What AI-assisted financial statement review actually catches

A large portion of a CA's client work involves reviewing financials without conducting a full audit — because the entity is below the threshold, because the review is for a bank loan application, because management wants an independent eye on the numbers before they file.

These engagements are valuable but often done without a formal procedure. The CA looks at the P&L, checks the major line items, compares to last year, asks a few questions. It's judgment-based, experience-based, and largely undocumented — which means if something is missed, there's no paper trail showing what was checked.

What a structured FS review covers

A properly structured financial statement review goes through several layers:

• Analytical review — ratio analysis across years, margin trends, working capital movement, comparison to industry benchmarks where available.
• Disaggregated revenue review — revenue by segment, by product, by geography, checking for unusual concentration or one-off items that inflate the period.
• Cost structure analysis — COGS percentage, key expense ratios, identification of any significant year-on-year shifts that aren't explained by obvious business events.
• Balance sheet quality — receivable aging quality, inventory obsolescence provisions, loan classification and interest accrual, deferred tax position.
• Cash flow quality — reconciliation of reported profit to operating cash flow, capex vs depreciation relationship, borrowing trends.

What tends to be missed without a structured approach

The areas most commonly skipped in a quick review are the ones that require cross-statement analysis. A revenue recognition issue, for instance, shows up not just in the P&L but in the receivables aging (unusually high end-of-year debtors), the cash flow (high profit, low operating cash generation), and potentially in the notes to accounts (contingent liabilities for customer disputes).

Without a structured approach that connects observations across statements, the pattern is easy to miss in isolation. Each individual statement looks unremarkable. Together, they tell a different story.

How FS Review works in DribbleAudit

FS Review takes the client's financial statements — in any standard format — and runs the full analytical review procedure. It identifies anomalies: ratios that are materially outside the prior-year trend, line items that warrant specific inquiry, cash flow to profit discrepancies that need explanation.

The output is a review findings document: each analytical procedure run, the result, and whether it requires follow-up inquiry. Findings that need a management explanation are flagged separately so you can prepare your client meeting questions in advance.

The working paper is structured to SA (Standards on Auditing) requirements, so it serves as proper documentation for the engagement file. If the review is for a loan, the bank-ready summary is available as a separate export.

When FS Review is not enough

FS Review is analytical review — it's not an audit and doesn't substitute for one. It catches things that are analytically unusual. It doesn't substitute for physical stock verification, third-party balance confirmations, or legal review of contracts.

A FS Review finding that suggests a potential material misstatement is a trigger for a deeper conversation with the client — and possibly a scope change to a limited review or full audit engagement. The tool is explicit about this in the output: findings are categorised as 'requires inquiry' or 'may require extended procedure', not as audit conclusions.