Salesforce News

Agentforce is Salesforce’s AI-powered platform for building and using autonomous agents across the Salesforce ecosystem. It helps organizations bring AI into everyday business processes and experiences. Agentforce is not just for your users or customers anymore. Salesforce has introduced a new feature called Setup with Agentforce to help Salesforce admins inside Setup. In this post, you will learn what Setup with Agentforce is, how it works, and what you can do with it. Introduction to Setup with Agentforce Salesforce admins are responsible for configuring new features, building customizations, maintaining security, and keeping their org running smoothly. Many of these tasks require navigating through multiple Setup pages and completing several steps along the way. Setup with Agentforce simplifies this experience by letting admins collaborate with an AI agent directly in Salesforce Setup. You can use natural language to complete supported tasks, find information, troubleshoot issues, and get guidance based on Salesforce best practices. Instead of manually navigating through different Setup pages, you can simply have a conversation with the agent and let it help you get the job done. How to Enable Setup with Agentforce To turn on Setup with Agentforce, first make sure Data Cloud is enabled in your org. Then, search for Setup with Agentforce in Salesforce Setup and enable the feature. Finally, assign the required permissions (Use Setup with Agentforce) to yourself and any other admins who need access. How to Use Setup with Agentforce Using Setup with Agentforce is simple. You can chat with the agent using natural language, and it brings the relevant Setup information and tools directly to you. It reduces the number of clicks and pages you need to navigate. Go to Setup Home or click the Ask Setup Agent button from any Setup page. Then, enter your question or describe the task you want to complete. For example, you can ask, “Why can’t Ariel Ibrahimzade see the Case records?” You can also start with one of the suggested conversation prompts and continue from there. After enabling this feature, Ask Setup Agent buttin appears on every setup page. Depending on what you ask, the agent can also display records, Salesforce documentation, and other relevant components directly in the conversation. You can expand these components to explore the information in more detail. When you create or edit something in your org, it can also show the changes directly within the flow of work. The agent can then recommend next steps or direct you to the relevant Setup pages when you need further configuration. What You Can Do with Setup with Agentforce Setup with Agentforce can help with a variety of common admin tasks. You can ask questions about your org, troubleshoot user access and configuration issues, find relevant Salesforce documentation, and get guidance on how to complete Setup tasks. The agent can also make supported configuration changes. Such as: Creating custom objects and fields Managing users and permissions Working with sharing settings Creating and managing flows Building custom report types Building Lightning pages Creating or updating formulas It can also recommend next steps and take you directly to the relevant Setup page when needed. Let’s ask it to create a new custom object with a few fields. Here is the prompt: I need to track periodic reviews for our customer accounts. Can you create a custom object called Account Review, relate it to the Account object, and add fields for the review date, review type, account health, annual revenue reviewed, next review date, review notes, and whether follow-up is required? For Review Type, use Quarterly, Annual, Renewal, and Risk. For Account Health, use Good, Needs Attention, and At Risk. As you can see, Setup with Agentforce proposes a data model for the new object according to the request. Importantly, it doesn’t create the object and fields immediately. Instead, it asks you to review the solution and confirm it before making any changes to your org. Only after you confirm the proposed solution, Setup with Agentforce proceeds with creating the object. As you can see, it creates the custom object and all of its fields for you, including descriptions for each one. This is a nice touch, since it helps keep the data model documented from the start without requiring additional manual work. Final Thoughts Setup with Agentforce brings AI directly into Salesforce Setup and can make everyday admin work faster and easier. Instead of navigating through multiple pages and figuring out each step manually, you can describe what you need, review the proposed solution, and let the agent help you complete supported tasks. It is still important to review every suggested change, and keep in mind that enabling Setup with Agentforce consumes Data 360 credits. Even with that consideration, it can significantly reduce the amount of manual work involved in many common admin activities. The post Setup with Agentforce: Everything You Need to Know appeared first on Salesforce Time.

