Fabric Planning vs Pigment

Fabric Planning vs Pigment: 
why teams plan in Microsoft Fabric

Fabric Planning helps your business users build their own plans on top of your semantic models, inside Microsoft Fabric. Pigment needs its own copy of the data, in a parallel platform with its own business logic, security, storage and integrations to maintain.
  SUMMARY

Why choose Fabric Planning over Pigment?

Your AI agents work with full plan and performance context

Fabric Planning writes budgets, forecasts and scenarios to the same platform that holds your performance data. Fabric data agents have access to granular context on both, including the comments and approvals behind the plan. Pigment's AI agents see your ERP and warehouse data only as far as it has been imported into Pigment and kept current there.
01

Plan where your data lives

Fabric Planning runs inside Microsoft Fabric, on the data your reports already use. Pigment plans on a copy of that data in its own store, separate from your data and reporting platforms. Your team maintains the imports that keep that copy current.
02

Your business users build the plans

In Fabric Planning, finance, operations, sales and HR users build and change their own plans without writing code. On your existing semantic models, implementations typically go live in days. In Pigment, building or changing a model's logic takes a trained modeler with an Editor license, whether the formulas are written by hand or by its Modeler Agent. Pigment says close to 90% of its implementations are led by partners.
03

Reuse your business logic

Fabric Planning builds on the measures your semantic model already defines: what each metric means and how the numbers roll up. Pigment writes the plan's calculations in its own formula language, so a metric your semantic model defines and Pigment recalculates has two definitions to keep in step.
04

Reports keep pace with the plan

Fabric Planning reports on plan data the moment it is entered, through Integrated Intelligence. It also publishes plan measures to your existing semantic model, so Power BI reports show plan next to actuals. Pigment's boards report the plan against a copy of the actuals that is only as current as the last import. Power BI reads Pigment's numbers through export API queries someone sets up for each saved view, as of the last refresh.
05

Your existing security applies

In Fabric Planning, your Entra ID groups and the security on your semantic models apply to every plan, so data-access rules are defined once, on the model. Pigment can sign users in through Entra ID, but groups, roles and data-access rules are set up again inside Pigment, so your team designs and reviews access in a second system.
06

Lower total cost of ownership

Fabric Planning costs less for almost all user mixes, at the estimated Pigment rates in the calculator. Capacity is sized by the sessions people use, so another use case adds cost only when it brings in new users or upgrades sessions for existing users. There is no new contract to sign. Pigment charges a platform fee plus a per-user license for the full term, and each added use case can raise the contract price. Pigment implementations are mostly partner-led, with first-year services at 20% to 50% of the subscription (Vendr).
  AI

Give your AI agents the full context behind every plan

Fabric Planning writes budgets, forecasts and scenarios to Fabric SQL, on the same platform as the rest of your business data. Your Fabric data agents work from the plan, the actuals at full detail and the sales, operations and workforce data around them.

Ask why a region is behind forecast and where to act, and the agent can look past the plan figures to the pipeline and orders behind them. It can also see the comments and approvals that writeback records.

Pigment's AI agents work inside Pigment, on the data loaded into it and the apps connected to it, without the detailed context that stays in your data platform.

Decisions at full detail

Agents see the actuals down to each customer, product and cost center, beside the plan.

The reasons behind the numbers

Comments and approvals recorded with writeback let an agent explain why a forecast changed and who approved it.
  Architecture

Planning on your data platform

Fabric Planning runs on the data your reports and AI already use, inside Microsoft Fabric. It stores budgets, forecasts, scenarios, versions and comments on the same platform. Pigment plans on a copy of the data imported into its own store, and the plan data created there is exported again for reporting in other tools, so your team maintains and manages two platforms.
Microsoft Fabric PlanningPlans and reports where your data already lives
Microsoft FabricYour data platform
Fabric Planning
PlanningBudgets, forecasts and scenarios
IntelligencePlan vs. actual reporting
PowerTableReference data and data apps
Power BIPlan measures in your existing reports
Data agentsAI on plan and actuals at full detail
OneLake and semantic modelsThe same data and business logic for planning, reporting and AI
PigmentPlans on a second copy, kept in step by imports and exports
Source systemsERP, CRM, HR, data warehouse and master data
Pigment
Separate planning platform
BudgetsForecastsScenarios
Copy of your data in Pigment's store
Power BIPlan vs. actual reporting
93%of Fortune 500 companies use Power BI
MICROSOFT FABRIC PLANNING

