Clean and validate accounting data before migration
Prepare reliable customer, supplier, inventory and transaction records by correcting format, duplicates and incomplete fields before transfer.
Section 3.4 · Computerised accounting systems
Exam scope and assessment notes
Cambridge International AS & A Level Accounting 9706, 2026–2028 syllabus, Version 2. These lessons cover the process of transferring business accounts to a computerised accounting system and ways to maintain data integrity during conversion. No specific accounting application is required. Practice is self-checked, not an official Cambridge grade. Full timed Papers 1–4 and human subject-teacher review remain outstanding.
Before you begin
- Use the Topic 1.2 accounting-system skills: understand double entry, trial balances, control accounts, source documents and why complete records matter.
Work on paper. Try each question before revealing help, and explain why your method works.
Quick readiness check
A valid invoice date must be in the required format and within the reporting system's permitted dates. What kind of check is this?
What you will learn
- Identify common source-data defects
- Apply validation and duplicate checks
- Preserve an audit trail of corrections and unresolved exceptions
Inspect the source data before importing it
Source records may contain duplicates, inconsistent names, invalid dates, missing account codes, inactive customers, obsolete inventory items or amounts entered in different formats. Cleanse records against agreed rules before import and keep a copy of the original extract for comparison.
Validation checks test whether fields are present, correctly formatted and within reasonable ranges. Duplicate checks can compare account numbers, tax references or combinations of name and address. A potential match needs review; automatic deletion can merge two real customers or suppliers by mistake.
Record every correction and exception
Document the rule used, the original and corrected value, who approved the change and why it was made. Separate errors corrected before import from unresolved items that must be investigated or excluded under an approved rule.
Data cleaning should not change a transaction merely to make the totals agree. Preserve supporting documents and change history so an authorised reviewer can trace the migrated record back to its source.
Pause and explain
Why retain a copy of the original source extract?
Worked example
A supplier list contains 500 rows, 12 missing payment terms and 8 possible duplicate names. Describe the checks before import and explain why the duplicates should not all be deleted automatically.
Show the worked solution
- Validate required fields such as supplier account code, name and payment terms; resolve or formally flag the 12 missing terms.
- Compare duplicate candidates using stronger identifiers such as supplier number and address, then review the evidence.
- Record each correction or exclusion with source, reason, preparer and approval.
Answer Validate missing fields, investigate the 8 possible duplicates against reliable identifiers, and preserve a correction log. Similar names alone do not prove two records are duplicates.
Try a topic test
Choose Easy, Normal or Difficult. Save your answers and open worked review after finishing.
Test this topic →From guided practice to a new situation
Write your answer and reasoning first. Use a hint only if you are stuck. The model solution is for self-checking; your written work is not automatically marked.
An import file has 120 inventory items. Four have blank item codes, three use negative quantities despite the business rule that stock cannot be negative, and two may be duplicates. Give a check and action for each exception type.
Give me a hint
Separate missing fields, invalid values and possible duplicate records.
Compare my reasoning
- Reject or resolve the four missing item codes using approved master-data evidence.
- Investigate the three negative quantities against stock records and the source ledger; correct only with evidence.
- Review the two duplicate candidates using item identifiers and descriptions; merge only after approval and preserve the log.
Resolve missing codes, investigate invalid negative quantities and verify possible duplicates. Record the source, correction and approval for each change.
Look for these in your work
- Reject or resolve the four missing item codes using approved master-data evidence.
- Investigate the three negative quantities against stock records and the source ledger; correct only with evidence.
- Review the two duplicate candidates using item identifiers and descriptions; merge only after approval and preserve the log.
A sales extract has 2,400 invoices. The file contains 2,398 unique invoice IDs and two repeated IDs with different totals. What should the team do before import?
Give me a hint
A repeated identifier with different amounts may indicate a correction, a duplicate or an error; inspect evidence.
Compare my reasoning
- Compare the repeated records with the original invoices and source-system audit trail.
- Determine whether each is a valid amendment, two separate documents with a bad ID, or an accidental duplicate.
- Correct or exclude only under an approved rule and reconcile the final invoice count and value to the source report.
Investigate the source documents and history, resolve each repeated ID with approval, then reconcile the final count and value. Do not delete or overwrite by guess.
Look for these in your work
- Compare the repeated records with the original invoices and source-system audit trail.
- Determine whether each is a valid amendment, two separate documents with a bad ID, or an accidental duplicate.
- Correct or exclude only under an approved rule and reconcile the final invoice count and value to the source report.
A new system rejects 17 customer records because of address format. The team suggests shortening names and deleting apartment numbers to speed conversion. Explain a safer response.
Give me a hint
Preserve information that identifies the customer; correct the format without silently losing meaning.
Compare my reasoning
- Identify the required field format and convert punctuation or spacing consistently while preserving address details.
- Check a sample and the full rejected-record list against source files after transformation.
- Record the transformation rule, retain original values and escalate records that still cannot be mapped.
Use a documented format conversion that preserves identifying details, compare results with the source and record exceptions; do not discard address information to force an import.
Look for these in your work
- Identify the required field format and convert punctuation or spacing consistently while preserving address details.
- Check a sample and the full rejected-record list against source files after transformation.
- Record the transformation rule, retain original values and escalate records that still cannot be mapped.
Common mistakes
- Deleting all records with similar names without review
- Changing source values with no correction log
- Assuming a successful file import means the data is valid
What can you explain now?
What should happen to a possible duplicate customer record?
Compare with the explanation
Compare reliable identifiers and review before merging or deleting
Two records can have similar names while representing different customers; preserve evidence and approval for changes.
After trying it yourself, choose your next review. This is your self-assessment.
Your review choice appears on Today. Sign in to sync it across devices.
Make it stick
What evidence should remain for each material correction made during data cleansing?
If you needed the solution, close it and solve the problem again from a blank page. A correct answer is stronger when you can explain the reason for each step.