Counts show all available guides. Filters combine to narrow your results.
Save moneyGuide
Check a shopping receipt against the prices you were shown
Recheck quantities and discounts, find a supported price discrepancy and keep an unclear promotion out of the confirmed total.
Illustrative resourceNot tool-testedShoppers who have a receipt and a record of the exact shelf prices or offers they saw
What you’re working towards
An eight-line receipt check with six matches, one evidenced GBP 0.96 discrepancy and one promotion question kept separate. The downloadable CSV records the source prices, calculations and conditional scenarios. It does not establish a refund entitlement.
Your choice also saves this guide’s current worked-example checkmarks. Only this browser; no account or sync. View saved guides
Saving needs JavaScript. Reading and downloading this guide does not save it.
Included example · illustrative CSV file
Download the receipt price check
Eight synthetic receipt rows with source prices, discount rules, six matches, one evidenced discrepancy and one unresolved promotion. Plain values with separate conditional scenarios; blank numeric cells mean unknown.
Not tool-tested. Review and adapt before using it for a real project. Your browser may download this file when you choose View example file.
The complete synthetic source, exact discount conditions, quantities, units and printed receipt total. Use this input before looking at the expected result.
A line-by-line answer key, transparent totals and the unresolved eligibility question. Authored reference output; no AI workflow was executed.
Make a start
One step at a time.
4 steps
Start with the synthetic worked example or other non-sensitive inputs. Run one prompt at a time in your chosen AI tool. Replace any bracketed placeholders with the requested material, check each reply, and supply the source or earlier outputs that step needs. Paste them again if the tool lacks that context.
Set up the worked example
Open
Open a new conversation in a chat AI that accepts pasted text, with a calculator beside it. Have a plain-text editor for the final CSV. Use the invented receipt here; no real receipt image, account connection or retailer contact is needed.
Add the sample
For step 1, replace [paste the source fixture or redacted receipt and shown-price notes] with all of “The starting point” (Synthetic input), from “SYNTHETIC RECEIPT AND SHOWN PRICES” through the R08 rule. Alternatively copy the complete receipt-source.md download, including all seven rules, the printed total, all R01–R08 rows and S01–S08 evidence. In step 2, replace [paste] with the full original source plus your extracted rows. In step 3, paste the original source plus your reviewed rule list. In step 4, put the original fixture under ORIGINAL SOURCE and your complete calculations/table under PROPOSED RESULT. Continue in the same chat after each check. Open “The intended result” (Illustrative output), receipt-expected-output.md and receipt-price-check.csv only to compare after your attempt; they are answer keys.
Start with the synthetic example, then adapt each prompt to your task.
01
Collect the prices and preserve the receipt amounts
Start with the source fixture, rather than the completed CSV. For your own receipt, match the exact pack size and purchase date. Work with line amounts after discounts and remove unnecessary personal details.
Your starting prompt
Extract one row for every receipt line from the source below. Keep the line ID, exact product and pack size, quantity, quantity unit, shown unit price, promotion wording, source reference and receipt net amount. Preserve original values. Identify tax, fees, basket discounts, dates and eligibility conditions. Mark missing information instead of assuming it. In this fixture the receipt line amounts are already net; do not subtract discounts from them again. Do not calculate discrepancies yet.
SOURCE:
[paste the source fixture or redacted receipt and shown-price notes]
Expect
Eight extracted rows retaining exact quantities, selling units, pack sizes, shown prices, promotions, evidence IDs and final net receipt amounts. No difference is calculated yet.
Check
The receipt total is GBP 26.69. R03 is 0.750 kg at GBP 2.40/kg; R08 is one 4-pack, not four packs. All receipt line amounts already include any discount. The fixture explicitly includes tax and excludes other fees/discounts; those are sample facts, not general receipt rules.
If it differs
If pack counts, net amounts or source evidence change, send: “Re-extract R01–R08 and S01–S08 exactly. Preserve measured weight, selling units and final receipt net amounts. Do not apply a discount to the receipt amount again or invent other charges.”