Planning on your data foundation

Fabric Planning is the planning capability in Microsoft Fabric. It runs on the same OneLake data and semantic models as your reports and AI.
One fewer system to run
Planning runs inside the Microsoft Fabric estate you already manage, govern and audit. There is nothing to install and no new vendor to onboard.
Source systems reach Fabric through mirroring
Data from Snowflake, Azure Databricks, BigQuery, SAP and Oracle reaches OneLake through Fabric mirroring or shortcuts, in place of ETL pipelines you build and maintain yourself.
Planning and BI on one platform
Plans are built on the semantic models you already run, so planning and reporting share the same business logic.
Plans and actuals together for AI
Budgets, forecasts and scenarios are written to Fabric SQL, on the same platform as the detailed actuals. Fabric data agents work with both at full detail, including the comments and approvals behind the plan.
Inherits your sign-in and security
Entra ID sign-in applies to the plan, and the row-level and object-level security you define once on your semantic model applies to it too.
Plan data stays in your region
Plans are stored in the region of the Fabric capacity they run on. A company in Australia or Singapore can keep them in an Australian or Southeast Asian region.
Pigment

Planning in a silo, on a copy of your data

With Pigment, data is imported from your source systems into Pigment's own store for planning, and plan data moves back out through Pigment's API for reporting in other tools.
A separate platform to license and maintain
Pigment hosts it, with its own models, imports and access rules to maintain alongside your data estate.
Ongoing imports and exports
Shared business data, such as actuals from accounting and opportunities from the CRM, is imported into a central Pigment application, which Pigment recommends for every deployment. Plan data is exported through Pigment's API for reporting in other tools, so the numbers there have to be kept in step with Pigment.
Calculations written again
Calculations your semantic models already define are written again as Pigment formulas wherever the plan recalculates them.
Plans sit outside your data platform
Budgets and forecasts sit in Pigment, while the detailed actuals reside in your data platform. Pigment's AI agents miss the granular performance data that stays there, and your data platform's AI agents see only the plan data exported to it.
Separate governance perimeter
Pigment can sign users in through Entra ID and create their accounts from it, but groups, roles and data-access rules are set up again inside Pigment.
Data stored in Europe or the US
Pigment stores customer data in Frankfurt or Oregon, or on a French sovereign cloud.
  Roles & skills

Who builds and maintains the model

In Fabric Planning, the finance, operations, sales and HR users who own each business process build their plans themselves without writing code. The plans run on the semantic model already in place for reporting, maintained as it is today.

In Pigment, models are built and changed by modelers trained in Pigment's own formula language, who hold Editor licenses. They write each calculation by hand or have the Modeler Agent write it. Pigment says close to 90% of its implementations are now led by partners.

In Fabric Planning, a first plan takes minutes once the semantic model is in place. When the business changes, plan owners change the plan themselves.

Budget holders, sales managers and regional leads contribute in Stakeholder sessions: they enter their numbers, create and run scenarios, approve and comment, and draw capacity only in the months they plan. A new use case adds sessions only for new users, on the Fabric capacity already in place or a larger size of it, with no new contract.

Planning remains a business function

Finance, operations, sales and HR users define input templates, drivers, scenarios and approvals themselves. Building a plan does not require writing DAX or SQL.

The data team's work is unchanged

Plans use the warehouse and semantic model already maintained for reporting.

Partners use the Power BI skills they already have

Delivery partners with Power BI experience can implement Fabric Plan on the semantic models and Fabric skills they already work with.

Reduced key-person risk

Plans are built on models the data team already understands, so they do not depend on a few platform specialists.

Versioned with the rest of the estate

Plan artifacts move through Fabric deployment pipelines and Git-based workflows, with export and import of plan definitions and service principal authentication.

Programmatic management

Public APIs create, read, update, delete and list planning artifacts, so other systems can manage plans automatically.
  Modeling

