Included with every capability
The agent-ready foundation under every capability: context, business rules and trusted data in one governed model.
See Data StudioThis page
One governed model, every way in: ready-made and self-service Power BI, AI Studio and any AI agent over MCP, with Microsoft Copilot in build.
A digital workforce you can audit, working on your own governed model. Included, not an add-on.
See AI StudioWhy it matters
Most platforms make you choose their reporting tool, then ask you to trust that it matches everything else. Analytics works the other way round. The hard work happens once, in the governed model built in Data Studio: every cost loaded, allocated and signed off. After that, how you read it is up to you.
Ready-made Power BI report sets cover the CFO, the CIO, cloud, application TCO, the bill and variance from day one. Business-unit owners and analysts copy them and shape their own, self-service, on the same governed numbers. A CFO asks AI Studio a question in plain language. Your own AI agents query the model over MCP to build the answers and views you need, and a route inside Microsoft Copilot is in build with Microsoft. Your data team pulls the model into its own warehouse. Every route reads the same allocated numbers, so the answer never depends on where it was asked, and any figure can be followed back through every allocation to the ledger line, the invoice or the metered event underneath.
Read forwards, the model tells you what technology cost. Read backwards, it tells you which business units you’re funding a vendor for, and whether another vendor is already doing the same job.
One model, every way in
Ready-made reports give the whole organisation the same starting point. From there, people copy and shape their own in Power BI, ask AI Studio, or build a view through any AI agent over MCP, and a route inside Microsoft Copilot is in build. Every route reads one governed model, so the number is the same everywhere.
Wherever the question is asked
Everyone, from day one
Start from the shipped reports, then copy and shape your own. Filters carry over into the board pack.
Our Assistants and Specialists
Plain-language questions, and analysis sent on a schedule.
Our agents, in Microsoft 365
The governed answer, where your people already work.
In build with MicrosoftClaude, ChatGPT or your own agents
Your agents query the model to build the answers and views you need.
Your data team
Feed source data in through the same checks and sign-off, and take the governed model out to your own warehouse or BI tool.
Standard objects
Every source lands in the same standard shape, whoever supplied it. That is why reports, AI Studio and any agent understand your data from day one, without a build for each customer, and why every model reads it the same way.
Your cost model
Run the framework you report on, from IT to the whole enterprise. Costs flow from the ledger through your model to the people who consume them and what they deliver.
One governed platform for every source. Data loads from the ledger, CMDB, HR, service management and cloud and AI bills into a Databricks lakehouse, moves from raw, to cleansed, to business-ready, is mapped to standard objects and is signed off before any route reads it. Built for the volume that cloud and AI usage now generate.
What Analytics does
Follow a cost from where it started, through every allocation stage, to the business unit that received it, and from any business unit back to the ledger line, the invoice or the metered event underneath, in the app or in a report. That trace is what keeps a chargeback standing when it’s challenged.
It is Power BI, so anyone with access can copy a shipped report and make it their own, and your BI governance carries straight over. No separate report requests, and no second version of the numbers.
Built-in views for the CFO, the CIO, cloud, the IT portfolio, application TCO, IT finance, business spend, the bill and variance, ready to use rather than built in your first quarter.
The Model, FinOps and Planning Assistants in AI Studio answer plain-language questions against the same model. Every figure comes from the model’s own calculation, never from the language model.
Subscriptions run an assistant’s analysis on a schedule and send the result to the people who need it, without anyone asking. Built, and generally available in release 6.2.
An outbound API hands the governed model, financials and journals included, to your own warehouse or BI tool, so reporting outside the platform starts from the same allocated numbers.
Claude, ChatGPT or the agents your own team builds can query the governed model over MCP, instead of guessing from a warehouse. They ask, the model calculates, nothing is written back. Release 6.1, enabled on request.
In build with Microsoft. MagicOrange agents, reachable from inside Copilot, so people get the governed answer where they already work rather than opening another product.
Vendor and Contract Optimization, and Project and Resource Optimization, are complete report sets, each aimed at a decision rather than a metric.
A fast default view, front-to-back cost flow, a finance-aligned journal view and a full-detail model that traces every allocation stage. Choose the one the question needs.
They’re ledger types on the same model, so comparing actual with budget, or budget with forecast, is a filter rather than a rebuild.
Vendor, cost pool and cost variability each have their own hierarchy, so spend can be cut by supplier, by finance’s own groupings or by how fixed or variable it is.
One layer controls what a person can do with a report. A separate layer controls which data they can see, by cost center, account, structure or dimension. AI Studio answers under the same permissions as the person asking, so a question never opens up data a report wouldn’t.
Cost per user, per server, per gigabyte and per transaction sit beside allocation coverage, forecast accuracy and budget adherence. Analysts define their own measures once, in the model, so every report uses the same definition.
Because it’s Power BI, any report view exports to PowerPoint or PDF with its filters and scenario intact.
Every route shows the data prepared in the model and none can change it, so reporting can never become a back door into the numbers.
The four data models can’t be mixed inside one report. That’s deliberate. Each one answers a different kind of question, and quietly blending them is how a reporting tool ends up with a total nobody can defend.
Questions Analytics answers
What it covers
Analyst view
Forrester calls MagicOrange “a strong fit for organizations seeking modern analytics, speed, and ease of use without unnecessary complexity.”The Forrester Wave™: IT Financial Management Software, Q2 2026
Forrester gave MagicOrange the highest possible score, 5.00, for reporting and dashboarding in The Forrester Wave™: IT Financial Management Software, Q2 2026.
Read the Forrester Wave reprintDisclaimer: Forrester does not endorse any company, product, brand, or service included in its research publications and does not advise any person to select the products or services of any company or brand based on the ratings included in such publications. Information is based on the best available resources. Opinions reflect judgment at the time and are subject to change. This report is part of a broader collection of Forrester resources, including interactive models, frameworks, tools, data, and access to analyst guidance. For more information, read about Forrester’s objectivity at forrester.com/about-us/objectivity.