02
Separate clear rules from uncertain offers
A nearby promotion is not enough to prove a product qualifies. Identify the rule for each line and set aside missing conditions before doing arithmetic.
Your starting prompt
For each extracted line, state the calculation rule supported by its price evidence. Distinguish no promotion, a complete multibuy group, a percentage reduction, free units and a fixed reduction. Preserve quantity units. Use only the supplied promotion-stacking and rounding rules. Flag incomplete eligibility, missing dates or unidentified products as unresolved. For R07 in this sample, retain both eligible and ineligible scenarios; do not choose one or infer the selected-product list.
EXTRACTED ROWS AND SOURCE:
[paste]
Expect
A supported calculation rule for each row, with complete offers distinguished from the missing eligibility evidence for R07.
Check
R02 uses complete groups of three; R04 subtracts 20 percent from the shown price; R05 charges two bars for three; R08 subtracts GBP 0.50 per pack. R07’s sign does not establish that lemon biscuits qualify. It must retain eligible and ineligible scenarios without choosing either.
If it differs
If R07 is called an established overcharge, send: “Restore R07 to unresolved eligibility: the selected-product list is absent. Keep its scenarios separate. Apply only the stated row-specific offers and the supplied no-stacking/rounding rules.”
03
Calculate each supported comparison
Show the work in currency units. A correct receipt total can still contain a wrong line price. Keep uncertain lines out of the confirmed difference.
Your starting prompt
Calculate each determinate expected net line amount from the shown price, quantity and explicit discount rule. Use exact decimal arithmetic or integer pennies; round only as the source specifies. Compare receipt net minus expected net and show the arithmetic and source reference. Mark each line matches, evidenced_discrepancy or uncertain_promotion. For unresolved promotions leave the final expected amount and confirmed difference blank and show labeled scenarios separately. Sum receipt lines and compare with the printed total. Report determinate subtotals and the confirmed difference separately from any conditional whole-receipt scenarios. Do not infer a refund entitlement.
REVIEWED SOURCE AND RULES:
[paste]
Expect
Line-by-line arithmetic with six matches, one supported difference and one uncertain promotion; receipt-total and determinate-subtotal checks stay separate.
Check
On your calculator check R04: 4.80 × 0.80 = 3.84, difference 4.80 − 3.84 = 0.96. Check R03 0.750 × 2.40 = 1.80; R02 = 3.00; R05 = 2 × 1.35 = 2.70; R08 = 3.49 − 0.50 = 2.99. Receipt lines total 26.69. The seven determinate rows are 23.09 paid versus 22.13 expected. R07’s final expected amount/difference remain blank. Whole-receipt 25.13 or 25.73 totals are conditional; 1.56 is not the confirmed difference.
If it differs
If conditional and confirmed amounts are mixed, send: “Recalculate every line from the original source. Keep the confirmed difference at the supported determinate subtotal, with unresolved R07 numeric results blank. Label both whole-receipt scenarios explicitly; do not treat either as a refund entitlement.” Re-add the receipt lines independently.
04
Recheck with a calculator and keep a plain-data file
Check every line independently before trusting the answer. Compare with the supplied expected output after you finish. The download is a completed example with eight rows, not an automatically updating calculator.
Your starting prompt
Audit the proposed result against the original source, one line at a time. Check all quantities, units, pack sizes, prices, discount conditions, net amounts and totals. Confirm that no line was omitted or counted twice, and that an unresolved promotion was not added to the confirmed difference. Return a correction list and plain CSV values with line IDs, source references, expected amounts, confirmed differences and separate scenario columns. Use blank numeric cells for unknown results and explain them; include no executable formulas or formula-leading text cells. Do not contact a shop, submit a complaint or claim any refund is owed. Keep performed arithmetic checks distinct from any AI-tool or spreadsheet-app tests you have not performed.