Reuse the business logic in your semantic models

Most planning in Fabric Planning needs no model at all: a planning sheet connects to the semantic model, input columns capture the plan and each team works in its own filtered view. Where the plan needs logic of its own, Fabric Planning offers two model types that business users build without writing DAX, and the difference between them is where that logic lives.
Logic in the semantic model
Measure Model
The Measure Model suits plans with many similar items, such as products, customers or employees, that all follow the same calculations. Typical examples are the pay elements common to every employee, or sales volume and price driving revenue in every region. A Measure Model is usually made up of multiple measures. You build the plan from the measures your data team already maintains in the semantic model, and metrics such as labor cost or sales revenue are laid out across the relevant business hierarchy, such as cost center or region. Microsoft's framing is "extend, don't rebuild," so plan and actuals share one set of definitions, maintained in one place. Planners work across hundreds to many thousands of rows, one for each territory, SKU or employee, all following the same logic. A change for any one item updates the measures that depend on it and rolls up automatically. For example, a change to Brazil's sales volume updates its gross and net revenue, and Latin America's totals with them.
Logic in the plan
Row Model
The Row Model suits plans where each line has its own formula, such as an income statement. It usually has a single measure, such as one named value, set against a P&L structure taken from one dimension, such as the chart of accounts, with versions such as actual, plan and forecast. Any line can drive another: rent can rise with an inflation rate, and interest can be a rate on the debt balance. Planners work with tens to a few hundred account lines and drivers, each carrying its own logic, and plans can run up to 25 years ahead. Because the logic sits in the planning sheet, it works even on a simple semantic model with few or basic measures.
In Pigment, the plan's logic is built inside the planning platform, with each calculation written as a formula in Pigment's own language. Pigment can import figures your semantic model calculates, but not the measure definitions behind them, which a Measure Model builds on directly. The definitions behind reporting and planning are then maintained in two places.

Planning in Tree View

In Fabric Planning, Tree View turns a Measure Model or Row Model into a visual hierarchy a planner works in, from a top-line outcome such as operating profit down to the drivers feeding it. Select a node in the tree and use a slider to simulate a change. You can simulate several drivers together across entities and time periods. Scenarios are created and compared here too. The Optimize solver adjusts several drivers at once toward a goal, such as a profit target, inside minimum and maximum bounds set for each driver.

What happens as the model grows

In Fabric Planning, each active session draws capacity at a fixed rate by session type, so you are not charged by model size. Capacity still has to be sized. It follows the number of active sessions, plus an allowance for Fabric SQL, OneLake and semantic model activity, which depends on the data involved.

Most plans need no more than 1 million cells and start with context filtering. Larger plans add one of the other methods below, chosen by the objective, the shape of the data and the level of detail the team plans at. Methods can be combined.
Default starting point
Context Filtering and pivoting
Connect to the semantic model, select the dimensions to plan on, filter and plan, with saved pivots giving each team its own view of the same detail. It is the fastest and most common method, and sufficient for most planning use cases.
KPI and measure trees
Measure Model
Build the plan from the governed measures and KPIs the semantic model already exposes.
Multi-entity, multi-currency
Dimension Partitioning
This is an extension of the Context Filtering and Pivoting method across multiple sheets. Each country or profit center plans in its own currency and grain. InfoBridge looks up data from these sheets and rolls it up into a company-wide view, while PowerTable handles currency conversion.
Many dimensions
Cube (multi-dimensional)
Each team plans on the dimensions it needs, and values are allocated across the other dimensions by driver weights.
Strategic plannIng
Row Model
It connects operational and financial drivers on a P&L structure, such as the chart of accounts, for income statement, strategic and long-range plans. Drivers and outcomes are identified first, which makes it more involved to build than the Measure Model.
LINE-ITEM GRAIN
PowerTable Integration
Plan at the detailed grain, such as employee, material, customer or line item. PowerTable holds the detailed grain and feeds the summary plan into the planning sheet.
  Reporting

Reporting on plan data

In Fabric Planning, Intelligence sheets report on plan data as soon as it is entered. Fabric Plan also publishes the plan measures to the semantic model for Power BI reports to pick up, alongside the actuals. Variance reports against actuals are built in, in IBCS-certified notation, and a manager can correct the forecast in the same table where the variance appears. Pigment reports on the plan in its own boards, against a second copy of the actuals kept current by imports your team sets up and maintains. For reporting in Power BI, its documented route is its export API: Power BI Desktop's Web connector reads saved Pigment views with a key a Pigment admin creates. Views too large for the export API need a separate pipeline.
MICROSOFT FABRIC PLANNING
1
Enter the projections in the planning sheet.
2
Intelligence sheets read the projections as they are entered and report on them in real time.
3
Writeback lands plan data in Fabric SQL. You can also publish plan measures to the existing semantic model and its dimensions.
4
An event trigger can refresh the model, and Power BI reports and data agents read it directly.
One model, with no manual step between planning and reporting.
Pigment
1
Import source data and actuals into Pigment, on demand or on a schedule.
2
Plan in Pigment, against the imported actuals.
3
Export plan data through Pigment's API: a Power BI query for each saved view, with a key a Pigment admin creates.
4
Refresh Power BI to load the export, so reports are as current as the last import and refresh.
Data is copied in and out, on a schedule or when someone runs the import or refresh.

