Accepting a count result: seven checks before you sign

The count is finished and the report is in your hands, but the job is not over. Before you sign the report or load it into 1C, there are seven checks to make on your side. Each takes ten to thirty minutes, yet it can save you months of asking "where did this figure come from?" The checks are: completeness, totals, a spot recount, the largest differences, the explanations, the file format and the approval routine.

Completeness and totals

First confirm that the number of rows in the report matches the book list. Positions that were not counted should be shown separately as "not counted", not as "0 pieces". A not counted position is not the same as zero: zero means "the goods are not there", while not counted means "nobody looked". Then check the zones: is every part of the warehouse in the report?

Next check the totals by hand. The book total, the counted total and the total difference must agree: difference = counted − book. Look at it both in pieces and in so'm. In Excel a plain SUM formula is enough. Check the sign too: shortages and surpluses should be marked the same way throughout, so that the sign does not flip somewhere down the column.

Look for duplicates too. If one article was counted in two zones and landed in the report twice, the total is inflated. A filter or the remove-duplicates tool in Excel finds this in a minute. Check the reverse as well: are there rows that exist in the books but are missing from the report?

A spot recount

The most reliable check is a spot recount: count 20 to 30 positions yourself on the spot. Choose them at random and mix positions with and without a difference. When your figure matches the one in the report, it is easier to trust the rest. For example, if 29 of 30 positions match, the report is dependable; if only 20 match, ask the contractor to recheck the whole zone. The sample should be chosen by you, not by the contractor. Be sure to include expensive goods, fast-moving goods and look-alike variants with similar names, because errors most often occur there. Record the positions you checked in a table with the date and your name.

Differences and explanations

Sort the list of differences by value and examine the ten largest rows separately. Most of the shortage is usually concentrated there. Every row should carry an explanation, and one that can be checked: if it says "found in another bin", ask which one and go and look. Give surpluses the same attention. A surplus is often the trace of a shortage on another position: the goods were entered under the wrong article, and two similar rows offset each other. An entry such as "cause not established" is acceptable too, but it should state what was checked.

The file and loading into 1C

If, besides the Excel and PDF report, you were given a file for 1C, load it into a copy, not the live database. The most common faults are these: a long barcode turns into "4.6E+12" in Excel; an article number with a leading zero loses the zero; the unit of measure (box or piece) differs from the book unit; the encoding breaks and names become unreadable. Finding and fixing this in a copy is far easier than repairing the live database. A small trial is enough here too: load 20 to 30 rows and compare the name, quantity and unit of measure with the report.

Approval and storage

Once you accept the report, do not rush to post the adjustments. First decide who approves them and from which date: the warehouse manager, the accountant and the director. Agree a threshold in advance as well: for example, differences below a small amount are closed with the manager's signature, while large ones stay open until the director has reviewed them. The adjustment document should carry the count date, not the date the report was delivered. Keep the original file unchanged and save the corrected copy under a different name, so you can compare the two versions at the next count.

Conclusion

Accepting the result is not a formality but the last and cheapest control point of the count. Completeness, totals, a spot recount and the explanations take under an hour and keep errors out of the books. StockCheck delivers the report with a list of differences and is ready to recheck the positions you choose together with you.