Last updated: November 16, 2026
By ZapAI Team
Dynamics 365 Sales forecasting rolls opportunity revenue up a hierarchy of sellers, territories or products, so a sales leader can see committed, best case and pipeline numbers next to quota in one grid. Our Dynamics 365 Sales services team configures these forecasts for mid-market US companies, and this guide covers setup, what premium forecasting adds, Copilot, Power BI and the review cadence we run, checked against Microsoft Learn on November 16, 2026.
Dynamics 365 Sales forecasting is a configurable grid in Sales Enterprise and Sales Premium that rolls up opportunity revenue by forecast category and hierarchy. Premium adds an AI Prediction column. Accuracy depends mostly on clean close dates and categories.
A hypothetical to start. The VP of Sales opens the forecast on the last Monday of the quarter. Committed says $2.1 million. Finance says $1.6 million. Same CRM.
Who’s right? Usually neither, and the gap rarely comes from the forecasting engine. It comes from close dates nobody touched since spring, a deal still sitting in Pipeline after it signed last week, and a manager field that still lists a rep who left in June.
So this is a deployment guide, not a license guide. If you’re still choosing between Professional, Enterprise and Premium, start with our Dynamics 365 Sales pricing breakdown. Here we cover how forecasting works once you own it.
What Is Dynamics 365 Sales Forecasting, Exactly?
It’s a configurable grid. Each row is a person, territory or product in a hierarchy, each column is a bucket of revenue such as Committed or Won, and each cell adds up the opportunities underneath it for one forecast period.
Two versions exist. Every user gets an out-of-the-box forecast for the current month, built on the org chart, with zero setup. Admins can then build configured forecasts with their own hierarchy, periods and columns. Microsoft’s guide to viewing forecasts lists the differences, and they matter more than they look.
| Feature | Out-of-the-box forecast | Configured forecast |
|---|---|---|
| Setup | None | Built by an admin in Forecast configuration |
| Periods | Current month only | Every period you schedule |
| Refresh | Near real time | Every 24 hours, or a manual recalculation |
| Record limit | Up to 50,000 records | No stated limit |
| Quota | Last month’s won amount | Values you upload |
| History | Not saved | Saved for each period |
Look at that quota row again. The default forecast measures this month against last month’s wins. Fine for a demo. Useless for a board meeting.
Licensing, briefly. Microsoft’s Dynamics 365 Sales pricing page lists forecasting under advanced sales force automation, which ships with Sales Enterprise at $105 and Sales Premium at $150 per user per month, paid yearly, and not with Sales Professional at $65. Forecasting also isn’t supported on mobile devices or in Government Community Cloud.
How Do You Configure a Forecast in Dynamics 365 Sales?
Seven steps, in the Sales Hub under App Settings, then Forecast configuration. You’ll need an admin or forecast manager role. Microsoft’s forecast configuration guide, updated May 29, 2026, also ships a Sample Forecast you can activate as is, just to see how the pieces fit.
- Pick a template. Org chart, product or territory.
- Define the model, meaning rollup entity, hierarchy, the relationship between them, top of hierarchy and periods.
- Set permissions, including the forecast owner and who can view it.
- Build the layout from an option set, almost always Forecast Category.
- Add drill-downs so a manager can break a row down by a related table, like account.
- Set advanced options such as multi-currency and prediction factors.
- Activate, then upload quotas from the Excel template.
Rollup entity, hierarchy and the path between them
The rollup entity is what you add up. The hierarchy entity is how you group it. Each template starts with defaults.
| Template | Rollup entity | Hierarchy entity | Relationship path |
|---|---|---|---|
| Org chart | Opportunity | User, through the Manager field | Opportunity owner to user |
| Product | Opportunity Product | Product | Opportunity product to existing product |
| Territory | Opportunity | Territory | Opportunity to account to territory |
That territory row hides a trap. The forecast follows the account’s territory, not anything on the opportunity, so a deal on a misassigned account lands in the wrong manager’s number. Fix the account. Only tables with change tracking turned on appear as rollup options, which is why a custom table can go missing from the list.