Prompt Templates in Salesforce let you create reusable prompts that use generative AI to perform various tasks with Salesforce data. You define the instructions, provide the relevant context or data, and let the AI model generate the desired output. Since Prompt Templates use AI models, you don’t have to teach them everything from scratch. These models already understand many concepts and languages, which opens the door to many useful use cases. One of them is translating text from one language to another. In this post, you will learn how to translate text in Salesforce using Prompt Templates. Creating a Prompt Template to Translate Text Flex Prompt Templates give you more control over the inputs and instructions you send to the AI model. They are a good fit for reusable use cases that are not tied to a specific Salesforce record or predefined action. Create a Flex Prompt Template and add a Free Text input that will contain the text you want to translate. Then, enter a prompt like this: You are an AI-assisted translator. If the Text Input is not in English, translate it into English. Instructions: Follow these instructions precisely and do not add any information that is not provided. * If the Text Input is not in English, translate it into English. * Do not explain what you are doing. Return only the translated text. Text Input: {!$Input:Text_Input} Your prompt template is ready. You can now enter any text that you want to translate into English. In this example, the prompt template translates a text from Spanish into English. Calling the Prompt Template from Flow Now that we have a Prompt Template that can translate any text into English, it is time to automate the process using Flow. You can call this Prompt Template from any type of Flow, depending on your use case. In this example, we will build an on-demand solution that translates the text when the user clicks a button from a Case record. Therefore, we will use a Screen Flow. 1- Create a Screen Flow and add an input Case record variable called recordId. We will use this variable to pass the entire Case record into the Flow. 2- Let’s translate the Description field of the Case. Add the Translate Text Prompt Template that you created in the previous step to the Flow. This Prompt Template has only one input: Text Input. Set the Case Description field as the value for this input. 3- Once the text is translated, you can use the result however you want. For example, you can update the Case record with the translated text or simply display it to the user. In this example, we will display the translated text on the screen. Add a Screen element with a Display Text component, then select the output of the Prompt Template. 4- Save and activate the flow. Here is how it works. The post How to Translate Text in Salesforce Using Prompt Templates appeared first on Salesforce Time.

Meet Claudeforce, the newly expanded partnership between Salesforce and Anthropic. The companies announced it on August 26, 2026, as Salesforce kicked off its Q2 ’27 earnings call. Marc Benioff and Dario Amodei took the stage together to unveil the partnership, and CRM shares jumped in after-hours trading following the announcement. What’s actually launching The headline product is Salesforce in Claude, a plugin with 37 prebuilt sales skills that lets sellers and agents work with live Salesforce data straight from inside Claude. Think meeting prep, deal health checks, and pipeline reviews, without ever opening the Salesforce app. Under the hood, AIforce powers the experience, exposing Salesforce business data and workflows to agents through MCP servers, APIs, and CLI tools. The pitch: Claude brings the reasoning and judgment, Salesforce brings the governed data, business rules, and trusted workflows to turn that reasoning into real action. Admins can set up authentication and permissions centrally, once, instead of configuring them for each user. It’s rolling out first to a select group of pilot customers. Why this isn’t really new This is an escalation, not a first date. Claude has been a foundation model inside Agentforce since late 2025. It already powers the Atlas Reasoning Engine, runs Agentforce Vibes and Agentforce Coworker by default, and shows up in Agent Builder. Claudeforce takes that relationship and runs it in the other direction too. Salesforce data now lives inside Claude, not just Claude inside Salesforce. Both companies say more integrations across Claude, Salesforce, and Slack are coming. Benioff framed it as a direct answer to “SaaSpocalypse” chatter. The idea that agentic AI kills the need for traditional SaaS interfaces. His line: he’s done with that narrative, and Claudeforce is Salesforce’s way of proving the CRM stays relevant by becoming the trusted action layer underneath the AI, not the screen you must open. Why Salesforce professionals should care If you build on the platform, this is worth watching closely: Governance is the differentiator. Salesforce is explicitly selling “actions routed through Salesforce so business rules are always enforced”. That’s a direct pitch to admins and architects nervous about agentic AI going rogue on production data. Skills, not just prompts. 37 prebuilt sales skills suggests a packaged, opinionated skill library. It is worth comparing against what you’d build custom with Flow or Apex. Watch the MCP angle. AIforce leaning on MCP servers is a strong signal for where integration architecture is heading platform-wide. Pilot access only for now. But if you’re already deep in Agentforce or experimenting with Claude, this is one to get on the waitlist for early. Wait, isn’t this just MCP? If you’ve already connected Claude to Salesforce through a generic MCP server, you might be wondering what’s actually new here. Fair question. AIforce, the harness underneath Salesforce in Claude, is itself built on MCP servers, APIs, and CLI tools. The difference is what sits on top of that plumbing: Raw MCP gives Claude general-purpose tools: SOQL queries, record CRUD, schema describes. Claude has to improvise the workflow itself every time: figure out the right objects, chain the right calls, interpret the results. Salesforce in Claude ships 37 prebuilt skills engineered specifically for sales motions: meeting prep, deal health review, pipeline review. They use Claude’s reasoning, agentic tool use, and generative UI. Not just an API wrapped in a prompt. Governance is centralized. Salesforce routes every action through its business rules, while admins set up authentication and permissions once instead of configuring them for each user or connection. So think of raw MCP as the plumbing, and Salesforce in Claude as a finished, governed product built on top of it. The post Claudeforce: Salesforce Meets Claude appeared first on Salesforce Time.

