Home · Business · Operations & Logistics · Inventory & stock

YES

As of 13 August 2026, AI can find errors in your stock counts.

This still needs a person who signs their name to it.

Can you do it?

5 minutesto a draft.

30 minutesto something you’d act on.

Cost, all in£0

Skill neededchat-fluent

Who has to check ita colleague

What the alternative costsThe available tool data gives no price for a human stock-counting alternative.

If this goes wrong: a genuine discrepancy is missed or a correct count is changed, causing avoidable purchasing, picking or production problems.

What to actually do

  1. Hand it to a person

    The route this page recommends

    A person who owns the outcome does this end to end, worth it when the failure is dear.

  2. Use a tool built for this

    Second choice

  3. Do it yourself

    The distant third

    A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

    How to actually do it

    1. Export the inventory-system records for the same locations and count period, including item code, description, unit, location and recorded quantity.
    2. Gather the completed count sheets and any notes about damaged, quarantined, opened or incoming stock.
    3. Put both datasets into separate clearly labelled tables, keeping item codes, units and location names unchanged.
    4. Paste the tables and any relevant counting rules into the prompt, then ask the model to produce the exception list and manual checks.
    5. Compare each flagged item with the original count sheet and system export to remove flags caused by formatting, duplicate rows or mismatched units.
    6. Have a colleague physically recount the remaining flagged locations and check goods received, dispatches, transfers and adjustment records before changing the stock system.
    7. Record the confirmed cause and approved adjustment in your normal stock-control system, leaving the original count and the AI-generated exception list attached to the audit record.

    Prompt

    You are checking an inventory count for a UK workplace. Compare the stock-count data below with the inventory-system data and identify discrepancies. Treat item codes, locations and units as exact matches unless the data says otherwise. For each discrepancy, show the item code, description, location, counted quantity, recorded quantity, difference, and percentage difference where the recorded quantity is not zero. Separate confirmed data mismatches from possible counting or data-entry errors. Do not invent missing values, merge similar item codes, or assume that different units are interchangeable. Flag duplicate rows, missing items, negative quantities and unusually large differences. Give each flag a reason and a specific manual check for a colleague to carry out. Finish with a clean exception list ordered by the largest absolute quantity difference. State any limitations caused by missing or inconsistent data.

Stock-count data:
[PASTE THE COUNT SHEET OR TABLE HERE]

Inventory-system data:
[PASTE THE EXPORT OR TABLE HERE]

Relevant rules:
[PASTE UNIT DEFINITIONS, LOCATION RULES OR COUNTING NOTES HERE]

    Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.

What it gets wrong

  • It cannot see stock that was put in the wrong location, left in a vehicle or counted in a different unit unless you provide that context.
  • It cannot distinguish a genuine loss from a timing difference without your goods-in, dispatch, transfer and adjustment records.
  • It can flag a discrepancy but cannot physically recount the shelf or decide whether a stock adjustment is authorised.
  • It may treat inconsistent item codes, duplicate rows or copied spreadsheet errors as real stock problems until you resolve the source data.

Even on a YES, the friction has a name: judgement under ambiguity, verification cost and stakes of error.

How we scored this

Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.

AxisScore (0–2)
Output2
Inputs2
Verification1
Liability1
Effort delta2
Total8 / 10

The methodology and its thresholds are published in full.

FAQ

Can ChatGPT check my stock count?
Yes. Give it the count sheet and the matching inventory export, and it can compare quantities, flag discrepancies and organise the checks. A colleague still needs to recount the stock and approve any adjustment.
Can AI spot stock discrepancies in Excel?
Yes, if the workbook contains consistent item codes, locations and units. AI can identify mismatches and duplicates, but you must check the source rows because spreadsheet formatting and timing differences can create false flags.
Can AI tell me why my stock count is wrong?
Only partly. It can suggest checks against goods received, dispatches, transfers and data-entry errors, but it cannot know what happened to the physical stock without records and a recount.
Is it safe to let AI adjust my stock levels?
No, not automatically. Use AI to prepare an exception list, then have an authorised colleague verify the count and supporting records before making the adjustment.

Nearby answers

All Inventory & stock rulings

Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.

The newsletter

AI news, new answers and product picks, straight to your inbox.