Custom rollups work too. Microsoft’s example pairs Opportunity Split with User, which is the route for teams that share credit on one deal across several reps.
Forecast periods
Monthly, quarterly or annual, on a Gregorian or fiscal calendar. One configuration can hold several period types. Microsoft’s example schedules 12 monthly periods for a fiscal year, and you can’t move past this step without at least one period.
Here’s a habit from our projects. We run quarterly periods for the leadership forecast and a separate monthly configuration for frontline managers. Two views, one data set. Dynamics lets several forecasts stay active at once, so there’s no penalty for splitting them.
Columns, adjustments and underlying records
Columns come from an option set. Pick Forecast Category, choose auto-configure, and Dynamics builds Pipeline, Best Case, Committed, Omitted, Won and Lost columns plus a Quota column that you fill by upload. Microsoft’s column configuration guide describes four column types.
| Column type | What it shows | Example |
|---|---|---|
| Rollup | Sum of records that match one option value | Committed revenue |
| Calculated | A formula built from other columns | Won plus Committed plus Best Case |
| Simple | Values you upload from Excel | Quota |
| Hierarchy related | An attribute of the row itself | Manager email or territory name |
Adjustments are where trust is won or lost. An admin marks a column adjustable, and then a seller or manager can override a cell and leave a note. Per Microsoft’s forecast adjustment guide, a manager can change their own row or a direct report’s, the change rolls up the hierarchy, and each override lands in a History tab. Nobody can adjust a row above their own level.
Two rules from that page deserve a printout. Don’t adjust to hide stale opportunities. Fix the records. And don’t zero out a dead deal. Close it as Lost.
One design constraint bites late. You can make a calculated column adjustable or the rollup columns inside it, never both, and adjustable formulas allow only addition and subtraction.
Underlying records are the opportunities behind any cell. Select a value and they appear, and edits made right there recalculate the forecast immediately. Edits made on the Opportunities page wait for the daily refresh or a manual Recalculate data click. That timing gap explains most of the forecast-doesn’t-match-my-pipeline tickets we get.

What Happened to Forecast Snapshots, Deal Flow and Trend Charts?
They’re gone. Microsoft’s list of deprecated Sales features says the snapshot capability was deprecated effective June 30, 2024, and that snapshots, deal flow, trend charts and forecast grid history are no longer available.
Partner blogs from 2020 to 2023 still teach them. Some old Learn links for the deal flow chart now redirect straight to that deprecation list. Expect confusion at kickoff.
Microsoft’s suggested replacement is the msdyn_ForecastApi custom action, which retrieves forecast configurations, periods and instances, and which Microsoft says you can use to store forecast data from time to time as a snapshot. Instance pulls cap at 4,000 records per page. Updates cap at 50 per request.
So build your own history. Our usual pattern is a scheduled Power Automate flow or Fabric pipeline that calls the API every Monday morning, writes each forecast row to a snapshot table, and feeds a Power BI page showing how Committed moved week over week. Small build. Big payoff, since week-over-week movement is the earliest warning a forecast gives you, and history you didn’t capture can’t be rebuilt later.
What Does Premium Forecasting Add, and Who Gets It?
Premium forecasting adds a Prediction column. It uses AI models trained on your closed opportunities and current pipeline to estimate what you’ll actually book this period, independent of what sellers chose in Forecast Category.
As of November 2026, Microsoft’s pricing comparison puts predictive forecasting under advanced sales engagement, a Sales Premium capability, while Sales Enterprise gets forecasting itself. Microsoft’s premium forecasting page, updated May 6, 2026, tells you to confirm entitlements in the Dynamics 365 licensing guide and says the feature isn’t available in every region or in Government Community Cloud. Check your tenant before you promise it to a CRO.
Prerequisites look modest on paper.
- Forecasting configured and active.
- More than 10 closed opportunities with actual value, actual close date, estimated value and estimated close date filled in.
- Open opportunities carrying an estimated value and close date.
- Real data. Test records skew the model.
- Opportunity as the rollup entity, with system users, not teams, as owners.
- Predictive opportunity scoring, which one Microsoft page calls recommended and another calls required, so just turn it on.