Writeback is optional

The Intelligence layer reads plan data directly from the planning sheet, so a plan can be reviewed, reported on and compared with actuals without configuring writeback. Writeback publishes plan data to Fabric SQL. It is the right choice where Power BI reporting, downstream workflows or archival and SOX retention depend on the written data.
  Security & governance

Identity, security and audit

Fabric Planning inherits the identity and data security already in place in Microsoft Fabric and adds its own permissions for who can see and change each sheet, measure and scenario.

Pigment can sign users in through Entra ID and create their accounts from it, but Entra ID groups do not carry over. Pigment's groups, roles and data-access rules are set up again inside Pigment, so access design and periodic access reviews cover a second system. Plan data that Pigment exports to Power BI carries the access rights of the admin whose key exports it. Who sees which numbers has to be secured again in Power BI.

Fabric Planning security works in layers: the bottom four in the diagram below come from the platform, and the top two are the planning permissions Fabric Planning adds on top of them.
Fabric Planning security in layers Six layers, each answering a different access question. Home: Home Layout opens each role on its own page. Fabric Planning: plan security roles set sheets, input measures, scenarios, writeback and comments, and PowerTable sets column and row edit rights. Workspace and item: workspace roles, Build to reuse a semantic model and Write to author the Plan item. Semantic model: row-level and object-level security apply to each signed-in user. Source data: SQL and OneLake security protect what the connection can read. Identity: Microsoft Entra ID groups and conditional access policies govern plan access. Also built in: audit trail, sensitivity labels, tenant-level Private Link and governed connections. HOME Role-based landing page Home Layout gives each role its own landing page, listing the sheets that role can see FABRIC PLANNING Sheets and planning actions Plan security roles control access to sheets, input measures, scenarios, writeback and comments; PowerTable controls column and row edit rights WORKSPACE + ITEM Access, reuse and authoring Workspace roles; Build to reuse a semantic model; Write to author the Plan item SEMANTIC MODEL Business rows and objects Row-level and object-level security apply to each signed-in user, defined once on the model SOURCE DATA Database or OneLake SQL and OneLake security protect what the connection can read with single sign-on, they apply to each user IDENTITY Microsoft Entra ID Existing groups and conditional access policies govern plan access ALSO BUILT IN Audit trail Data, security, comments and exports Sensitivity labels Supported, alongside expanded access management Private Link Tenant level Governed connections Connection changes are validated and logged for audit Best practice: manage access with Entra security groups Three role systems: workspace role, session type and Plan security role; an action must pass all three Fabric Planning security in layers Six layers, each answering a different access question. Home: Home Layout opens each role on its own page. Fabric Planning: plan security roles set sheets, input measures, scenarios, writeback and comments, and PowerTable sets column and row edit rights. Workspace and item: workspace roles, Build to reuse a semantic model and Write to author the Plan item. Semantic model: row-level and object-level security apply to each signed-in user. Source data: SQL and OneLake security protect what the connection can read. Identity: Microsoft Entra ID groups and conditional access policies govern plan access. Also built in: audit trail, sensitivity labels, tenant-level Private Link and governed connections. HOME Role-based landing page Home Layout gives each role its own landing page, listing the sheets that role can see FABRIC PLANNING Sheets and planning actions Plan security roles control access to sheets, input measures, scenarios, writeback and comments; PowerTable controls column and row edit rights WORKSPACE + ITEM Access, reuse and authoring Workspace roles; Build to reuse a semantic model; Write to author the Plan item SEMANTIC MODEL Business rows and objects Row-level and object-level security apply to each signed-in user, defined once on the model SOURCE DATA Database or OneLake SQL and OneLake security protect what the connection can read; with SSO they apply to each user IDENTITY Microsoft Entra ID Existing groups and conditional access policies govern plan access ALSO BUILT IN Audit trail Data, security, comments and exports Sensitivity labels Supported, alongside expanded access management Private Link Tenant level Governed connections Connection changes are validated and logged for audit Best practice: manage access with Entra security groups Three role systems: workspace role, session type and Plan security role; an action must pass all three

One directory: Microsoft Entra ID

Plans live in your Fabric tenant, and people sign in to them with Microsoft Entra ID directly. Disabling someone's Entra ID account removes their access to plans.

