setter in the world of technology. It is the platform that introduced the revolutionary Software-as-a-Service (SaaS) model to the world. At the time when most software required local installation, Salesforce delivered applications entirely over the internet. This was the dawn of the cloud computing era.
And now in the era of AI, Salesforce Agentforce is the company’s signal to the world that Salesforce is here to lead the next wave of AI agents. However, there is some dithering on the part of businesses in putting their trust and money into the platform.
But it all comes down to how companies measure their agentic AI ROI with Agentforce. Are they planning to play the long game? Or are they just content with making some quick gains and calling it a day? The Salesforce Agentforce platform has something to offer both of these customers, but it matters whether they know what they want before throwing in the towel.
If you’re also ambivalent about using Agentforce, Xavor can help you make the right choice by providing you with a roadmap. Our Salesforce consulting services work with clients to help them make the most out of their Salesforce investment in different products.
Partner with us by dropping a line at [email protected].