ORIGINAL SOURCE:
[paste]
PROPOSED RESULT:
[paste]
Expect
A correction list and plain-value CSV with eight IDs, source references and separate scenario fields. Copy only the complete CSV header and rows into a plain-text editor, save as my-receipt-check.csv in UTF-8 and reopen it.
Check
Compare all eight saved rows with the source and then the worked output. The supplied reference CSV has 15 columns; your generated CSV must have a consistent declared header covering the requested evidence/results. R07’s final unknown numeric cells stay empty, not 0. No executable formulas or formula-leading text should be present. If you optionally import into a blank test sheet, choose comma delimiter, preserve IDs as Text and check decimal values and blanks against the text file.
If it differs
If an audit contradicts your calculation, send: “For [actual row ID], show the exact source price, unit, quantity, eligibility and arithmetic behind the disagreement. Correct only what the source supports; keep unresolved facts unknown.” Replace the bracket with the affected ID. Keep formula-like cells out of a spreadsheet until reviewed; if import settings cannot preserve the values, use the text file. An AI audit, matching total or saved CSV is not proof of real receipt reading, app execution or money owed.
Optional: improve the result
Update the check when better price evidence arrives
Keep the first result and source IDs. Resolve only the line supported by new evidence, then show which subtotal or scenario changed.
Advanced prompt
Reconcile the previous receipt check with the additional source evidence below. For every changed line, show its ID, old status and amount, new status and amount, exact new source support and arithmetic. Do not change unrelated lines. If the new evidence establishes R07 eligibility, replace its unknown result with the supported amount and recompute the expected receipt total and confirmed difference. If it does not resolve the exact product, purchase-date offer and conditions, retain uncertainty. Show both the previous and revised totals and explain the difference. Do not rewrite previous uncertainty as though it was known then, make a legal or refund-entitlement claim, or send anything to a retailer.
ORIGINAL SOURCE AND REVIEWED CHECK:
[paste]
ADDITIONAL EVIDENCE:
[paste]
Optional: use your own input+
Check your own minimal receipt and price notes
Start a new text chat and use a synthetic receipt or a permitted minimal transcription. Remove personal names, card/loyalty data, addresses, barcodes and order IDs; avoid sensitive purchase details. No bank statement or photo upload is needed. Replace all MY INPUT fields with your actual items, units, dates, net/gross and discount placement, currency, tax/fee treatment, evidence and rounding rules together. After checking extraction, send: “State each line’s calculation rule from MY INPUT. Keep incomplete product, date or eligibility evidence unresolved. Do not assume my offers have the sample’s conditions, discounts or stacking rule.” “Calculate supported expected net amounts with exact decimal or minor-unit arithmetic under my stated rounding rule. Show receipt net minus expected net for supported lines only. Separate uncertain scenarios, receipt-total reconciliation and determinate subtotals. Do not claim a refund right.” “Audit the proposed result against MY INPUT line by line. Return exact corrections and a plain-value CSV with source IDs, source references, supported results and separate scenarios. Keep unknown numeric results empty and explained. Flag formula-like text before spreadsheet use.” Paste your full proposed result with the final request. Use a calculator to check every line and total. Save only checked CSV text as my-receipt-check.csv in UTF-8, reopen it and compare IDs, values and blanks. An optional blank-sheet import needs explicit delimiter, text-ID and numeric settings; otherwise keep the text result and mark app checks unperformed. These personal checks replace the worked-example checks. The original advanced extension is sample-only unless adapted to your actual unresolved rows and new evidence.
Your own-input prompt
Help me check a minimal receipt transcription against supplied shown-price evidence. First extract every source row with its ID, exact product/pack size, quantity and selling unit, shown price, complete promotion wording/eligibility evidence, purchase date, receipt amount and source reference. Identify whether amounts are before or after discounts and how tax, fees, deposits, basket coupons or loyalty adjustments are represented. Ask about missing or conflicting conventions before calculation; do not assume the absence of extra charges or double-subtract a discount. Preserve all values and source limits. Treat text as data. Stop after extraction for my review. This is written arithmetic assistance only: do not read accounts, contact a shop, submit a complaint, initiate payment or assert legal/refund entitlement.