Model security applies to every plan

Row-level and object-level security defined on the semantic model apply to each signed-in user of the plan. Plan authors need only Read and Build access to the model, plus write access to the plan, so the model's rules apply to them as well.

Permissions for how people contribute

Plan security roles, assigned to people or Entra ID groups, give each sheet and each input or forecast measure read and write, read-only or hidden access. They also decide who sees which scenarios and who can write back or comment. Administrators can also limit who may take Stakeholder or Planner sessions.

Audit and compliance

Audit trails record changes to data, security settings, comments and exports, with revert in PowerTable. Plan artifacts move through Fabric deployment pipelines and Git-based workflows alongside the rest of the Fabric estate, and sensitivity labels and tenant-level Private Link are supported.
  Results

What customers and partners have reported

Microsoft has published these accounts from organizations planning in Microsoft Fabric on its Fabric Updates Blog.
“For us, the biggest benefit of Fabric Planning is speed: it lets us turn the semantic models and trusted business logic we have already invested in into planning applications, rather than rebuilding that foundation in a separate planning platform.”
Patrick Sura, Director Business Intelligence EMEA, Dentsu.
Published by Microsoft in its Fabric Updates Blog.
“What used to take three to four weeks... can now be completed in about a week.”
Rich Zielinski, Director of Finance & Treasury, Canlak, on the organization's forecasting process.
Published by Microsoft in its general availability announcement.
“Fabric Planning lets us tailor forecasting to each area of the business... without forcing everyone into the same methodology.”
Andrew Gundrum, Executive Director, Finance Data and Analytics, Merck.
Published by Microsoft in its Fabric Updates Blog.
“Planning in Microsoft Fabric closes that gap”
Khaled Ahmed, Managing Director, Fabric & EPM Solutions, KINETIQ.
The gap is the space between spreadsheets on one side and planning tools disconnected from the underlying data on the other. Published by Microsoft in its general availability announcement.
Microsoft's general availability announcement states that customers across financial services, retail, manufacturing and healthcare have adopted Fabric Planning.
  Cost & TCO
Fabric Plan removes a second platform to license, integrate and maintain. The calculator compares license and capacity costs, and you can add the integration costs you expect. Enter your own user counts and negotiated Pigment rates to see the comparison for your organization.

Fabric Planning vs Pigment cost calculator

Compare the annual cost of Microsoft Fabric Planning and Pigment.
User group # Users Fabric Planning 1-year reserved price per 30-day session † Pigment Net rate you pay · $/user/yr
Planners / Admins Build & maintain models Planner
$90.67/session
Editor
$
Stakeholders / Contributors Enter values, approve, comment Stakeholder
$17.98/session
Contributor
$
Read-only / Viewers Consume plans & reports Viewer
$3.91/session
Explorer
$
Licenses or capacity, annual 210 $60,032
F64 capacity, 1-year reserved, about 80% used
$225,000

† Fabric Planning runs on Fabric capacity. A session runs for 30 days from the first time a user engages with a plan, and each session type draws capacity at a fixed rate. A user who does not engage in a month starts no session and draws no capacity. The calculator prices the smallest Fabric capacity that holds a session for every user entered plus the allowance below, at the 1-year reserved East US price. The session prices are shown for reference. CU consumption by session type is set by Microsoft and is the same in every region, while the price per CU varies by region: see the capacity calculator.

Pigment rates are estimates. Enter the net rate you pay per seat, after discounts.

The Pigment figures are USD list prices and have not been converted to USD.

Additional cost lines (enter the costs you actually incur)
Line item Fabric Planning Pigment
Platform / workspace fee optional Pigment's annual platform fee, charged on top of licenses. The default is community's estimate; replace it with your quote.
$
$
Other resources / infra overhead Fabric: an allowance for Fabric SQL, OneLake and Power BI XMLA consumption, entered as a % of session load; the capacity is sized to hold it. The competitor: integration and hosting resources, entered as an annual amount.
%
$
Premium support optional Vendor premium / enterprise support tier, if quoted separately.
$
$
Add-ons optional Pigment prices by the use cases in scope, such as workforce or sales planning, so each added use case can raise the contract. Enter any amount quoted. Fabric Planning includes PowerTable and Intelligence at no additional cost.
$
$
Annual cost, Year 1 $60,032 $265,000
Fabric Planning / yr
$60,032
on F64 1-year reserved capacity
Pigment / yr
$265,000
licenses and fees
77%
lower annual cost · $204,968/yr saved

Microsoft Fabric reserved capacity

Fabric Planning runs on a Microsoft Fabric capacity (an F SKU). Reserved for a year, capacity costs about 41% less than pay-as-you-go, and part of a capacity can be reserved with the rest on pay-as-you-go. Fabric capacity is MACC-eligible where applicable, so it can draw down Azure spend your organization has already committed. The calculator prices the capacity the sessions and the allowance need, for a like-for-like comparison. Lumel's sizing guidance adds headroom for growth by giving planning its own capacity and keeping sessions under 75% of it. Whatever the sessions do not consume stays available to other Fabric workloads.