Ten closed deals is a floor, not a target. Microsoft says more history means better predictions, and our practice opinion is to wait for four full quarters of clean won and lost data before showing the Prediction column to executives. Before that it’s noisy. Noisy numbers on a CRO dashboard get ignored for good.
Once the column is live, the predictive forecasting guide says predictions refresh automatically every seven days and can’t be recalculated on demand. Select a value and you get closed won, predicted from open deals, predicted from new deals that don’t exist yet, and the top five factors behind the number, such as the share of opportunities expected to slip into next period.
That predicted-from-new slice is the part sellers can’t produce by hand. Committed never includes deals nobody has created. The model does.
Should you buy Premium for this alone? Probably not. The $45 per user per month step up also pays for the sales accelerator and its work list, conversation intelligence and full predictive lead and opportunity scoring, and a forecast column is rarely the deciding line item. Bias disclosed, we sell Dynamics 365 projects. We’d still tell a 15-seller team with two years of messy data to fix the data first.
Where Does Copilot Fit in Pipeline and Forecast Work?
On the edges, for now. Copilot in Dynamics 365 Sales answers questions about pipeline data and summarizes opportunities. It doesn’t build the forecast.
What’s documented? Microsoft’s guide to asking Copilot questions lists sales questions like your pipeline with deals closing in the next 30 days, your average deal cycle, last quarter’s lead conversion rate and open deals above a set size. Copilot checks Dataverse first. Admins can add your own terms to a glossary, which helps when your team says commit and the field says Committed.
One Microsoft Learn training module also describes Copilot pipeline health summaries inside the forecast view. We couldn’t match that to a product documentation page as of November 2026. Treat it as unconfirmed until you see it in your own tenant.
Copilot comes with Sales Enterprise and Premium, not Professional. Three more facts shape planning.
- Microsoft’s Copilot in Dynamics 365 Sales FAQ says the Sales agent in Microsoft 365 Copilot will become the preferred way to get Copilot help in Dynamics 365 Sales, and custom summary fields need reconfiguring when you move.
- The Sales Opportunity Agent researches deals and flags risk using the predictive opportunity scoring model. It needs a Copilot Studio license and draws on consumption capacity.
- Sales Close Agent is done. New instances stopped September 30, 2026, and existing ones were removed October 30, 2026.
Our read? Use Copilot to prep for the forecast call, not to produce the forecast. A manager who asks for open deals closing this month gets a list in seconds and walks into the meeting with questions ready. Our Copilot in Dynamics 365 use cases cover where else it pays off.
What Data Hygiene Does a Trustworthy Forecast Need?
Five fields and one habit. Dynamics forecasting is arithmetic over opportunity records, so every flaw in those records shows up in the grid with a dollar sign in front of it.
| Field or setting | Why it matters | Our default rule |
|---|---|---|
| Forecast Category | Decides which column a deal lands in | Required on save, reviewed weekly |
| Est. Close Date | Decides which period a deal lands in | No past dates on open deals |
| Est. Revenue | The amount that rolls up | Required before Best Case |
| Owner and Manager | Decide whose row the deal hits | Manager field audited monthly |
| Actual Revenue and Actual Close Date | Drive Won totals and premium training data | Set only by closing the opportunity |
The habit is closing deals properly. Microsoft’s forecast category guide warns against picking Won or Lost by hand, and its forecasting FAQ explains why, since a forecast only counts Won when actual revenue and actual close date exist, so a manual Won can make a deal vanish from the grid. Use the Close Opportunity dialog. Every time.
Stage and category aren’t the same thing. Sales stage tracks the process. Forecast Category tracks the rep’s confidence. We map them loosely, with a business rule that blocks Committed before the proposal stage, because tying them one to one strips out the judgment that makes the category worth having. Our CRM customization best practices cover that tradeoff in more depth.
Watch for two technical gotchas. If a forecast column uses a field with field-level security, the # Dynamics 365 Sales Forecasting system user needs read access in the field security profile, or the column shows blank, a detail that fits with our Dataverse security roles guide. And if you swap Forecast Category for a custom option set, Microsoft says you need a workflow that keeps it in sync with opportunity status.
