The Salesforce Manufacturing Cloud data model - now part of Agentforce Manufacturing - is built around three object families that work together: commercial objects for sales agreements, forecasting objects for account-based demand planning, and incentive objects for partner programs. Understanding how these objects connect is the difference between a Manufacturing Cloud implementation that gives you real demand visibility and one that becomes a glorified spreadsheet.
This guide walks through each object family, explains the field-level relationships between sales agreements and account forecasts, covers both standard and Advanced Account Forecasting, and shows how the data model bridges your CRM and ERP data. Whether you're an admin configuring Manufacturing Cloud for the first time or an architect planning an implementation, this is the reference you'll keep coming back to.
The Manufacturing Cloud data model isn't one flat set of objects. It's organized into three distinct families, each serving a different business function. Every object ultimately connects back to the Account object - that's the hub of the entire model.
The beauty of this design is that sales agreement data feeds directly into forecast calculations. When a sales rep enters a product schedule in a sales agreement, those planned quantities automatically roll up into the account forecast. When actual orders come in from your ERP, the forecast compares planned vs actual at every level — by product, by period, by account. That closed loop is what makes Manufacturing Cloud more than just another CRM feature.
Before you create a single sales agreement or forecast, five standard Salesforce objects need to be configured. These aren't Manufacturing Cloud-specific — they're standard platform objects that Manufacturing Cloud depends on.
Account is the root of everything. Every sales agreement, every forecast, every rebate program links back to an Account record. Your account hierarchy matters here: if you sell to a parent company with multiple subsidiaries, your account structure determines how forecasts roll up.
Product2 holds your product catalog. Each product you include in a sales agreement must exist as a Product2 record. If you have 500 SKUs, all 500 need to be in Salesforce before you start building agreements.
Pricebook2 and Price book Entry define your pricing. Sales agreements reference price book entries for each product's negotiated price. You can use a standard price book for list prices and custom price books for customer-specific pricing.
Opportunity connects your sales pipeline to Manufacturing Cloud. Account forecasts can pull pipeline data from open opportunities to give a fuller picture of expected demand — combining committed volumes (from sales agreements) with potential volumes (from opportunities).
Order captures actual purchase orders. When your ERP sends order data into Salesforce, the Order object is where it lands. These actuals are what the forecast compares against planned quantities.
The AP-213 Manufacturing Cloud certification exam specifically tests whether candidates understand that Accounts, Contacts, Products, Opportunities, and Price Books must all be in place before sales agreements can be created. Get this foundation wrong and everything downstream breaks.
The commercial object chain follows a strict master-detail hierarchy: Sales Agreement sits at the top, Sales Agreement Product in the middle, and Sales Agreement Product Schedule at the bottom. Here's what each one does.
This is the parent record representing the entire contract between you and your customer. It has a lookup relationship to Account (not master-detail, so one account can have multiple agreements).
Key fields you'll work with:
A Sales Agreement can have multiple statuses throughout its lifecycle. You'll typically create it in Draft, move it to Activated once both parties agree, and the system sets it to Expired when the end date passes. This lifecycle tracking is something you'd normally manage in a separate contract management tool — having it inside Salesforce means your sales reps see agreement status right on the account page.
This is a master-detail child of Sales Agreement. Each record represents one product line within the agreement. If your customer commits to buying three product families over two years, you'll have three Sales Agreement Product records.
Key fields:
The planned vs actual fields here are aggregates - they sum up from the schedule records below. This gives managers a quick view of whether the customer is on track to meet their contractual commitments without drilling into monthly details.
This is the most granular level - a master-detail child of Sales Agreement Product. Each record represents one time period (typically monthly) for one product within the agreement.
Key fields:
The actual fields on the schedule records are where your ERP integration writes data. When SAP or Oracle processes a purchase order, the integration updates the Actual Quantity and Actual Amount on the corresponding schedule record. This is the bridge between your CRM and your ERP — the planned side comes from your sales team, the actual side comes from your back-office systems.
Let's put this in concrete terms. Say you have a two-year agreement with a chemical distributor for three product lines, with monthly schedules. That's 1 Sales Agreement, 3 Sales Agreement Product records, and 72 Sales Agreement Product Schedule records (3 products x 24 months). At scale — a manufacturer with 200 accounts, each with 5 product lines and 12-month agreements - you're looking at 12,000 schedule records. The data model handles this volume without issues, but it's worth understanding the record count implications for data storage and reporting performance.
This is the connection that most articles skip over. Here's how it actually works.