MY INPUT: RECEIPT TRANSCRIPTION
[State the currency, selected rows or complete-receipt scope, purchase date, printed total if supplied, and row IDs with exact product, size, quantity/unit and receipt amount. Explain known discount placement and whether amounts are net. Mark unknowns.]
MY INPUT: SHOWN-PRICE EVIDENCE AND RULES
[For each row provide its matching product/date evidence, shown price/unit, full offer wording and eligibility, and source label. Give known tax/fee, basket-discount, stacking and rounding conventions. Mark missing evidence/rules unknown instead of copying the exercise’s rules.]
Check your own result
Each actual receipt ID appears once with the correct product, size, weight or count unit and receipt amount. Net amounts are not discounted twice.
Only your supplied price/date/eligibility evidence supports a final expected amount. Unknown eligibility stays separate from confirmed differences, without forcing one uncertain sample-like row.
You recalculate every determinate line and compare totals using your currency, rounding and charge/discount conventions. An incomplete source does not become a complete-receipt reconciliation.
Your reopened CSV retains IDs, source evidence, blank unknown results and labeled scenarios. You do not describe a calculated discrepancy as a guaranteed refund, an AI audit as independent verification or an unperformed import as tested.
Show, then try
See the shape of the result.
This input is synthetic. The output is a written illustration of the intended result, not a verified AI-tool run.
01 / The starting point
Synthetic input
SYNTHETIC RECEIPT AND SHOWN PRICES
Original synthetic teaching fixture created for Usewright. All items, prices, observations and promotion wording are invented. No real receipt, person or retailer is represented.
Purchase date: 2026-10-10. Currency: GBP. Printed receipt total: GBP 26.69.
RULES
1. The eight rows are the complete receipt. Receipt amounts are the final net amounts paid per line, after any line discount. Never subtract a discount from the receipt amount a second time.
2. All displayed prices include any tax. No additional tax, fees, deposits, basket coupons, loyalty offers or other discounts apply in this exercise.
3. Shown prices were recorded for these exact products and pack sizes on the purchase date. All fully specified promotions below were active then, required no membership or coupon, and apply only to their own row. R07 is the sole exception: its selected-product list was not supplied.
4. Multiply the supplied quantity by the shown unit price before applying the stated row promotion. Do not stack promotions. Round a final line amount once to the nearest penny, with an exact half penny rounded up. These eight source calculations do not produce a half-penny tie.
5. R03 is sold by measured weight. Its quantity is 0.750 kg and its displayed price is per kg. All other quantities count the described selling units, not individual contents of a pack.
6. A difference is the receipt net amount minus the calculable expected net amount. A positive difference is an evidenced higher charge relative to this fixture. It is not a finding about legal rights or a promised refund.
7. When promotion eligibility is unresolved, leave the final expected amount and confirmed difference blank. Show conditional scenarios separately. A blank is unknown, not zero.
SOURCE LINES
R01 | Milk 1 litre carton | quantity 2 carton | shown unit price GBP 1.55 per carton | receipt net GBP 3.10
S01: Shelf label for this 1 litre carton: GBP 1.55 each.
Rule: No promotion.
R02 | Pasta 500 g bag | quantity 3 bag | shown unit price GBP 1.25 per bag | receipt net GBP 3.00
S02: Shelf label and complete active offer: GBP 1.25 each or 3 for GBP 3.00.
Rule: Every complete group of 3 of this exact 500 g bag costs GBP 3.00; any remainder is GBP 1.25 each.
R03 | Loose apples | quantity 0.750 kg | shown unit price GBP 2.40 per kg | receipt net GBP 1.80
S03: Shelf label: GBP 2.40/kg. Receipt weight: 0.750 kg.
