Data cleansing

Procurement data rots in predictable ways. "Acme Industries," "ACME IND," and "Acme Industries LLC" enter the vendor master as three suppliers, usually because three plants onboarded the same vendor separately. Units of measure flip between pieces and thousands. The same part carries different item numbers in two ERP instances after an acquisition. None of this blocks transactions, which is exactly why it accumulates: the PO still clears, the invoice still pays, and the damage only appears when someone tries to analyze.

The damage is concrete. A supplier that looks like $400K across three records is actually a $1.2M relationship negotiated as three small ones. Two plants buying the same bracket under different item numbers never see the price gap between them. Spend analysis on dirty data produces precise-looking answers to the wrong questions.

Cleansing is the corrective pass: match and merge duplicates, standardize naming and units, repair categories. It differs from data enrichment, which appends new external attributes rather than fixing existing ones, and from master data management, the ongoing governance that keeps records from drifting again. Cleansing without that governance is mopping the floor with the tap running.

Frequently asked questions

What is data cleansing in procurement?

Data cleansing in procurement is the work of detecting and correcting errors in supplier, item, and transaction records: deduplicating supplier entries, normalizing names and units of measure, fixing miscoded categories, and reconciling records across systems. Clean records are the prerequisite for trustworthy spend analysis, supplier consolidation, and price comparison.

Why does procurement data go bad in the first place?

Procurement data rots in predictable ways because errors do not block transactions. The same vendor enters the master file three times when three plants onboard it separately, units of measure flip between pieces and thousands, and part numbers diverge across ERP instances after an acquisition. The PO still clears and the invoice still pays, so the damage only appears when someone tries to analyze the data.

What problems does dirty supplier data cause?

Dirty supplier data fragments the picture of what a company actually buys. A supplier that looks like $400K across three duplicate records is really a $1.2M relationship negotiated as three small ones, and two plants buying the same bracket under different item numbers never see the price gap between them. Spend analysis run on dirty data produces precise-looking answers to the wrong questions.

What is the difference between data cleansing and data enrichment?

Data cleansing fixes what a company already knows about its records: merging duplicates, standardizing names and units, repairing category codes. Data enrichment appends new external attributes, such as financial health scores or certifications, that the company does not hold internally. Cleansing also differs from master data management, which is the ongoing governance that keeps records from drifting again.

Is data cleansing a one-time project?

Data cleansing without ongoing governance is mopping the floor with the tap running. A cleansing pass corrects the records that exist today, but new duplicates and inconsistencies accumulate unless master data management rules control how records get created, so the corrective project needs to be paired with governance that keeps data clean.