MICROSOFT FABRIC ACCOUNTABILITY & VALUE VERIFICATION

Know who consumes shared Fabric capacity, agree what should change, and verify the result.

FabOps turns Microsoft-native usage and billing evidence into a customer-owned accountability process: reconcile the commercial picture, apply the agreed ownership policy, expose confidence and exceptions, assign approved actions and retain the evidence of what happened next.

Book a Fabric Economics Assessment See How the Accountability Close Works

FabOps is in paid validation. Software pricing, recurring cadence and product build priorities are confirmed through real customer evidence rather than public legacy tiers.

  • Native-first: works with Microsoft Fabric evidence rather than replacing it
  • Transparent confidence and exceptions
  • Customer-owned policy and decisions
  • Verified action/outcome record
  • Raw-telemetry target architecture subject to implementation and review evidence

THE CUSTOMER PROBLEM

The bill, the usage and the decision rarely live in the same place.

Microsoft gives customers increasingly capable native evidence for capacity health, usage, billing and allocation. The unresolved management job is usually customer-specific:

  • how the commercial bill and commitments relate to actual usage;
  • which team, cost centre or owner accepts shared consumption;
  • how unallocated and disputed amounts are treated;
  • which interventions are approved under operational guardrails;
  • whether the expected economic or operational value materialised;
  • what evidence Finance, platform leadership and Risk accept at the next review.

FabOps is designed around that accepted process, not around duplicating another telemetry dashboard.

FABRIC ECONOMICS ASSESSMENT

Establish the first defensible accountability baseline before buying a recurring product.

The FEA is a fixed-scope professional-services assessment. It uses the customer’s billing, commitment, native usage and ownership evidence to create the first reconciled view, draft policy, exceptions ledger, intervention plan and verification methods.

Outputs

  • Evidence-quality and Data-Ready report
  • Bill and commitment reconciliation
  • Draft ownership and shared-cost policy
  • Confidence-rated showback
  • Unallocated and disputed amounts
  • Action and verification plan
  • Executive readout with platform and Finance participation
  • Written recurrence decision
Explore the Fabric Economics Assessment

THE ACCOUNTABILITY CLOSE

A recurring management process — not passive dashboard access.

Where customers choose recurrence, FabOps Enterprise supports the customer-run Fabric Accountability Close.

Each close:

  1. reconciles the period’s billing and commitments to native evidence;
  2. applies the current ownership and shared-cost policy;
  3. surfaces low-confidence, unallocated or disputed consumption;
  4. reviews capacity pressure and economic opportunities;
  5. assigns approved actions to named owners;
  6. verifies previous interventions against the pre-agreed method;
  7. produces one accepted platform-and-Finance record.

Initial cadence hypothesis

Quarterly customer-run by default. Monthly only for sufficiently complex enterprise estates at enterprise-custom economics. Every FEA records what cadence the customer would actually adopt.

See the Customer Runbook

ENTERPRISE VS OPERATE

Software and services stay separate.

FabOps Enterprise

Software and evidence system.

Designed to include

  • Policy and ownership model
  • Close periods and source evidence
  • Reconciliation and allocation outputs
  • Confidence, exceptions and disputes
  • Approvals, actions and due dates
  • Verification cards and customer sign-off
  • Historical accountability ledger
  • Approved standard product documentation and support

Designed not to include

  • Substantive human close facilitation
  • Implementation of approved interventions
  • Managed monthly operation
  • Custom consulting
  • Formal audit or assurance opinion

FabOps Operate

Separately contracted professional services.

May include

  • FEA delivery
  • Policy and close setup
  • Bounded facilitation
  • Implementation of approved interventions
  • 90-day verification
  • Defined governance and operating-model work
  • Separately scoped customer success

Commercial boundary. Enterprise and Operate remain separate line items and ledgers. Service work is never used to inflate software ARR or disguise labour as SaaS.

NATIVE-FIRST

Microsoft-native evidence is the starting point, not the enemy.

FabOps is designed to consume sufficient native evidence such as Capacity Metrics, Chargeback, Workspace Monitoring, Cost Management/FOCUS, tags, domains and approved customer ownership data. Where Microsoft improves the evidence layer, FabOps should integrate it and reduce duplication.

FabOps must prove incremental value over the native stack in each paid engagement

  • Customer-specific commercial and rate context
  • Policy and exception treatment
  • Accepted owners and decision rights
  • Accountable actions
  • Pre-agreed verification
  • Historical evidence and sign-off

GOVERNANCE AS GUARDRAIL

A lower-cost answer is not automatically a safe answer.

FabOps retains governance scoring, best-practice evidence and operational context where they change the decision. “Safe to change” means the customer approves an intervention against stated constraints such as:

  • Workload criticality
  • Schedules and seasonality
  • Service levels and headroom
  • Resilience and rollback
  • Growth plans
  • Commitments and reservations
  • Security and regulatory requirements

FabOps never guarantees that an algorithmic recommendation is safe.