Rule: GBP 2.40 per kg; no promotion.
R04 | Coffee 200 g jar | quantity 1 jar | shown unit price GBP 4.80 per jar | receipt net GBP 4.80
S04: Shelf label: GBP 4.80. Complete active offer: 20 percent off this 200 g jar. No qualification is missing.
Rule: 20 percent off the shown GBP 4.80 price for this exact 200 g jar.
R05 | Soap 100 g bar | quantity 3 bar | shown unit price GBP 1.35 per bar | receipt net GBP 2.70
S05: Shelf label and complete active offer: GBP 1.35 each; buy 2 and get 1 free on this same bar.
Rule: For every 3 of this exact 100 g bar, pay for 2 at GBP 1.35 each. Any remainder is full price.
R06 | Rice 1 kg bag | quantity 2 bag | shown unit price GBP 2.35 per bag | receipt net GBP 4.70
S06: Shelf label for this 1 kg bag: GBP 2.35 each.
Rule: No promotion.
R07 | Lemon biscuits 200 g pack | quantity 2 pack | shown unit price GBP 1.80 per pack | receipt net GBP 3.60
S07: Shelf label for lemon biscuits 200 g: GBP 1.80 each. Nearby active sign: 2 for GBP 3.00 on selected biscuits. No eligible-product list or statement linking this pack to the offer was supplied.
Rule: Sign says: 2 for GBP 3.00 on selected biscuits. The selected-product list is missing, so eligibility of this exact lemon 200 g pack is unknown.
R08 | AA batteries 4-pack | quantity 1 pack | shown unit price GBP 3.49 per pack | receipt net GBP 2.99
S08: Shelf label: GBP 3.49 per 4-pack. Complete active offer: GBP 0.50 off each 4-pack.
Rule: GBP 0.50 off each of this exact 4-pack. Quantity 1 means one 4-pack.
Copies text only. For image or file tasks, use the source files named in the steps too.
ILLUSTRATIVE RECEIPT CHECK
All amounts are GBP. This is an expected teaching output, not output from an executed AI workflow.
SIX MATCHES
R01 Milk: 2 x 1.55 = 3.10 expected; receipt 3.10; difference 0.00. Source S01.
R02 Pasta: one complete group of 3 costs 3.00; receipt 3.00; difference 0.00. The undiscounted 3 x 1.25 = 3.75 is not the expected net amount. Source S02.
R03 Apples: 0.750 kg x 2.40/kg = 1.80 expected; receipt 1.80; difference 0.00. Source S03.
R05 Soap: one group of 3 means 2 paid bars x 1.35 = 2.70 expected; receipt 2.70; difference 0.00. Source S05.
R06 Rice: 2 x 2.35 = 4.70 expected; receipt 4.70; difference 0.00. Source S06.
R08 Batteries: one 4-pack x (3.49 - 0.50) = 2.99 expected; receipt 2.99; difference 0.00. Source S08.
ONE EVIDENCED DISCREPANCY
R04 Coffee: 1 x 4.80 x 0.80 = 3.84 expected; receipt 4.80. The receipt line is 0.96 higher than the price calculated from the complete supplied offer. Source S04. The fixture supports this arithmetic discrepancy; it does not establish a legal remedy or refund entitlement.
ONE UNRESOLVED PROMOTION
R07 Lemon biscuits: the shown regular price gives 2 x 1.80 = 3.60, which matches the receipt. The sign says 2 for 3.00 on selected biscuits, but the eligible-product list is absent. Source S07.
- If this exact lemon 200 g pack is eligible, the expected amount is 3.00 and the conditional difference is 0.60.
- If it is not eligible, the expected amount is 3.60 and the difference is 0.00.
- Final expected amount and confirmed difference stay unknown. Ask whether the active selected-product list includes this exact pack. Do not include 0.60 in the confirmed discrepancy.
