Managing audit evidence means handling the several hundred documents an engagement generates so that each one can be found, is known to be current, and is linked to the figures that depend on it. Four things have to work: requests specific enough that what arrives is usable, classification and naming applied on arrival rather than later, version control so that a superseded file cannot be relied on, and a live view of what remains outstanding. Where these fail, the audit work is unaffected in quality but the file becomes undefendable, because nobody can show which document supports which number.
The scale of the problem
A mid-market statutory audit typically involves a trial balance and ledgers, a dozen registers and schedules, twelve monthly GST statements per registration, bank statements and confirmations for every account, supplier and customer confirmations, contracts, board minutes, statutory registers, fixed asset records, stock records, and a sample of transaction-level documents that can run to several hundred invoices on its own.
Most firms manage this with a shared folder and an email thread. That works until a document is superseded, until two people are working on the same section, or until somebody asks in November which version of the stock sheet the valuation was tested against.
Requesting so that what arrives is usable
Most document problems are request problems. A vague request produces whatever the client's accountant happens to generate, and the cost of that lands three weeks later.
- 01Name the document, the format and the periodNot “the purchase data” but “purchase register for 1 April to 31 March, in Excel, one row per invoice line, filtered on voucher date, with the columns listed below”. Include a sample file the first year. The specification pays for itself immediately.
- 02Say what it is forA client who knows the register is being matched against their GST statements will produce it with the GSTIN column populated. A client who does not will send whatever the default report gives.
- 03Separate what you need now from what you need laterA single list of ninety items sent in week one produces ninety items in week six. A first tranche of the eight things that unblock the population reconciliation produces those eight in days.
- 04Name a person, not a companyRequests addressed to the finance department are addressed to nobody. Identify who maintains the accounting system, and route data requests to them directly rather than through the partner's contact.
- 05Make the outstanding list visible to the clientThe single highest-return change available. A client who can see what is outstanding will close most of it without being chased, and the arguments about what was asked for disappear.
Common practice, not a requirement
Some firms gate the start of fieldwork on a minimum set of documents having been received. This is a commercial decision rather than a professional requirement, and it is unpopular, but it prevents the most expensive failure mode in the whole process: testing performed on data that is later replaced.
Classification and naming
A document that cannot be found is evidence that does not exist. Naming has to happen on arrival, by whoever receives it, and has to encode enough to be searched.
| Element | Why | Example |
|---|---|---|
| Entity and period | Files migrate between engagements and years | ABC-FY26 |
| Document type | So a category can be listed without opening anything | purchase-register |
| Sub-period or scope | Monthly statements and multi-registration clients | 2025-11-27AAAPL |
| Version or received date | So the current one is identifiable without opening both | recd-2026-06-14 |
| Source | A client-supplied schedule and a portal download are different evidence | client / portal |
The last row matters more than it looks. A file named so that anyone can see whether it came from the client or was downloaded by the audit team preserves a distinction that goes directly to the reliability of the evidence, and that distinction is otherwise lost the moment the file lands in a folder.
Versions, and why the second file is dangerous
Clients re-run reports. A purchase register sent in May and re-sent in July will differ, because entries were passed in between. The risk is not that the second file is wrong; it is that half the testing was done on the first and half on the second, and the working paper says neither.
- Record, for every file, when it was received, from whom, and the control total and row count. Two registers with the same name and different row counts are then distinguishable in seconds.
- Never overwrite. A superseded file stays, marked superseded, because work already performed refers to it.
- When a file is replaced, identify the work already performed on the old one and decide explicitly whether it must be redone. That decision belongs in the file.
- State in the working paper which file the testing was performed on, by name and received date.
The most damaging version failure is silent: someone helpfully replaces a file in a shared folder with an updated one. The testing does not change, the working paper does not change, and the figures in the paper no longer agree to the document in the file. Nobody notices until a reviewer tries to agree one.
Linking documents to the figures that depend on them
Storage is not the hard part. The hard part is that a figure in a working paper should lead to the document and page it came from, and a document should lead to everything that relies on it.
The second direction is the one usually missing, and it is the one that matters when something changes. If a client re-issues a bank confirmation in week five, the question is which papers relied on the old one. Without a link from the document back to the work, answering that means opening every paper in the section.
| Direction | Question it answers | Needed when |
|---|---|---|
| Figure to document | Where did this number come from? | Review, and any later examination of the file |
| Document to figures | What relies on this document? | A document is replaced, found to be unreliable, or turns out to be the wrong version |
Knowing what is outstanding
At any moment in an engagement, someone should be able to answer three questions without opening a folder: what has been requested, what has arrived, and what is blocking work that cannot start. Most firms cannot answer the third at all, which is why teams discover a blocking gap in week four rather than week one.
- Requested, with the date and the person it was requested from.
- Received, with the date and what was actually received, which is not always what was asked for.
- Rejected or returned, with the reason, because a file received in the wrong format is not received.
- Outstanding, sorted by what it blocks rather than by when it was requested.
- Not required after all, marked as such, so it stops appearing on the chase list.
What to fix if you only fix one thing
Give the client a single view of what is outstanding, and stop chasing by email. It costs nothing in audit skill, removes most of the friction from the relationship, and reclaims the largest single block of elapsed time in the engagement. Everything else on this page is a refinement by comparison.
How we think about this
Lekhya treats a document as the end of an evidence chain rather than an attachment: finding, procedure, transaction, extracted field, source document, page, navigable in both directions. Where a link is missing it is shown as a gap rather than omitted, because a blank in a schedule is ambiguous and a gap is not. None of that improves the audit judgement. It means the judgement can still be defended two years later.
This is practical guidance written for audit teams, not an authoritative statement of any standard or statute. Standards, forms and timelines change. Check the current text issued by the ICAI and the relevant authority before relying on anything here in an audit file.
Related resources
- Audit documentationAudit evidence: what makes it sufficient, appropriate and traceableSufficiency and appropriateness without the textbook circularity, a reliability hierarchy for Indian practice, and what has to be recorded for evidence to be traceable.
- Audit workflow & automationThe statutory audit workflow, from evidence collection to sign-offAn engagement end to end: acceptance through to sign-off, with where the time actually goes and the four points at which engagements reliably stall.
- Audit documentationAudit trails: the client's, and your ownThe edit log the Companies Act now requires in the client's software, and the record of what the audit team did to its own file. What each must show, and how to test the first.
Free tools, no signup
Each one reads a file entirely in your browser — or, for two of them, just answers a few questions — and nothing is ever uploaded.
- Purchase register checkDuplicate payments, round-sum amounts, weekend-dated invoices, incomplete records.
- GSTR-2B reconcilerMatches your purchase register against GSTR-2B — missing invoices either side, amount mismatches.
- Bank reconciliation checkerMatches your cash book against a bank statement — what's outstanding on each side.
- Fixed asset register checkDuplicate asset codes, missing location or useful life, negative net book value.
- Payroll red-flag scannerBank accounts shared across employees, and records missing an ID, PAN or salary.
- CARO applicability checkerAnswer a few questions about the company and see which CARO 2020 clauses likely apply.
- MSME payment delay checkerMSME-flagged invoices paid — or still unpaid — beyond the 45-day Section 22 limit.
- Related-party transaction screenerTransactions above a threshold you set with no approval reference, grouped by counterparty.
- Depreciation useful life checkerCompares a useful life in use against the Schedule II, Part C indicative figure.
- Materiality calculatorOverall and performance materiality against PBT, revenue, assets or equity, with reasoning shown.
- Sample size calculatorStatistical sample size for a test of controls or a substantive test of detail.