Power BI Data Readiness Assessment

Executive position

The assessed landscape provides a sound base for the stated core finance requirements, but it does not yet support the full reporting ambition.

Across 45 scope-bearing requirement instances:

The counts are requirement instances, not unique data objects. Some common finance objects appear in more than one workbook.

The position differs materially by reporting area:

• Core finance structures and measures are substantially present. One client workbook is fully matched, with all 14 declared requirements found. A second is largely ready, with 10 of 13 requirements found.
• Three finance enrichments remain open: cost centre, transaction currency and account group.
• The customer and sales reporting requirement is not supported by the assessed landscape. None of its eight requirements was found.
• The month-end adjustment control is only partially supported. Source availability is indicated for journal source and approval status, but the completed control, exception measure, entry-level traceability and acceptance evidence are not established.

This supports a targeted funding decision rather than treating the estate as uniformly ready or uniformly deficient. The evidence points to three distinct bodies of work: closing the limited finance-model gaps, establishing the customer and sales capability if it remains in scope, and designing and delivering the month-end close control.

Important assessment boundary

The comparison is against the current gold wiki, which is limited to what the Power BI audit could access. “Not found” means not found within that discovered landscape. It does not prove that the data or capability does not exist elsewhere in the client’s estate. Before committing to build missing items, the relevant source and system owners should therefore confirm whether they already exist outside the assessed boundary.

Status of confirmation

None of the 45 scope-bearing findings has an explicit consultant decision recorded against it.

The findings below are therefore suitable for planning and scope definition, but they are not yet agreed item-by-item facts. Two of the partial findings are specifically marked as weaker assumptions.

Readiness by assessed input

What is already there

The assessed landscape contains the principal objects requested for the core finance model:

• account key, name and type;
• calendar date, date key and fiscal month;
• ledger links to calendar and account;
• signed ledger amount;
• adjustment indicator; and
• the requested Actual Net, Actual Revenue, Budget Net and Net Variance measures.

The fully matched finance workbook contains 14 direct matches. The largely ready finance workbook contains 10 direct matches.

This is meaningful coverage, but it is structural coverage rather than proof of production correctness. The available evidence did not establish ordinary cell values, source reconciliation or exact calculation logic for the four observed measures.

Finance gaps requiring targeted work

Three requested finance attributes were not found in the assessed landscape.

These are bounded additions to an otherwise substantially matched finance structure. Their source availability and governance remain to be confirmed before implementation scope can be finalised.

Customer and sales reporting

None of the eight customer and sales requirements was found within the assessed landscape:

• customer identifier;
• customer segment;
• relationship owner;
• sales-order identifier;
• gross order value;
• sales channel;
• Customer Lifetime Value; and
• Gross Order Intake.

The current evidence is finance-ledger oriented. It does not establish customer or sales-order entities, their relationships or the two requested commercial measures.

If this reporting area remains a priority, the work is broader than adding two calculations. It requires confirmation and definition of the underlying customer and sales-order data, including:

• governed customer and order identifiers;
• customer segmentation and ownership history;
• the treatment of order amendments, cancellations and returns;
• sales-channel attribution;
• gross-value and currency rules;
• the event and date that define booked order intake; and
• the approved commercial definition of Customer Lifetime Value.

Leaving this area open means the assessed landscape cannot currently be relied upon to provide the requested customer segmentation, relationship accountability, order-intake reporting or customer-value analysis.

Month-end close adjustment control

The month-end control has a viable starting point, but it is not ready as a completed control.

Partially supported components

Six requirements have some supporting evidence:

Requirements not yet evidenced

The four remaining requirements should not all be treated as the same type of gap:

Until these matters are resolved, the assessed landscape does not establish a close control that finance can use to identify unapproved manual adjustments before pack sign-off, trace exceptions to original entries and reconcile the result for a closed month.

Validation still required for matched finance items

The remaining questions against matched objects are validation activities, not evidence that the objects are missing.

The principal checks are:

• whether Calendar.DateKey follows the required YYYYMMDD encoding and is unique;
• whether fiscal month numbering uses the intended July start;
• whether Calendar.FiscalMonth should be configured not to summarise;
• whether signed amounts follow the intended reporting-currency and sign conventions;
• what operational rule populates the adjustment indicator;
• whether AccountType covers the intended balance-sheet and profit-and-loss classifications; and
• whether the four observed finance measures implement the required calculation definitions and filter behaviour.

Exact calculation expressions for the observed measures were not available. Their names and presence support a match, but do not prove calculation equivalence.

A Management P&L prototype also raised questions about separate Actual Expenses, Budget Revenue and Budget Expenses measures. This was classified as contextual only because those measures were not part of the authoritative register. It creates no current scope obligation and is excluded from all readiness counts.

Funding priorities

The evidence supports the following order of decision and investment:

1. Confirm the working findings. Obtain consultant decisions on the 45 scope-bearing items, prioritising the 21 items currently marked as proposed work and the two weak-signal assumptions.
2. Close the three defined finance gaps. Confirm source availability and governance for cost centre, currency and account grouping before committing to model changes.
3. Decide whether customer and sales reporting remains in scope. If it does, fund source discovery and business-definition work before treating the eight missing requirements as build-ready.
4. Treat the month-end control as a complete control deliverable. Scope its data attributes, entry-level traceability, calculation, report view, late-approval definition, reconciliation and acceptance evidence together.
5. Validate rather than rebuild matched finance objects. The matched objects require targeted semantic and value-level checks; the evidence does not identify them as missing implementation.

No cost or effort estimate was supplied in the evidence bundle, so this assessment does not price these work areas.

Evidence limitations relevant to the decision

The client workbooks document requirements and prototype logic. They do not establish implemented source connectivity, refresh paths or embedded semantic models. Their working records are described as fictional or synthetic.

Ordinary numeric, date and Boolean cell values were deliberately excluded from the evidence. Sample values, amounts, dates, adjustment flags, budgets and formula results therefore could not be reconciled. Stored formulas show intended logic, but do not prove successful recalculation or current results.

Nothing in this assessment asserts that a target solution has been built, deployed or adopted.

Received but not assessed

Three received workbook records were inspected but never compared with current gold. They support no finding in this report and have not been silently included in the readiness position.

These records bear the same filenames as assessed inputs but were separately recorded as not assessed. No conclusion is drawn from their contents.

How this document was produced

The narrative above was written by a language model from a fixed evidence bundle. The figures in this section are written by the application itself and are not model output.

• Document: Data Readiness Report (report_cf34516922387cee8dbad68776604ec8)
• Generated: 2026-07-31T05:01:12.793Z
• Evidence bundle: ev_sha256_a2548f7efa309fbea88ffdcf3ae4482de06fd6d1dbed8780404d6c80f08e9fd6 (fingerprint d1286dc87de686be)
• Prompt: data-readiness-report.md (4c089bb4599f3482)
• Model: codex
• Current gold build: wiki_7c268dbf7f8b78844080cae53f6f0a0e · 11 pages
• Data profile: synthetic

Basis of the items in this document

• Items assessed: 45
• Carrying an explicit consultant decision: 0
• Carried on the model's unreviewed finding: 45, of which 2 rest on a possible or ambiguous match

Where a consultant decision exists it governs the item. Where none exists the model's finding has been carried forward unchanged; such items are the working position, not an agreed one, and may move when reviewed.