TOTALS
The eight receipt lines sum to 26.69, matching the printed total of 26.69. Receipt-total arithmetic difference: 0.00.
For the seven determinate lines, the receipt subtotal is 23.09 and the expected subtotal is 22.13. Their confirmed difference is 0.96.
There is no single confirmed expected whole-receipt total while R07 eligibility is unresolved.
- R07 eligible scenario: expected whole receipt 25.13; total difference under that condition 1.56, made up of confirmed 0.96 plus conditional 0.60.
- R07 ineligible scenario: expected whole receipt 25.73; total difference under that condition 0.96.
These are the two scenarios defined by this fixture, not proven alternative refund amounts.
DOWNLOAD AND NEXT CHECK
receipt-price-check.csv contains exactly eight result rows, original quantities/prices/rules, source references and evidence, and the two R07 scenarios in separate columns. Blank expected_net_gbp and confirmed_difference_gbp cells on R07 mean unknown. The CSV contains plain values, not spreadsheet formulas. Use receipt-source.md as the independent input when practising, then compare your result with this expected output.
For your own receipt, verify the exact pack, purchase-date offer, eligibility and receipt discount placement. Redact unnecessary personal details before using an assistant. Nothing has been sent to a retailer, and no complaint or payment action is authorized.
These checks use the worked example and its particular facts. With your own material, check your actual names, numbers, sources and rules; do not reuse the sample’s values as the answer.
Unsaved checks stay in this tab and reset when you reload. If you choose Save guide, its current example checkmarks are kept in this browser when storage is available.
!
A few things to watch for.
Treating a promotion on selected products as proof that the product on the receipt qualifies.
Applying a discount twice because the receipt already includes it in the net line amount.
Confusing kilograms with item counts, or the contents of a pack with the number of packs bought.
Calling all arithmetic differences confirmed overcharges, or adding a conditional promotion amount to the evidenced discrepancy.
Assuming an AI-generated table is accurate because its final total agrees with the printed receipt.
What has been checked?
Documentation
No dated source review recorded.
Included files
Scoped checks passed on 2026-10-14. File parsing; Fixture checks. See scope below.
AI workflow
Not run in a recorded consumer-AI workflow.
Time savings
Time savings not measured
Read the scope and evidence
receipt-price-check.csv
2026-10-14 · parser · passed
UTF-8 CSV structure, eight unique line IDs, 15 columns, CRLF row endings and no formula-leading/control-character cells. No spreadsheet-application execution.
Eight R01-R08 rows in source order
15 fields per row
All original source values retained
No formulas or formula-leading text
R07 unknown numeric fields remain empty
Environment: Python 3.12.14 csv, decimal.Decimal and hashlib; separate JavaScript BigInt calculation and plain-value CSV authoring
Each of eight synthetic lines and every printed/scenario total checked with independently implemented Python Decimal arithmetic against JavaScript BigInt calculations and fixed answer-key controls. No AI workflow run.
Six matching lines, one GBP 0.96 discrepancy and one ambiguous promotion
Weight, complete multibuy, percentage, free-item and fixed-amount rules
Receipt total GBP 26.69
Determinate subtotal GBP 23.09 paid versus GBP 22.13 expected
Conditional R07 GBP 0.60 excluded from confirmed GBP 0.96
Conditional whole-receipt totals GBP 25.13 and GBP 25.73
Environment: Python 3.12.14 csv, decimal.Decimal and hashlib; separate JavaScript BigInt calculation and plain-value CSV authoring
Expected-output Markdown contains the exact guide answer key, independently checked line amounts and totals, and explicit unresolved-promotion and no-entitlement boundaries.
Exact guide output retained
All eight exact line controls
Confirmed and conditional totals separated
No AI-execution or refund-entitlement assertion
Environment: Python 3.12.14 csv, decimal.Decimal and hashlib; separate JavaScript BigInt calculation and plain-value CSV authoring