Audit evidence: reliable and sufficient

An audit conclusion does not start with "I think" but with "here is the evidence". If the statement "the shortage is 12 million so'm" is not backed by documents, a count and confirmation, the manager may reject it or, worse, make a wrong decision. The quality of audit evidence is defined by two words: reliability and sufficiency. The first says where the evidence came from, the second how much of it there is.

Types of evidence

In a warehouse the auditor usually works with six kinds of evidence:

• Physical: a count, observation, seeing the goods in person. Example: "38 boxes stand on the shelf".
• Documentary: a delivery note, an invoice, the 1C log, a write-off certificate.
• External confirmation: written confirmation of a balance or settlement from a supplier, a bank or a third-party warehouse operator.
• Inquiry: an employee's explanation. Weak as sole evidence, but it gives direction to the work.
• Recalculation: the auditor recomputes quantities and sums personally.
• Analysis: relationships and changes between figures, such as an unexpected deviation.

The strongest conclusion comes when several kinds of evidence confirm each other: when the count, the document and the external confirmation all show the same result, no doubt remains.

Reliability: where the evidence comes from

The rule: the more independent the source, the more reliable the evidence. A rough order:

1. What the auditor saw or counted personally.
2. A document from an external source: a supplier's delivery note, a bank statement.
3. An internal company document under strong control.
4. An internal document with no control.
5. An oral explanation.

An example. The storekeeper says "the goods are not with the supplier, they are in our warehouse". The supplier's written confirmation is far more reliable than his word. Likewise an original beats a copy, and a document obtained directly beats "a screenshot sent to me", because a screenshot or a Telegram chat is easy to edit. The reliability of an internal document also depends on the strength of control: a document with two signatures that is locked in 1C is more reliable than an Excel sheet one person can edit.

Sufficiency: how much evidence is needed

Sufficiency is about quantity, and risk sets it. One delivery note cannot show that "receiving works correctly". To judge receiving control, for instance, you might pick 40 delivery notes at random and compare each with the actual count and the 1C entry. If two errors turn up, that is 5 percent, and the conclusion has a basis. In a high-risk process (say, stock adjustments) the sample is larger, or all entries are reviewed.

One conclusion is often built from several pieces of evidence. Suppose an auditor wants to write "shortage of 12 million so'm". First they take the count sheet signed by two teams (physical evidence), then the 1C balance report at the count date (documentary), the supplier's confirmation for the latest receipts (external), they recount items with a variance and recompute the value of the shortage at cost themselves. Only then does the figure "12 million" become a conclusion.

How to collect and keep evidence

Each piece of evidence is recorded in the working papers: what, from where, when, who obtained it, what it relates to. In practice the following help:

• photographs are taken so that date and place are visible;
• a screenshot shows the time and the user name;
• a copy of a document is marked "true copy";
• 1C logs are exported to a file so that they cannot be altered;
• files get clear names, for example count_2026-06-30_zone-A.xlsx.

An employee's explanation is recorded too, but it is marked as a lead, not as evidence: until a document or a count confirms it, it does not enter the conclusion.

When pieces of evidence contradict each other

If 1C shows one balance and the delivery note another, the auditor relies on the order of reliability and digs deeper: the external document, the count and the log are reviewed again. The contradiction is not hidden but described in the report. Sometimes the contradiction itself is the most important finding, showing that the accounting system cannot be trusted.

Conclusion

A conclusion without evidence turns into an opinion, and an opinion does not improve control. OmborAudit aims to support every finding with a reliable source and a sufficient sample, and to keep working papers in a form another auditor can review. If another specialist, reading those records, reaches the same conclusion, the evidence can be considered sound.