Run this list before go-live.
- No open opportunity with a close date in the past.
- Every open deal carries an estimated revenue above zero.
- Manager field set for every active seller.
- Departed reps’ deals reassigned before the quarter starts.
- Test and demo records deleted, or set to Omitted.
- Won and Lost set only by closing the opportunity.

When Should You Use Power BI or Fabric Instead?
Use Power BI alongside the forecast, not instead of it, unless your forecast logic lives outside the CRM. The native grid handles hierarchy, adjustments and quota. Power BI handles history, data from other systems and visuals.
Microsoft nudged things this way itself. Its Sales Analytics and Process Analytics Power BI template apps were deprecated in May 2025, and with snapshots gone, any trend line now needs a report you own.
Connection choice decides how far that report scales. Microsoft’s Dataverse connector documentation puts typical retrieval at roughly 500 rows per second, gives each query a fixed 5-minute timeout, and recommends Azure Synapse Link for bulk extraction. For a 2,000-opportunity pipeline the connector is fine. For five years of deal history plus ERP invoices, use Link to Fabric, which our Dataverse to Fabric routes guide compares.
Bring in Power BI or Fabric when any of these is true.
- Bookings live in Business Central or another ERP, and finance wants forecast versus actual on one page.
- You need week-over-week history.
- Leaders want win rate, slippage and cycle time trends, which the grid doesn’t chart.
- Some executives hold no Dynamics 365 license. Power BI Pro runs $14 per user per month, paid yearly.
- Renewals or recurring revenue aren’t tracked as opportunities.
Our Power BI consultants build these reports on Dynamics data, and if you want the broader argument for doing it at all, read how Power BI improves decisions. Copilot can draft visuals once the model is clean, and our Copilot in Power BI setup guide lists what that takes.
Basic Forecasting vs Premium Forecasting vs a Power BI Pipeline Report
Here’s the side-by-side, as of November 2026. Setup effort is our estimate, not Microsoft’s.
| Factor | Basic forecasting | Premium forecasting | Power BI pipeline report |
|---|---|---|---|
| License | Sales Enterprise or Premium | Sales Premium | Power BI Pro or capacity, plus access to Dynamics data |
| Typical setup effort | Days for a standard org chart | Basic setup plus enough closed history | Weeks for the model and a snapshot pipeline |
| Hierarchy and rollup | Native, by user, territory, product or custom | Native, Opportunity rollup for predictions | You model it |
| Adjustments with history | Yes, with notes | Yes, with notes | Only if you build it |
| AI prediction | No | Prediction column and top five factors | Only if you build it |
| Trend history | Retired since June 2024 | Retired since June 2024 | Yes, from your own snapshot table |
| Data outside Dynamics | No | No | Yes |
| Refresh | Daily, plus manual recalculation | Predictions every seven days | Whatever schedule you set |
| Best for | Weekly commit calls | Challenging seller judgment | History, finance reconciliation, cross-system views |
How Should Sales Leaders Run a Forecast Accuracy Review?
On a fixed rhythm, scored against what actually closed. In a 2020 survey, Gartner found fewer than half of sales leaders and sellers had high confidence in their organization’s forecast accuracy. In our experience confidence follows measurement. Teams that score last quarter’s forecast trust this quarter’s more.
Below is the cadence we set up with clients. It’s ZapAI practice. Not a Microsoft feature.
| When | Who | What to check |
|---|---|---|
| Weekly, Monday | Frontline managers and reps | Past close dates, category changes, Committed deals with no recent activity |
| Weekly, Tuesday | Sales leadership | Committed and Best Case against quota, plus adjustments and their notes |
| Monthly | RevOps and sales leadership | Snapshot trend, slippage rate, Prediction against Committed if you run Premium |
| Quarterly, week two of the next quarter | Sales, RevOps and finance | Accuracy by manager, category definitions, hierarchy cleanup |
Keep the math simple. Take the Committed number from your week 6 snapshot of a 13-week quarter, divide actual closed-won by it, and track that ratio per manager over four quarters. Then repeat it for Committed plus Best Case. Wildly different ratios between managers usually mean different definitions of Committed, not different talent.
