Investigate differences between a stock count and your system

Compare our physical stock count with the system quantities below and help {{audience}} investigate the differences before any adjustment is posted. The goal is to separate counting and recording errors, which should be fixed at the source, from real losses, which should be reported.

Count results and system quantities (item or SKU, location, system quantity, counted quantity, unit cost):
"""
{{count_data}}
"""

Activity around the count (deliveries received, orders shipped, returns, transfers, adjustments, with dates and times):
"""
{{activity}}
"""

When and how the count happened: {{count_details}}
Our tolerance for differences: {{tolerance}}

Produce:
1. Summary: items counted, items with differences, net value difference and total absolute value difference, with the calculations. Leave out any line you suspect is in the wrong unit, and show its effect separately.
2. A table of differences outside tolerance, largest value first, with the columns Item, Location, System, Counted, Difference, Value. If our tolerance has two tests (units and value), treat a difference as outside tolerance if it fails either, and say so.
3. For each item in the table, the explanations the data supports, most likely first, such as a delivery received but not booked in, an order shipped before the count but booked after it, a unit mix-up (cases counted as units), or stock sitting in a second location. Quote the activity line that supports each. A line explains a difference only if the stock physically moved before the item was counted while the system figure did not yet reflect it (for example, shipped before the count but booked after the system quantities were taken, or received but never booked). A movement already booked before the system quantities were taken cannot explain a difference.
4. Small recording errors: differences inside tolerance that an activity line clearly explains, since they are worth fixing even when small.
5. Recount list: items to count again before adjusting.
6. Unexplained after checking: items where nothing in the activity explains the difference.

Rules:
- Use only the data given. Do not conclude that stock was stolen or that a person made a mistake; give explanations to check, not blame.
- Do not propose posting adjustments. That decision belongs to {{audience}} after the recount.

Keep the text outside the tables under 250 words. If a quantity looks like a different unit (for example exactly 12 times too high), say so and ask, rather than converting it yourself.