Salesforce is adding another certification to its growing credential portfolio: Salesforce Certified Marketing Cloud Next Consultant. The new certification is designed for professionals who work hands-on with Marketing Cloud Next in consulting and implementation roles. It validates the skills needed to implement, configure, and optimize Marketing Cloud Next solutions. The certification has already been announced, but the exam is not available yet. Salesforce will launch the Marketing Cloud Next Consultant certification on August 21, 2026, with registration opening on the same day. In this post, let’s take a look at who this certification is for, what the exam covers, and how you can prepare for it. What Is the Salesforce Certified Marketing Cloud Next Consultant Certification? Marketing Cloud Next brings marketing capabilities together with the Salesforce platform, Data 360, Agentforce, and Flow. Instead of focusing only on individual marketing channels, Marketing Cloud Next enables organizations to connect customer data, build audiences, orchestrate cross-channel campaigns, and use AI agents as part of their marketing processes. The Salesforce Certified Marketing Cloud Next Consultant certification focuses on the skills required to implement these capabilities for customers. According to Salesforce, a Marketing Cloud Next Consultant should be able to work with areas such as platform setup and governance, consent management, data modeling and identity resolution, segmentation, campaign orchestration, content personalization, Agentforce, and analytics. Who Should Take the Marketing Cloud Next Consultant Exam? This certification is primarily intended for professionals who implement or administer Marketing Cloud Next in a consulting or customer-facing role. Salesforce recommends that candidates have 6 to 12 months of hands-on experience with Marketing Cloud Next. Therefore, although there are no certification prerequisites, this isn’t necessarily an entry-level exam. Candidates should already be comfortable working with Marketing Cloud Next concepts and implementing solutions in real-world scenarios. The expected skill set includes areas such as: Configuring Marketing Cloud Next environments Designing governance and Business Unit strategies Working with Data 360 Building segmentation and identity resolution strategies Managing customer consent Designing marketing campaigns and flows Personalizing marketing content Implementing Agentforce marketing use cases Working with reporting and analytics The focus is clearly on platform-native Marketing Cloud Next implementation skills, rather than traditional Marketing Cloud Engagement development skills such as extensive AMPscript or SQL development. Exam Format The Salesforce Certified Marketing Cloud Next Consultant exam includes: 60 scored multiple-choice questions Up to 5 additional unscored questions 105 minutes to complete the exam 72% passing score No prerequisite certifications Marketing Cloud Next Consultant Exam Topics The exam is divided into six sections: Campaign Design, Flow Orchestration & Content – 30% This is the largest section of the exam. Candidates need to understand how to design and configure campaigns, use Flow for marketing orchestration, work with messaging channels, and personalize content. Salesforce’s official preparation trail specifically includes resources covering marketing campaigns and flows, Flow Builder elements for Marketing Flows, campaign optimization, and email personalization. Data Modeling, Identity Resolution & Segmentation – 25% Data is another major part of the certification. Candidates should understand how Marketing Cloud Next works with Data 360, including data modeling, segmentation, and identity resolution. This includes understanding how customer data from multiple sources can be unified and then used for segmentation, personalization, and activation. Salesforce’s certification preparation trail includes dedicated material for Data 360, identity resolution match rules, and reconciliation rules. Together, this section and Campaign Design, Flow Orchestration & Content account for 55% of the entire exam. Therefore, these are likely to be the two most important areas to focus on when preparing. Platform Setup & Governance – 13% This section focuses on setting up and governing a Marketing Cloud Next implementation. Topics include areas such as Business Units, permissions, Data Spaces, content governance, and email domain configuration. One particularly important concept is how Business Units and Data Spaces can be used to implement data isolation strategies. Consent – 13% Consent management is a significant part of Marketing Cloud Next and represents 13% of the certification exam. Candidates should understand how consent data is collected, stored, and applied when communicating with customers. Salesforce has included several consent-specific resources in its official certification preparation trail, covering