Another hypothetical. A 40-seller distributor snapshots Committed at $3.0 million in week 6. The quarter closes at $2.55 million won, which is 85% of that commit. One region lands at 104% and another at 61%, and the 61% region’s managers were marking deals Committed on a verbal yes. Same tool. Different definitions.
Write the definitions down. One page, three categories, an example deal for each. Revisit it every quarter.

Our Decision Rules for Dynamics 365 Forecasting
These rules are ZapAI practice opinions from Dynamics 365 Sales projects, not Microsoft guidance. Bend them to fit your team.
- Under 10 sellers with one manager? The out-of-the-box forecast plus a clean pipeline view is enough.
- Two or more management layers, or quotas by territory, means a configured forecast.
- Turn on adjustments for Committed only at first, because opening every column invites sandbagging nobody can trace.
- Require a note on any adjustment above 10% of the row.
- Hold the Prediction column back until four quarters of clean closed data exist.
- Build the weekly snapshot table in month one.
- Bring Power BI in the day finance asks to reconcile bookings.
- Buy Premium for the bundle. Never for one column.
What Could Change After Ignite 2026?
Copilot and agents, mostly. Microsoft Ignite runs November 17 to 20, 2026, in San Francisco, and the Sales agent’s role inside Dynamics 365 Sales is the item forecasting teams should watch. We’ll refresh this post if forecast or Copilot entitlements change.
Wrapping Up
Think of Dynamics 365 Sales forecasting as a solid grid wrapped around whatever your opportunity data says. Configure the hierarchy and periods to match how leaders actually review numbers, lock down Forecast Category and close dates, turn on adjustments with notes, and build your own snapshot history since Microsoft retired its own. Add Premium’s Prediction column when the history supports it. Add Power BI when finance shows up.
ZapAI is a Microsoft Dynamics 365, Power Platform, Fabric and AI consulting firm in Irving, Texas, working with companies across Dallas-Fort Worth and nationwide. Our Dynamics 365 Sales consultants set up forecasts, data rules and Power BI forecast history for mid-market sales teams, and our wider Microsoft Dynamics 365 practice covers the rest of the stack. Not sure your forecast deserves trust? Book a forecast configuration review, and we’ll score last quarter’s forecast against actuals before recommending anything.
Questions Sales Leaders Ask About Dynamics 365 Forecasting
Does Sales Professional include forecasting?
Not as of November 2026. Microsoft’s pricing page lists forecasting under advanced sales force automation, which comes with Sales Enterprise and Sales Premium. Professional teams typically track pipeline with views and charts instead.
So what does premium forecasting actually add?
A Prediction column, plus the reasons behind it. The model estimates what you’ll close from open deals and from deals not yet created, refreshes every seven days, and shows the top five factors driving the number. It needs Sales Premium and clean closed-deal history.
Can we still get forecast snapshots and the trend chart?
Short answer, not natively. Microsoft retired snapshots, deal flow, trend charts and grid history on June 30, 2024. The workaround is the msdyn_ForecastApi custom action on a weekly schedule, writing to your own table, with Power BI on top. It’s a small build, and we’d start it in month one so the history begins early.
Who can adjust a forecast, and is it tracked?
Sellers and managers, on columns an admin marked adjustable. Managers can change their own row or a direct report’s, never a row above them. Every adjustment and reset sits in a History tab with its note.
How much history does predictive forecasting need?
More than 10 closed opportunities is Microsoft’s minimum, with actual and estimated values and dates filled in. That’s the floor. We wait for about four quarters of clean won and lost data before putting predictions in front of executives, since a noisy number early on can sour leaders on the feature for a long time.
Should the forecast category follow the sales stage automatically?
Wrong question, slightly. Stage tracks process and category tracks confidence, so we keep them separate and add guardrails, like blocking Committed before the proposal stage.
Can Copilot build the forecast for us?
Not today. Copilot in Dynamics 365 Sales answers pipeline questions and summarizes opportunities, and Microsoft says the Sales agent in Microsoft 365 Copilot will become the preferred route. The forecast itself still comes from your opportunity data and configuration.
