Your own count before the auditor arrives: what and when

Trying to recount everything the day before the auditor arrives is the most expensive and least reliable way. There is another: a few weeks before the visit you run an internal count yourself. If the auditor finds a difference, there is no time to correct it, and the finding goes into the report. If you find it in advance, there is time to correct it, establish the cause and write an explanation. Below we cover what an internal count should include, when to run it and how to document the result.

Why count in advance, and when

An internal count gives four benefits. First, there is time to correct differences. Second, the cause of each difference is found and the book data is cleaned up. Third, you are ready for the auditor's questions: what, why and what was done are all clear. Fourth, the team sees real problems in the process, which helps next year too.

As a practical guide for timing, start at least 3–4 weeks before the auditor's visit. Week one is the count. Week two is the analysis of differences and the correcting documents. Week three is a recheck of the corrected areas. Week four is a buffer. This is an approximate schedule; it varies with the size of your site.

What to count: choosing by risk

You do not have to count everything. Start with the higher-risk groups:

• high-value groups: few items, large sums;
• fast-moving goods: errors build up here fastest;
• groups with differences last year or at the previous check;
• balances off site: on consignment, in processing, in transit;
• goods with expiry dates and batches;
• third-party and commission goods;
• a random sample from the remaining items.

The last point matters. If you choose only the "convenient" groups, the result looks good but does not show reality. A random sample gives an honest picture of the general state of the warehouse.

Who runs the count matters too. Ideally it is done by people who do not keep the records: staff from another department or an outside contractor. Warehouse staff show the way and answer questions, but do not count their own zone.

An illustrative example: a sample count

Suppose the warehouse has 9,000 items. You pick 1,200: 400 high-value, 300 fast-moving, 200 with differences last year, 100 off-site balances and 200 as a random sample (400 + 300 + 200 + 100 + 200 = 1,200). The count shows 54 differences: 1,200 × 0.045 = 54, that is 4.5 per cent. After analysis they split like this: 38 document errors, 11 location errors, 5 real shortages (38 + 11 + 5 = 54). By the time the auditor comes, the document errors are corrected, the locations put in order, and a written explanation is prepared for the 5 shortages.

Documenting the result

The internal count must also be documented, otherwise for the auditor it does not exist. Usually the following is enough:

• count sheets with dates and signatures;
• a list of differences with a cause column on every line;
• correcting documents with the actual date;
• notes for differences whose cause was not found;
• a one-page summary for the manager: what was counted, how many differences were found, how they were corrected, what remains open.

An important rule: do not correct documents retroactively. A correction is made with today's date and an explanation. The auditor needs a clear trail, not a hidden and "wound back" one. Then it is visible how the count went and what was done with the differences. Do not hide unresolved issues either: if every open issue has an owner and a deadline written down, it looks not like a flaw but like a managed process.

Typical mistakes

• Counting a day before the visit. Even if a difference is found, there is no time to fix it.
• Counting only the convenient groups.
• Posting differences to the books without establishing the cause. The problem returns next year.
• Trying to hide findings. An auditor usually soon spots contradictions in the documents.
• Entrusting the count to the person who keeps the records. Independence is lost.

Conclusion

An internal count turns the audit from a test into a process that has been checked in advance. Inventory can run an external, independent count before the audit: we count free from internal interests and hand over a report with an analysis of the differences, and you have time to fix shortcomings before the auditor arrives.