The Account Forecast object pulls data from three sources: Sales Agreement planned quantities, Order actuals, and Opportunity pipeline amounts. It combines these into a single rolling forecast for each account.
When the system generates an account forecast, it reads the Sales Agreement Product Schedule records for each period and uses the Planned Quantity and Planned Amount as the baseline forecast. Then it overlays actual order data (from the Order object) to show where the customer is ahead or behind plan. Finally, it can layer in open opportunity amounts to show potential upside beyond committed volumes.
The result is a three-dimensional view of demand for each account: what the customer committed to (agreements), what they've actually bought (orders), and what they might buy (opportunities). No other CRM gives you this out of the box for manufacturing. It's the kind of analysis that typically requires a separate S&OP tool or custom-built reports pulling from multiple systems.
This data flow is automatic once you configure account forecasting and link it to your sales agreements. There's no manual export-import step. When a sales rep updates a planned quantity on a schedule record, the account forecast recalculates.
Standard Account Forecasting is the simpler of two forecasting approaches in Manufacturing Cloud. It works well for manufacturers with straightforward forecasting needs — one product dimension, monthly periods, and no multi-enterprise complexity.
The parent forecasting object. One record per account per forecast period.
Breaks the account-level forecast down by product. One record per account per product. This gives product managers visibility into which product lines are tracking to plan and which are falling behind.
Key fields include cumulative planned quantities and amounts, actual quantities and amounts, and opportunity amounts - all aggregated across all forecast periods for that account-product combination.
The most granular standard forecasting object. One record per account, per product, per time period. This is where you see month-by-month planned vs actual comparisons at the product level.
Think of the hierarchy this way: Account Forecast gives you the big picture for the account. Account Product Forecast breaks it down by what they're buying. Account Product Period Forecast tells you when they're buying it and whether it matches the plan.
When a demand planner or account manager needs to override the system-calculated forecast - say they know a customer is about to place a large one-time order that isn't in the sales agreement - they create an adjustment record. This preserves the original forecast while layering on the manual change, so you always have audit traceability.
Advanced Account Forecasting is a significant step up in complexity and capability. It's built on the Data Processing Engine (DPE) and supports multi-dimensional analysis across products, regions, time periods, and custom dimensions.
If Standard Account Forecasting is a spreadsheet, Advanced Account Forecasting is a pivot table that refreshes automatically.
This is the container for an entire forecast scenario. Think of it as a "forecast run" - it defines which accounts are included, what time periods are covered, and which measures and dimensions apply.
A forecast set can include partner accounts (via Advanced Account Forecast Set Partner), allowing multi-enterprise forecasting where distributors or channel partners contribute their demand signals alongside your direct sales data.
The data records within a forecast set. Each fact record represents one intersection of dimensions - for example, "Account X + Product Family Y + Region Z + Q1 2026." Adjustments to fact records are tracked separately via Advanced Account Forecast Fact Adjustment, maintaining the same audit trail as standard forecasting.
Advanced Account Forecast Dimension defines what axes your forecast analyzes - product, territory, time period, custom dimensions. Advanced Account Forecast Measure Definition defines what you're measuring - planned quantity, actual revenue, forecast variance, etc.
Advanced Account Forecast Dimension Source tells the system where to pull dimension values from - whether it's a Salesforce field, a formula, or an external data feed.
Custom calculations that run within the forecast engine. For example, you might create a formula that calculates forecast accuracy percentage: (Actual Quantity divided by Planned Quantity) times 100. These formulas run inside the DPE and update automatically when underlying data changes.
These define the time grain of your forecast. Period records represent individual time buckets (weeks, months, quarters). Period Group records aggregate periods - so you can view data at monthly granularity but roll it up to quarterly or annual views.
A Salesforce Stack Exchange discussion highlighted this exact confusion - users asking "What's the difference between Account Forecast and Advanced Account Forecast?" The short answer: standard forecasting gives you account-by-product-by-period views. Advanced forecasting gives you multi-dimensional analysis with custom calculations, partner collaboration, and DPE-powered automation.
If you're unsure which approach fits your organization, a Salesforce consulting partner like Minuscule Technologies can assess your forecasting complexity and recommend the right path. Their team has configured both standard and advanced forecasting for manufacturers across industries.
Manufacturing Cloud includes a target-setting layer that lets sales leadership define revenue or quantity targets for each account manager, broken down by time period.
(Account Manager Periodic Target Distribution) is the primary object here. It links to a User record (the account manager) and distributes targets across periods. For example, an account manager responsible for $10M in annual revenue might have monthly targets of $700K in January through March, $900K in Q2 (peak season), and so on.
