Does DISE change the face of the income statement?

No. FASB describes DISE as a disclosure requirement. It centralizes expense detail in the notes but does not require a new income statement format.

What is the first ERP design question finance teams should answer for DISE?

Whether each material expense can be traced from source system and journal entry into both a functional caption and a natural expense category with reviewable evidence.

FASB DISE is not a footnote project. It is an ERP expense data project.

The thesis: DISE is downstream from transaction design.

The official description of the standard is deceptively calm. In its own November 4, 2024 summary, the FASB says DISE is disclosure-only and does not require a new face-of-income-statement presentation. That is true in accounting literature and dangerously incomplete in operations. If the note needs new caption-by-category detail, the systems feeding the note need new structure, governance, and evidence first.

For public business entities, the timing is no longer abstract. FASB’s January 6, 2025 clarification and Deloitte’s March 13, 2026 implementation guide align on the operative dates: annual reporting periods beginning after December 15, 2026, and interim periods within annual reporting periods beginning after December 15, 2027. That means calendar-year filers are designing toward a 2027 annual note and a 2028 quarterly rhythm now, not later.

What changed, exactly.

The new regime requires tabular disaggregation of relevant expense captions that contain prescribed natural expense categories. The official categories highlighted by FASB and KPMG are purchases of inventory, employee compensation, depreciation, intangible asset amortization, and depletion or DD&A where relevant. FASB also added a separate selling-expense disclosure and kept the rule as a note requirement rather than an income-statement redesign.

The technical trap is that none of those outputs exist automatically just because the general ledger balances. Inventory can be tracked one way in the subledger and disclosed another. Payroll may be booked at a department or legal-entity level while the note needs a defensible functional split. Depreciation and amortization may be posted through top-side journals with weak functional classification. Selling expense is especially awkward because it is effectively a controlled, management-defined reporting concept that needs a stable boundary.

What current filings already show

Companies are already telegraphing the implementation work. In its quarter ended March 31, 2025,Pegasystems told investors that ASU 2024-03 expands disclosures around employee compensation and selling expenses and becomes effective for the year ending December 31, 2027. That is the market signal: this is in the plan-of-record window now.

The finance-ops implication: the footnote is only as good as the lineage.

DISE is easiest when the disclosure table is a reproducible output of ERP facts, not a late reporting rewrite.

Deloitte’s 2026 implementation note says many entities will need information that is not currently readily available and may need estimates, IT changes, reporting changes, and control changes. That is exactly the right framing. DISE becomes easy only when every material expense fact can carry both a natural category and a functional meaning before close reporting begins to aggregate it.

This is why the “we can handle it in reporting” instinct is weak. Reporting teams can format a table. They cannot manufacture clean source lineage for payroll allocations, depreciation attribution, inventory basis choices, or selling-expense policy after the fact without creating a control problem.

The data model finance actually needs.

A DISE-ready fact is not merely a posted amount. It needs enough context to support both the note and the reviewer’s challenge later: which relevant caption it belongs to, which required natural category it belongs to, whether it is in selling expense, where it came from, what rule allocated it, and how to retrieve the evidence packet.

Example disclosure dataset row

{
  "entity_id": "us_parent",
  "period": "2027-Q4",
  "relevant_expense_caption": "SG&A",
  "natural_expense_category": "employee_compensation",
  "source_system": "payroll",
  "source_run_id": "payroll_2027_12_31_main",
  "gl_account": "6100",
  "cost_center": "sales_ops",
  "function_code": "selling_general_admin",
  "selling_expense_flag": false,
  "allocation_rule_id": "alloc_payroll_042",
  "amount_minor": 184502331,
  "evidence_ref": "s3://close-packets/2027-q4/payroll/alloc_payroll_042.json"
}

If that structure looks more like a data contract than a spreadsheet, good. That is the point. FASB’s own proposed taxonomy implementation guide already previews caption-and-category-level tagging logic. Once the filing taxonomy thinks at that grain, your ERP and close architecture should too.

Where each required disclosure usually breaks.

Required output	Usual system source	Common failure mode
Purchases of inventory	Inventory subledger, purchasing, manufacturing cost rollforward	No documented choice between cost-incurred and expense-incurred basis; note does not reconcile cleanly to COGS and inventory change.
Employee compensation	Payroll, bonus accruals, stock comp, benefits, labor allocations	Payroll posts as one summary journal with no durable mapping to SG&A, R&D, cost of sales, or selling expense.
Depreciation	Fixed asset subledger and lease-related depreciation outputs	Depreciation is booked top-side with missing functional dimensions or manual reclasses late in close.
Intangible asset amortization	Intangible register, amortization schedule, acquisition accounting support	Amortization lives in an Excel schedule with no link back to caption-level reporting.
Selling expenses	Management policy, chart-of-accounts mapping, cost-center ownership	Definition changes quarter to quarter, or sales and marketing costs are mixed into generic SG&A with no policy trail.