consent fundamentals, segmentation and consent, consent concepts, and considerations for managing marketing consent data. Agentforce & AI Innovation – 11% Not surprisingly, Agentforce also has its own section in the new certification. Candidates should understand how Agentforce can be used for marketing use cases, including creating and refining campaigns and supporting conversational marketing. Salesforce’s official preparation material includes Agentforce in Marketing Cloud Next as well as conversational marketing scenarios involving channels such as SMS, email, and WhatsApp. Analytics & Performance Insights – 8% The final section covers reporting and measurement. Candidates should understand the reporting options available in Marketing Cloud Next and how marketers can monitor campaign performance and measure results. How to Prepare for the Marketing Cloud Next Consultant Certification Salesforce has already published an official Prepare for Your Marketing Cloud Next Consultant Certification trail. Hands-on experience will also be especially important for this certification. Many of the objectives are implementation-oriented, so understanding why you would choose a particular architecture or configuration is likely to be just as important as remembering individual features. Additional Practice Exams from the Community Besides Salesforce’s official study materials, there is also a very useful community-built resource created by Rodrigo Santander and Nobuyuki Watanabe. They built an unofficial practice exam platform specifically to help candidates prepare for the new Salesforce Certified Marketing Cloud Next Consultant certification. What makes this resource especially useful is that you can choose between two different learning modes: Study Mode: See the explanation for each answer as you work through the questions. Test Mode: Simulate a more realistic exam experience without immediately seeing the explanations. The platform includes three separate practice exams with increasing difficulty levels: Practice Exam Part 1 – Easy A good starting point for reviewing the fundamentals and checking your overall understanding of Marketing Cloud Next. Practice Exam Part 2 – Intermediate A more challenging set of questions designed to test your knowledge

The Decision element is one of the main elements in Salesforce Flow, and has been there since the early days of Flow. Flow builders use the Decision element to evaluate conditions and determine which path the flow should follow. It supports both simple and complex logic, making it a key element for controlling a flow’s behavior. In the Winter ’27 release, Salesforce introduced two new elements called Split by Field Value and Split by Date. Although Salesforce introduced them as brand-new elements, they essentially work as simplified versions of the Decision element. They split the flow based on a field value or a date. In this post, we’ll take a closer look at these two new elements and compare them with the traditional Decision element. Split by Field Value Element The Split by Field Value element is essentially a simplified version of the Decision element that splits the flow based on a field value. You simply choose the field that you want to evaluate and then configure the possible field values. Each value that you add creates a separate path in the element. It is important to mention that the Field to Evaluate doesn’t have to be an actual field. You can also select a resource, and the value is not limited to text. If none of the configured values match, the flow follows the Default Path. However, unlike the Decision element, this element does not let you rename the default outcome. Moreover, you can easily clone existing paths using the Clone button. Split by Date Element The Split by Date element works similarly to Split by Field Value. It is essentially a simplified version of the Decision element that splits the flow based on a date value. When you add a new path, you can choose from four date operators: Before, Between, After, and Is On. For the comparison value, you can either enter a specific date or select a resource. Unlike the Split by Field Value element, you don’t need to select a single resource to evaluate at the beginning. Each path can evaluate a different date value. Split Elements vs. Decision: Which One Should You Use? The Split by Field Value and Split by Date elements work best for simple scenarios. Unlike the Decision element, they don’t let you define multiple criteria within the same path. Even if your use case starts simple, you may quickly reach this limitation. For example, if you want to add another condition that doesn’t depend on the same field, you cannot add it to the existing path. In that case, you need to switch to the Decision element. Keep these limitations in mind when choosing the right element for your flow. The post Split by Field Value and Split by Date Elements in Salesforce Flow appeared first on Salesforce Time.