These targets connect to account forecasts so managers can compare their forecast against their goal - are they on track, ahead, or behind? The comparison happens automatically in Manufacturing Cloud reports and dashboards.
The incentive object family handles channel partner programs- rebates, volume bonuses, and market development funds.
Rebate Program is the parent record defining the incentive program. It specifies the program type, qualification criteria, and payout schedule.
Program Tier defines the performance thresholds within a rebate program. For example, Tier 1 might be 0 to 100 units at 2% rebate, Tier 2 is 101 to 500 units at 3.5%, and Tier 3 is 500+ units at 5%.
Program Member is a junction object linking a Rebate Program to an Account. Each participating channel partner gets a Program Member record.
Rebate Payout records the calculated incentive amounts. When a partner hits a tier threshold, the system generates a payout record with the earned rebate amount.
The Winter '26 release added Rebates Explainability -giving partners visibility into exactly how their rebate was calculated, which tiers they qualified for, and what's still outstanding. The Spring '26 release added Stock Rotation Incentives (partners can return aging inventory for rotation credits) and Incentive Budget Management (define and monitor incentive budgets across partners and programs).
The Spring '26 release introduced several new objects and capabilities to the data model.
Inventory allocation across locations is the biggest addition. This gives manufacturers real-time visibility into on-hand stock, allocated quantities, and available-to-promise inventory across multiple warehouses. The data model now supports soft reservations (tentative holds) and concurrent transactions - multiple sales reps allocating from the same inventory pool without double-counting.
Stage management was extended to additional manufacturing objects including Asset, Fleet, and Appraisal. This means you can now track lifecycle stages on these objects the same way you track agreement status - with configurable stage transitions and validation rules.
New Connect REST API resources were added for manufacturing, along with new Apex namespace classes. These give developers programmatic access to manufacturing operations that previously required UI interaction.
For teams working with the Agentforce Manufacturing data model in 2026, keeping up with quarterly releases is essential. Salesforce adds one to two new manufacturing objects or capabilities per release cycle. Minuscule Technologies maintains a release readiness practice that helps clients adopt new features without disrupting existing configurations.
Here's the practical setup sequence. Each step depends on the one before it, so order matters.
Step 1: Verify your foundation objects. Confirm that Account records exist for every customer you'll create agreements for. Confirm that Product2 records exist for every product you'll include. Confirm that at least one Pricebook2 exists with Price book Entry records linking products to prices.
Step 2: Enable Manufacturing Cloud in Setup. Navigate to Setup, search for "Manufacturing Cloud," and enable the feature. This makes the sales agreement objects available in your org.
Step 3: Create a Sales Agreement record. On the customer's account page, click New Sales Agreement. Enter the agreement name, start date, end date, and select the price book. Set the status to Draft.
Step 4: Add Sales Agreement Products. Within the agreement record, add product lines. For each product, specify the planned quantity, planned amount, and unit price. These should reflect the negotiated contract terms.
Step 5: Create Product Schedules. For each product, create monthly (or quarterly) schedule records. Enter the planned quantity and planned amount for each period. This is where your sales team captures their month-by-month demand expectations.
Step 6: Configure ERP integration for actuals. Set up your MuleSoft or middleware integration to write actual order data from SAP, Oracle, or your ERP back into the ActualQuantity and Actual Amount fields on Sales Agreement Product Schedule records. This is the critical bridge between the systems.
Step 7: Activate the agreement. Once the product schedules are complete and both parties have signed off, change the status to Activated. The agreement now feeds into account forecasts automatically.
Once sales agreements are in place, setting up forecasting takes these steps:
Step 1: Enable Account Forecasting in Manufacturing Cloud Settings. Choose your forecasting type - standard or advanced. For first-time implementations, start with standard.
Step 2: Define forecast periods. Choose your time grain (monthly is most common for manufacturing) and set the forecast horizon (typically 12 to 18 months rolling).
Step 3: Configure forecast data sources. Tell the system which data to include: sales agreement planned quantities, actual orders, open opportunities. You can enable or disable each source depending on your forecasting methodology.
Step 4: Set up forecast sets (for advanced only). If using Advanced Account Forecasting, configure your DPE templates, define dimensions and measures, and set up the forecast set with the appropriate partner and account selections.
Step 5: Generate forecasts. Run the forecast generation process. The system reads your sales agreement schedules, overlays order actuals and opportunity data, and creates the forecast records.
Step 6: Configure dashboards and reports. Build reports comparing planned vs actual at the account, product, and period level. Manufacturing Cloud includes pre-built report types for these objects.