The inventory row is particularly important. KPMG highlights that entities may choose a cost-incurred or expense-incurred basis when disaggregating relevant captions that include inventory amounts, with recasting implications if the basis changes. That is not a note-only choice. It affects how inventory movements, purchases, and income-statement expense roll together in your operational reporting.

Control design for a clean close and a clean filing.

Risk	Control	Evidence
Relevant caption drift	Quarterly recertification of which income-statement captions are in scope and why.	Controller sign-off memo, mapping version, effective date.
Unsupported payroll split	Rule-based allocation of payroll and benefits to functional captions with owner approval.	Payroll run ID, allocation rule ID, approver, posted journal batch.
Inventory basis inconsistency	One approved basis for inventory-related disaggregation with recast rules when changed.	Policy document, inventory reconciliation, prior-period recast evidence.
Selling expense creep	Locked definition of selling expense plus explicit review of included accounts and cost centers.	Annual definition memo, account list, change log, reviewer sign-off.
Top-side disclosure dump	Tie-out control from subledger and journal detail to disclosure table before XBRL tagging.	Disclosure package, exception log, preparer-reviewer checklist.
Opaque estimates	Documented estimation method, threshold, and retrospective true-up review.	Method paper, source extracts, variance analysis, controller approval.
XBRL mismatch	Pre-filing validation of caption/category tags against the approved disclosure table.	Tagged draft, validation report, filing reviewer evidence.

The right owner model is usually split, not centralized. Accounting policy defines the reporting logic. Controllership owns the relevant-caption map and selling-expense policy. ERP or data teams own fact capture and transformation reliability. SEC reporting owns the final disclosure assembly and filing review. Internal audit or SOX teams review whether the process changed the ICFR surface materially.

Implementation checklist before 2027 annual reporting.

Identify every relevant expense caption now, not during the first 2027 annual close.

Choose the disaggregation basis for inventory-related captions and document why it is operationally sustainable.

Break payroll, bonus, benefits, and stock compensation into reusable natural-expense outputs before they hit the disclosure table.

Version-control the selling expense definition, included accounts, excluded accounts, and effective dates.

Require every manual disclosure adjustment to carry an owner, reason code, timestamp, and reversal or permanence decision.

Run a dry disclosure for at least one comparative period so the close team can see where source data is missing.

Build the tie-out from note table to GL and subledgers before XBRL tagging begins.

Keep the evidence chain together: source extract, transformation logic, disclosure output, and reviewer approval.

Run the first dry disclosure on a closed historical quarter or year, not on a live first-close cycle. You want the missing-attribute conversation early, when payroll, inventory, and fixed-assets owners can still change their outputs without the clock pressure of the actual filing timeline.

Failure modes that create expensive cleanup work.

Treating DISE as an SEC-reporting spreadsheet project instead of a source-data project.

Leaving payroll allocations at a department-only grain when the disclosure needs functional-caption support.

Using a vague residual “other” line that becomes a dumping ground for unexplained balances.

Defining selling expense informally in meetings rather than as a controlled reporting policy.

Waiting for quarter-end to reconcile inventory basis choices that should be codified in the subledger design.

Letting XBRL tagging teams infer business logic that finance never documented upstream.

The common pattern behind all six failures is the same: finance discovers at review time that the note needs logic the ERP never stored. Every workaround after that point is slower, harder to review, and worse for audit evidence.

What ERP buyers should ask vendors right now.

Can the system store both functional-caption mapping and natural-expense attributes on the same fact without duplicating journals?

Can payroll, fixed asset, and inventory outputs be traced into the disclosure table with durable run IDs and reviewer evidence?

How are manual disclosure adjustments governed, approved, versioned, and explained later to auditors or the audit committee?

Can selling expense policy changes be versioned by effective date and rerun for comparative periods?

What is the XBRL handoff: a static spreadsheet export, or a tagged, reproducible disclosure dataset?

If the vendor answer is basically “export to Excel and let reporting handle it,” the product is not solving the actual DISE problem. It is moving it later in the process where controls are weaker and remediation is more expensive.

Practical takeaway.

The strongest interpretation of DISE is not that finance needs another disclosure memo. It is that the company needs a reproducible expense dataset with policy-governed classification, source lineage, and reviewer evidence. Teams that build that layer will make the note, the XBRL, and the audit easier. Teams that do not will end up reconciling payroll, inventory, and selling expense logic in quarter-end workbooks under filing pressure.