VERIFY WHAT CHANGED

Recommendations become valuable only when the outcome is measured and accepted.

Every intervention moves through distinct states:

  1. ObservedA condition exists in the evidence.
  2. AddressableThe condition may be actionable under the customer’s constraints.
  3. ApprovedThe customer accepts the intervention, owner, expected result and measurement method.
  4. RealisedThe result is measured and accepted at the appropriate platform and Finance level.

Verification card

  • Baseline period
  • Evidence sources
  • Expected mechanism
  • Expected result or range
  • Guardrails
  • Comparison window
  • Known confounders
  • Actual result
  • Adverse effects
  • Confidence
  • Customer verdict

Onyx does not unilaterally declare that its own remediation created realised value.

TRUST AND DATA BOUNDARIES

Describe the deployed route, not the desired architecture.

Current FEA route

For the assessment, work is performed in the customer tenant wherever technically and contractually feasible. Any required extract is minimised, approved, restricted, retained for the stated period and deleted under the engagement terms.

Target Enterprise architecture

The target product architecture keeps raw Fabric operational telemetry stored and processed in the customer tenant and provides FabOps only the minimum authorised results required for the service. This target may be described as current only for implementations whose deployed and review evidence substantiates it.

Cross-customer learning. Any contribution to the outcome corpus is separately permissioned, minimised and treated as pseudonymised or de-identified unless genuine anonymity is proven. Raw telemetry does not become the cross-customer asset.

Review the Current Trust Statement

THE LONG-TERM MOAT

The value compounds in the method, ledger and verified outcomes.

FabOps’s intended defensibility is not a temporary dashboard or API gap. It is the accumulation of:

  • Accepted allocation and decision methodology
  • Customer-specific policy and ownership history
  • The accountability ledger
  • Pre-agreed verification and customer verdicts
  • Permissioned intervention and outcome evidence
  • Category trust and distribution

Future outcome comparisons remain feature-off until rights, quality, cohort, privacy and statistical gates pass.

WHO IT IS FOR

A real accountability decision, not a dashboard request.

Strong fit

  • F64+, multi-capacity or demonstrably complex shared estates
  • Multiple teams or domains sharing capacity or budget
  • Recent renewal, reservation, throttling, bill, AI-growth or accountability trigger
  • Platform champion and Finance or FinOps methodology participant
  • Viable evidence and commercial route
  • Buyer values an accepted decision and follow-through

Not a fit today

  • Simple estate with no active trigger
  • Customer only wants another dashboard
  • No evidence or Finance route
  • Guaranteed savings or accounting-grade chargeback is required
  • General multi-cloud FinOps is the job
  • Customer expects free trial or product access before qualification

VALIDATION STATUS

What is proven — and what is still being tested.

Paid validation keeps prior consulting or product customer outcomes, technical environments, paid FEA customers, recurring Enterprise contracts and verified interventions distinct. No outcome is published here until the claims register approves its metric definition, source, period, offer and customer acceptance state.

Current base position

Until paid repeatability is evidenced, FabOps carries £0 in the company’s base wealth model and no public ARR or valuation claim.

What the first paid engagements test

  • Urgency and willingness to pay
  • Reconciliation and allocation method
  • Finance acceptance
  • Recurring-close demand
  • Delivery effort and gross margin
  • Productisation priorities
  • Data-rights feasibility

COMMON QUESTIONS

Before the assessment

Is FabOps another replacement for Microsoft Capacity Metrics or Chargeback?

No. FabOps is designed to use sufficient Microsoft-native evidence and add the customer-specific policy, exceptions, decisions, actions and verification process that must be accepted inside the organisation.

Does FabOps guarantee cost savings?

No. The method separates observed conditions, addressable opportunities, approved actions and realised outcomes. Financial treatment depends on the customer’s commitments, growth, policy, implementation and Finance acceptance.

Does raw data always remain in our tenant today?

The delivery route is described for the specific engagement and current implementation. Tenant-resident raw-data processing is the target Enterprise architecture and is claimed as current only where deployment and review evidence substantiate it.

Is human facilitation included in Enterprise?

Substantive facilitation or implementation is separately scoped through FabOps Operate. Enterprise is designed as the customer-run software and evidence system.

Is there a free trial or public monthly price?

Not during the current validation stage. The commercial entry point is the paid FEA. Recurring product and pricing are confirmed through customer evidence.

Start by establishing the evidence, policy and decision path.

The Fabric Economics Assessment is the current commercial entry point. We qualify the trigger, estate, Finance participation, evidence route and buyer decision before scope is issued.

Book an FEA Qualification Call