Step 7: Train your team. Account managers need to understand how to read their forecasts, create adjustments, and update sales agreement schedules when customer expectations change. This is often the make-or-break step - the data model is only as good as the data your team puts into it.
For organizations navigating this setup for the first time, having a partner who's done it before saves significant time. Minuscule Technologies has configured Manufacturing Cloud forecasting for manufacturers with complex product catalogs and multi-level account hierarchies.
After seeing Manufacturing Cloud implementations across industries, here are the patterns that cause the most problems:
Mistake 1: Skipping the account hierarchy cleanup. If your Account records are messy - duplicate accounts, parent-child relationships not set, subsidiary accounts missing - your forecasts will be unreliable from day one. Clean your account data before enabling Manufacturing Cloud.
Mistake 2: Not integrating ERP actuals. A sales agreement without actual order data is just a wishlist. The entire value of the planned-vs-actual comparison disappears if you don't connect your ERP. Prioritize the integration even if it delays your launch by four weeks.
Mistake 3: Starting with Advanced Account Forecasting. Unless you have a dedicated Salesforce team and clear multi-dimensional forecasting requirements, start with standard. You can migrate to advanced later. Starting with advanced adds weeks of DPE configuration that most organizations don't need initially.
Mistake 4: Over-customizing schedule periods. Monthly schedules work for most manufacturers. Weekly schedules create 4x the record volume without proportional value unless your business truly operates on weekly ordering cycles. Quarterly schedules are too coarse for demand planning.
Mistake 5: Ignoring the training step. We've seen implementations where the data model was configured perfectly but account managers didn't update their forecasts - so leadership had no confidence in the numbers. Invest as much time in training and adoption as you do in configuration.
Mistake 6: Treating actuals as a batch dump. ERP actuals should update in near real-time or at least daily. Monthly batch loads mean your forecast-vs-actual comparison is always stale. Use MuleSoft's event-driven architecture for timely updates.
The Manufacturing Cloud data model is a set of custom objects built on the Salesforce platform that manage sales agreements, account-based forecasting, account manager targets, and rebate programs for manufacturers. It's available in Enterprise, Unlimited, and Developer Editions. As of August 2025, Manufacturing Cloud was rebranded to Agentforce Manufacturing, but the underlying data model objects remain the same.
Sales agreement data flows into forecasts automatically. The Sales Agreement Product Schedule records provide planned quantities and amounts for each product in each time period. The Account Forecast system reads these schedules, overlays actual order data and open opportunity amounts, and generates forecast records that show planned vs actual at the account, product, and period levels.
Standard Account Forecasting supports single-dimension analysis (product by time period) and requires minimal configuration. Advanced Account Forecasting supports multi-dimensional analysis (product by region by time by custom dimensions), uses the Data Processing Engine for calculations, supports partner collaboration, and allows custom formulas and measures. Start with standard unless you have enterprise-scale forecasting needs.
You need Account, Contact, Product2, Pricebook2, and Price book Entry records configured before creating sales agreements. Optionally, Opportunity and Order objects should be set up to feed pipeline and actual data into forecasts. The AP-213 Manufacturing Cloud certification specifically tests this prerequisite knowledge.
For many mid-market manufacturers, yes - at least for the demand-side of S&OP. Manufacturing Cloud's account-based forecasting gives you bottom-up demand visibility that feeds into production planning. However, the supply-side (capacity planning, production scheduling, material requirements) still requires your ERP or a dedicated S&OP tool. Manufacturing Cloud is strongest as the demand signal source that feeds your supply planning process.
The rebate data model follows a program-tier-member-payout hierarchy. Rebate Program defines the incentive program and its rules. Program Tier sets the volume thresholds and rebate percentages. Program Member links participating channel partners (accounts) to the program. Rebate Payout records the calculated incentive amounts when partners meet tier thresholds.
The data model is the foundation of everything Manufacturing Cloud does. Get the object relationships right, integrate your ERP actuals, and train your account managers - and you'll have demand visibility that most manufacturers spend years trying to build in spreadsheets.
Minuscule Technologies helps manufacturers get this foundation right the first time. As a Trusted Salesforce Engineering Partner, they've implemented Manufacturing Cloud data models for organizations ranging from mid-market manufacturers to enterprises with complex multi-entity forecasting needs.
You've seen what's possible. Now, let's make it happen for your business. Whether you need an end-to-end Salesforce solution, a complex integration, or ongoing managed services, our team is ready to deliver.
Schedule a Free Strategic Call