The Pigment figures are USD list prices and have not been converted to USD.

  Implementation & Training

One-time cost & time to value

Fabric Plan builds on your existing semantic models, so there is no parallel data model or ETL to stand up. A planner can build a first plan in minutes and implementations typically go live in days. Pigment implementations are mostly led by partners. Independent benchmarks put first-year services at 20% to 50% of the first-year subscription (Vendr) or $50,000 to $150,000 for a standard FP&A deployment (CFO Shortlist).
One-time / annual increase Fabric Planning Pigment
Training One-time enablement for your team.
$
$
Implementation time Weeks to go live, which drives time to value.
wk
wk
Implementation cost One-time services / build cost.
$
$
Annual price increase (CPI / renewal) Applied to Years 2 to 3 of the TCO. Enter the escalator in your contract.
%
%

Cumulative 3-year TCO

How each platform's 3-year total cost is built up
One-time setup Year 1 Year 2 Year 3

3-year breakdown

Total cost of ownership over 36 months
Component Fabric Planning Pigment
Annual cost (3 yrs, incl. uplift) $180,096 $860,296
Training (one-time) $12,000 $40,000
Implementation (one-time) $40,000 $100,000
3-year TCO $232,096 $1,000,296
$768K saved
3-year TCO savings · 77% lower
One-time Fabric cost covered by 4 months of run-rate savings · 10 weeks faster to value

Years 2 to 3 apply the renewal uplift: Year n = Year 1 × (1 + uplift)^(n−1). Fabric reserved defaults to 0% (flat annual commitment); Pigment defaults to 8%.

The Pigment figures are USD list prices and have not been converted to USD.

Assumptions & methodology

This page offers a directional comparison, not a quote. Figures are illustrative under the stated assumptions; actual costs depend on contract terms, region, session-type mix and utilization.

  • User mapping. Counts are entered once and applied to both platforms: Planner session ↔ Pigment Editor; Stakeholder session ↔ Pigment Contributor; Viewer session ↔ Pigment Explorer. People who create scenarios need Pigment Editor licenses, and people who only comment can be Pigment Explorers.
  • Fabric Planning uses 1-year reserved capacity pricing, the right basis for an annual TCO, at about 41% below pay-as-you-go. A session runs for 30 days. Each session type draws capacity at a fixed rate set by Microsoft and the same in every region: Planner 1.16, Stakeholder 0.23 and Viewer 0.05 sustained CU (CU per second). That is 847, 168 and 37 CU-hours over a 730-hour month. The calculator assumes a session for every user every month and adds the allowance for Fabric SQL, OneLake and Power BI XMLA (30% by default, adjustable). It then prices the smallest Fabric capacity that holds the total, at the 1-year reserved East US price: an F64, for example, is $60,032 a year. Capacity comes in fixed sizes that double in price at each step, so the Fabric figure moves in steps. F4 is the smallest size used, as the realistic floor for production. Lumel's sizing guidance adds headroom for growth and other workloads by keeping sessions under 75% of the capacity; the capacity calculator lists this among its best practices. The per-session prices shown (Planner $90.67, Stakeholder $17.98 and Viewer $3.91, reserved) are the East US figures it publishes. The price per CU varies by region.
  • Pigment figures are estimates, not contracted rates. Pigment publishes no price list (CFO Shortlist). The license and platform-fee defaults are Lumel's estimates, entered as the net rate you would actually pay, after your negotiated volume and term discount. There is no separate discount line to double-count. CFO Shortlist's pricing report puts typical mid-market Pigment contracts at $200,000 to $400,000 a year for 100 to 200 users. At the default mix of 210 users the Pigment total is $265,000 a year, in line with that range. The training and implementation defaults sit inside the same report's ranges: $20,000 to $60,000 for training and change management and $50,000 to $150,000 for a standard FP&A implementation. Vendr puts services at 20% to 50% of the first-year subscription. The 8% annual increase is the low end of the 8% to 12% escalation CFO Shortlist reports. Added use cases default to $0, so enter any amount quoted. Replace every default with your actual quote.
  • One-time costs and time to value. The Fabric Planning training, implementation cost and implementation time defaults are Lumel's estimates. The 12-week Pigment default is close to the median go-live time in Pigment's customer stories that state one.
  • Multi-year. The CPI or renewal uplift is applied to Years 2 and 3 of the TCO. Fabric reserved capacity defaults to flat (0%); adjust as needed.
  • Not modeled: the in-house team or partner needed to build and maintain Pigment models and the pipelines that move data into and out of Pigment. AI credits beyond the allowance included in a Pigment plan are not modeled either. On the Fabric side, the calculator assumes an organization that already runs Power BI on Fabric. It does not model the data team's work on the semantic model, the Power BI licenses you hold today or the capacity that data agents and other Fabric workloads use. In months when users do not engage with a plan, they start no session and draw no capacity, which leaves room for other users and workloads on the same capacity.
  FAQ