**Written by Dorian Sabitov and Antonina Kharchenko After customer identity has been resolved, and the integration appears functional, and invoices and line items are created without errors. The issue then becomes accounting alignment. Line items may sync successfully, but revenue accounts, tax behavior, or classification mappings in QuickBooks may not reflect the intended logic from Salesforce. What the Problem Looks Like in Practice Teams typically notice these problems indirectly via finance reviews rather than sync logs: Invoices post to QuickBooks, but revenue lands in the wrong GL account. Tax rules configured in QuickBooks may not align with the assumptions made during quoting, leading to invoice totals that differ. Discounts, bundles, or usage-based charges appear flattened or misrepresented. Finance teams manually correct line items after the sync, which breaks traceability. From a Salesforce perspective, everything looks fine. From an accounting perspective, the invoice is structurally incorrect. Why It Happens This problem exists because Salesforce Products and QuickBooks Items are not equivalent concepts, even though integrations often treat them as such. In Salesforce: Products are sales abstractions Pricing logic may include bundles, proration, usage, amendments, or CPQ rules Line items are optimized for quoting and pipeline management In QuickBooks: Items are accounting constructs Each item is tied to revenue accounts, tax rules, and reporting behavior Consistency and auditability matter more than pricing flexibility Sync logic frequently pushes Salesforce line items downstream without preserving the accounting intent behind them. When that happens, QuickBooks accepts the data but interprets it differently from what Salesforce expects. This mismatch becomes more pronounced over time as product catalogs evolve, pricing logic changes, tax configurations are updated, or historical products are reused in new contexts. Because these invoices technically sync, the issue is rarely caught early. Instead, teams encounter: revenue misclassification discovered during close inconsistent tax reporting across periods dashboards that reconcile at a summary level but break down at the line-item level or dashboards that do not reconcile at all, forcing finance teams into manual spreadsheet reconciliation At that point, the sync is not broken; it is financially unreliable, which is harder to diagnose and harder to fix. How to Fix It Stable line-item sync requires a clear ownership model: Salesforce defines what was sold (products, pricing structure, quantities), while QuickBooks defines how those transactions post to revenue and tax accounts. Because the data models differ, the integration must explicitly translate Salesforce Products into QuickBooks Items. 1) Map Salesforce Products to specific QuickBooks Items using a stable identifier. QuickBooks Items define income accounts and tax treatment. The integration should store and use the corresponding QuickBooks Item ID (ItemRef) when creating invoice lines, rather than relying on product names or SKU alone. This ensures QuickBooks applies its configured accounting rules and prevents drift as product catalogs change. 2) Validate in Salesforce or middleware before calling QuickBooks.  Before creating the invoice in QuickBooks, validate upstream (in Salesforce logic such as Apex/Flow, or in middleware) that the mapped QuickBooks Item is resolvable, currency assumptions match the target company, and the required tax setup is present. Treat QuickBooks validation errors as a backstop, not as your primary control. 3) Normalize pricing before posting. Salesforce pricing structures such as bundles, proration, or complex discounts do not always translate cleanly into QuickBooks line items. Because each QuickBooks invoice line must reference a single Item with defined accounting behavior, decide upfront how complex pricing will be represented, so the invoice contains consistent, postable lines. Line-item sync problems are not data-type issues. They represent semantic drift. When QuickBooks integration with Salesforce assumes that a synced value carries the same meaning in both systems, it silently accumulates accounting risk. The longer the sync runs, the harder it is to correct. Why This Matters in Production Line-item issues are harder to detect than identity or connectivity problems because transactions continue to sync successfully. The failure is not technical but financial: invoices are posted, but the accounting intent is lost. In the next post, we examine how failures in partial data syncs and retry behavior introduce a different class of risk, where transactions are duplicated or corrupted during recovery attempts. Antonina Kharchenko is a Salesforce Admin with six certifications and a 2-Star Ranger on Trailhead. She works with Salesforce systems, automation, and process improvement, and enjoys turning her day-to-day experience into useful articles. Dorian Sabitov is a 4x Certified Salesforce Administrator and Developer with extensive experience in customizing Salesforce to the client’s needs. He started his journey in IT as a CRM admin and kept his focus on the Salesforce ecosystem. He loves exploring new integrations in Salesforce and spotting alternative ways to optimize business processes inside the CRM. He is currently working as a full-time Salesforce developer and contributing content to the SFApps.info educational portal.