Legacy catalog
Part numberAC820825
Technical fieldsIncomplete
Solutions / Catalog AssemblyLegacy strata → clean dataset
Product data cleanup
Clean Legacy Product Data Before It Creates New Problems
Old catalogs accumulate history.
Supplier conventions change, employees create new fields, units drift and incomplete records remain because the business has learned to work around them.
That data may still support daily operations, but copying it into a new PIM, ecommerce platform or AI workflow can preserve years of inconsistency.
ENRIVAQ supports catalog cleanup through audit, normalization, enrichment and validation before the cleaned data is exported for continued use.
Local fields
Outside Ø8.2 cm
CategoryUnmapped
Inherited values
Weight1.74 / 2.60 kg
SourceUnclear
Audit · normalize
enrich · validate
enrich · validate
Controlled recordFan impellerAC820825
IdentityStructured
Canonical fieldsTarget schema
Missing valuesRemain unresolved
Validation stateVisible
Reference cleanup case · incomplete source record → controlled product record.
Typical legacy problems
Legacy data usually contains several kinds of inconsistency at once
The problem is rarely one bad column. Older product datasets often combine structural and content issues. These archive drawers establish the starting scope for an audit.
Duplicate records
Duplicate or near-duplicate product records can fragment one product identity.
StructuralMatching
Field drift
Outdated supplier-specific attributes and duplicate fields accumulate over time.
StructuralNaming
Mixed conventions
Different units, category placement and unclear legacy values break comparison.
UnitsCategories
Open gaps
Required attributes remain empty while weak or copied descriptions hide the gap.
ContentMissing
Audit
Understand the current catalog before changing it
Cleanup should begin with an assessment of the existing data. This prevents the project from becoming a series of uncontrolled bulk edits.
Recurring patterns first
The audit identifies incomplete categories, inconsistent units, duplicate fields and records that need research rather than formatting.
IdentitySchemaUnitsCompletenessSources
Categories with incomplete required fieldsMeasured in auditEnrich
Attributes stored in mixed unitsMeasured in auditNormalize
Fields duplicating one anotherMeasured in auditNormalize
Records needing product researchMeasured in auditResearch
Normalize
Define the target structure before bulk cleanup begins
Existing values can be reshaped only after field meanings, approved units and naming conventions are agreed.
4.2 cm / 42mm
mixed unitsOD / Outside Ø / diam_ext
mixed labelsFan wheel / Gebläserad
mixed vocabularyPart No. / OEM / MPN
mixed identifiersTarget
mapping
mapping
outer_diameter_mm42 mm
outer_diameterApproved label
product_typeFan impeller
manufacturer_part_numberAC820825
Enrich
Keep formatting cleanup separate from information enrichment
Where relevant sources are available, the enrichment workflow can research additional product information and add structured attributes or content to the cleaned record.
OEM / MPNAC820825
Product typeFan impeller
CompatibilityMissing · research required
MaterialMissing · evidence required
DescriptionMissing · grounded content only
Validate
Check the cleaned record before treating it as finished
Bulk transformations can create new errors if mappings, units or product identities are wrong. Uncertain cases should remain reviewable.
MappingsA wrong mapping moves values into the wrong canonical field.Check
UnitsA wrong conversion changes the technical meaning of a value.Check
IdentityA wrong product match applies correct data to the wrong record.Check
ConflictsDisagreement must remain visible until resolved.Hold
Before / after
Show the catalog record before and after cleanup
One real legacy product makes the transformation visible without inventing customer results.
Fan impeller AC820825
ManufacturerKverneland
OEM / MPNAC820825
Technical attributesIncomplete
CompatibilityIncomplete
Legacy record with limited structure.
Kverneland AC820825 Fan Impeller
EAN8716106986118
CompatibilityOptima / Optima HD
MaterialMetal
Weight conflictReview / hold
Accepted fields only. No completeness or recovery metric is claimed.
Export clean dataset
Return a predictable dataset to the next system
The result should follow the agreed target schema and move into the next verified stage of the product-data workflow.
Supported export methods must match the actual implementation.
Catalog binderControlled output
Schema alignedProduct identityStructured
Canonical attributesAccepted values only
Unresolved valuesSeparated for review
Source contextRetained where implemented
Clean a representative category before touching the whole catalog
Start with a sample that contains the same legacy problems seen across the wider dataset.