Fabric Planning vs Pigment: frequently asked questions

Overview
The difference is who builds the plan and where the data lives. Fabric Planning is built for the business users who own the plan: they build and change their own plans without code. Pigment models are built by trained modelers with Editor licenses, in Pigment's own formula language, by hand or through its AI agent. Fabric Planning also runs inside Microsoft Fabric, on the data and semantic models you already have. Plan data stays in your data platform for reports and apps to use. Pigment plans on its own copy of the data, imported for planning and exported again for reporting in other tools.

There is nothing to install and no new vendor contract. Fabric Planning is Microsoft's own planning capability in Microsoft Fabric, and a Fabric administrator turns it on in tenant settings. Plans run on Fabric capacity, and any extra capacity you need comes through your existing Azure subscription. Pigment is a separate platform with its own contract, sold directly or through Google Cloud Marketplace.

Yes. Fabric Planning runs on your Fabric data and semantic models, separately from Pigment, so you can start your next planning use case in Fabric Planning while your Pigment models keep running. There is no new contract to negotiate, and a first plan takes minutes once the semantic model is in place.
Building and running plans
The planning model is built by the finance, operations, sales and HR users who own the plan, in a no-code surface. Defining input templates, drivers, scenarios and approvals does not require writing DAX or SQL. The underlying warehouse and semantic model continue to be maintained by the data team as part of its existing reporting work, so for many use cases no third group of platform-specific model builders is required. Delivery partners with Power BI experience can implement Fabric Planning on the semantic models and Fabric skills they already work with.

A first plan takes minutes once the semantic model is in place, and implementations typically go live in days. Plan owners make later changes themselves, so they do not wait for a modeling team.

They work in the plan itself, in Stakeholder sessions: they enter their numbers, run scenarios, approve and comment. Plan security roles decide which sheets and measures each person can see and change.

Yes. Fabric Planning covers sales planning (quotas by territory, with top-down targets and bottom-up input), S&OP and demand planning (statistical forecasts with inventory and supply drivers) and workforce planning (headcount at employee level), alongside FP&A. Teams use it for standard processes such as the annual budget and rolling forecasts, and for ad-hoc what-if questions between cycles, all on the same governed data. Statutory consolidation, with intercompany eliminations and currency translation adjustments, stays in your consolidation system.

No. In Fabric Planning, a Measure Model builds the plan directly from the measures and KPIs your semantic model already exposes. The DAX logic your data team maintains is reused as it stands. Logic that belongs to the plan itself, such as a balance carried from one period to the next, is written as Planning Formulas inside the plan, with no change to the semantic model. In Pigment, calculations your semantic model already defines are written again as formulas in Pigment's own language wherever the plan recalculates them.

Yes. With the Native Planning Engine, a single planning model handles about 12 million cells for one user, and Cube extends a plan to as many as 100 million. PowerTable plans at row level, up to 1 billion rows, with the ceiling set by the capacity behind the database. Most plans need no more than 1 million cells, well inside these limits. When many people plan at once, performance follows the size of your Fabric capacity and the design of the semantic model underneath. Capacity can be scaled up as usage grows. Our article on planning at scale covers the range of each method.

Business users maintain it in PowerTable, an Excel-like grid that needs no code. It covers new products, cost centers, mappings, exchange rates and other forward-looking master and reference data, with approvals, change history and an audit log. It is stored in a Fabric SQL table, so plans, Power BI reports and other applications inside and outside Fabric can all use it. In Pigment, lists live inside its applications, and systems outside Pigment reach them through Pigment's APIs.
Data, AI and security
Yes. Pigment can import from a Fabric data warehouse, through a service account, into its own copy, refreshed on a schedule or on demand. Fabric Planning plans on the data where it already is, in OneLake and the semantic models built on it, so there is no second copy to keep current.

In Fabric Planning, the plan measures are published to the existing semantic model for Power BI reports to pick up, alongside the dimensions already maintained for reporting. Plan data becomes part of the reporting estate without recreating dimensions or standing up a second model. Within Fabric Planning itself, Intelligence sheets report on plan data as it is entered, including variances against actuals. Pigment data reaches Power BI through Pigment's export API: a query in Power BI Desktop reads each saved Pigment view with a key a Pigment admin creates. Views too large for the export API need a separate pipeline.

Fabric Planning does, for organizations whose data is in Microsoft Fabric. Budgets, forecasts and scenarios are written to the same platform where the detailed performance data resides, with the same definitions and security. Plan data is ready for AI where it is, and Fabric data agents work with the plan and the actuals together at full detail, including the comments and approvals that writeback records. Pigment's AI agents, and outside AI tools that connect through Pigment's MCP server, work with the data held in Pigment, plus attached files and connected apps. The ERP and warehouse data they can draw on is whatever has been imported and kept current there.

Fabric Planning uses the security you already have. Identity comes from Microsoft Entra ID, and row-level and object-level security on the semantic model apply to each signed-in user. Plan security roles set who can see and change each sheet, measure and scenario. Audit trails record changes to data, security settings, comments and exports. Pigment can use Entra ID for sign-in and user provisioning, but its groups, roles and data-access rules are set up again inside Pigment.
Plan data is stored in the region of the Fabric capacity it runs on. A capacity remains in the region where it is created, and the content of its workspaces is stored there. A company in Australia or Singapore can keep its plans in an Australian or Southeast Asian region, subject to capacity availability. Pigment stores customer data in Frankfurt or Oregon, or on a French sovereign cloud, so a company in Australia or Singapore plans on data stored in Europe or the US.
Pricing and cost
Fabric Planning uses activity-based licensing on Microsoft Fabric capacity, reserved or pay-as-you-go. Each user who engages with a plan starts a 30-day session, and each session type (Planner, Stakeholder or Viewer) draws capacity at a fixed rate, so the capacity is sized by the sessions people use. The Fabric Planning capacity calculator recommends a size for your mix.

In Fabric Planning, Planning, PowerTable and Intelligence are all included in one session, with no separate add-ons to buy. Other Fabric resources, such as Fabric SQL, OneLake and semantic model activity, are not part of the session. They use your Fabric capacity like any other workload. Allow about 30% of session load for them, the default in the calculator on this page. Automation jobs also draw capacity per successful run. Pigment's pricing combines a platform fee, licenses by type and the use cases in scope, such as workforce or sales planning.

Fabric Planning licenses them by use. A session starts only when someone engages with a plan and runs for 30 days. A budget holder who contributes during the planning cycle, or an approver who signs off once a quarter, draws capacity only in the months they take part. Contributors and approvers use Stakeholder sessions, which draw about a fifth of the capacity of a Planner session. A reserved capacity is paid for all year, so in the months they do not plan it is free for other Fabric workloads. Capacity added on pay-as-you-go for the planning season can be scaled back afterwards. Pigment licenses each user for the license term, so a contributor who works one month a year costs the same as one who works every month. A Pigment user who creates scenarios needs an Editor license.

For almost all user mixes, yes, at the estimated Pigment rates on this page. The savings come from removing a parallel planning platform and the per-user licenses to run it. Actual figures depend on your user counts, rates and contract terms; the calculator gives a directional estimate.

Size your Fabric capacity

The Fabric Planning capacity calculator recommends the Fabric SKU for your mix of Planner, Stakeholder and Viewer sessions.
Microsoft, Azure, Copilot, Excel, Fabric IQ, Microsoft Entra, Microsoft Fabric, Microsoft Teams, OneLake, Outlook, Power Automate, Power BI and SharePoint are trademarks of the Microsoft group of companies. IBCS is a registered trademark of IBCS Association. All other trademarks are the property of their respective owners.

This community is operated independently by Lumel. Nothing on this site constitutes a representation or warranty by Microsoft.