Data quality Dimensions

Six dimensions, one question each.

Quality is a set of answerable questions about a record: is the field filled, is the value well formed, do the fields agree, is it one record, is it true of the world, is it current enough.

Completenessis the field filled?example checktax id present foractive suppliersValidityis the value well formed?example checkIBAN passes thechecksumConsistencydo the fields agree?example checkcountry matchesthe VAT prefixUniquenessis it one record?example checkone supplier pertax idAccuracyis it true of the world?example checkregistry confirmsthe nameTimelinessis it current enough?example checkbank detailsverified in 12monthsSIX DIMENSIONS, ONE QUESTION EACHScore of a record = checks passed ÷ checks that apply to it, per dimension and overallthe same arithmetic rolls up to a source, a domain and the estateeach check is a rule; each rule is an agent with a scope and a schedule — see the rule agents page
The six dimensions with the question each answers and one example check on a supplier record. Every check is a rule, and every rule is an agent.
6
dimensions
product
1
score per record, per dimension and overall
product
4
levels of roll-up: column, source, domain, estate
product
30
domains with sample checks per dimension in the catalogue
catalogue
The six, defined

What each one asks, and what a failing record looks like

The order is the order in which they are usually fixed: fill the fields, then make them well formed, then make them agree, then merge the duplicates, then verify against the world, then keep them fresh.

DimensionThe questionMeasured byA failing supplier record
CompletenessIs the field filled?Presence of a value in the fields that apply to this record type and statusAn active supplier with no tax id
ValidityIs the value well formed?Format, list membership, range and checksum rules on one fieldAn IBAN that fails the checksum; a country code outside ISO 3166
ConsistencyDo the fields agree?Row rules across fields of one record, and across records that refer to each otherCountry GB with a VAT prefix of DE; a payment term absent from the terms list
UniquenessIs it one record?Match rules on the domain keys; the cross-reference of source keysTwo supplier records with one tax id
AccuracyIs it true of the world?Verification against a registry, a postal service, a bank or a partner's own recordA registered name that the company registry spells differently
TimelinessIs it current enough?Age of the last verification or change against the freshness the consumer needsBank details last verified two years ago
The arithmetic

How a score is computed, and why it is weighted

A score is a ratio, so it can be compared across sources and over time.

For one record, the score on a dimension is the number of checks passed divided by the number of checks that apply to it. A check applies according to the record's type and status: a blocked supplier is excused the bank checks, an inactive material is excused the planning-parameter checks. The overall score of the record is the same ratio across all its applicable checks.

A column's score is the share of records passing the checks on that column; a source's score is the share across its columns; a domain's score across its sources. The roll-up is arithmetic, so a change in one rule shows at every level the next time the agents run.

Two weights refine the plain ratio. A field weight says how much a failure costs: a wrong bank account weighs more than a missing fax number, and the weights are set per domain by the owner. A consumer weight says who depends on the field: a field read by planning every night weighs more than one read by a quarterly report. The profiling page shows the weighted scorecard.

71%
completeness
74%
validity
54%
consistency
96%
uniqueness
88%
accuracy
91%
timeliness

One source's scorecard on the showcase estate: six dimensions, one number each. Consistency is the weakest, which points the first rules at the row checks. Showcase estate.

Dimensions by domain

What matters most differs by domain

The catalogue entry for each domain names the checks that carry the most weight. Three examples.

DomainDimension that carries most weightWhySample check
SupplierAccuracy and timeliness of bank and tax detailsPayments and compliance depend on them; changes are a fraud vectorBank account verified within twelve months; two-person rule on changes
Material and productCompleteness of planning and logistics attributesA material with defaults is planned wronglyLead time, minimum order and planner set for active plant materials
CustomerUniqueness and consistencyExposure and revenue add up only to one record per counterpartyOne record per tax id; billing and shipping countries consistent with the tax registration
Terms on this page

Applicable checkA rule that applies to a record given its type and status; excused checks leave the ratio untouched.
Field weightHow much a failure on a field costs, set per domain by the owner.
Consumer weightHow many processes depend on the field, and how often.
Roll-upThe same ratio computed for a column, a source, a domain and the estate.
VerificationA check against an outside service or registry that answers the accuracy question.
Questions

Why exactly six?

Six cover the questions the business asks; anything finer is a rule inside one of them. A lineage question (where did the value come from) is answered by provenance on every value, which is a property of the record itself.

Can a score be gamed by excusing checks?

Excusals are rules too, with an owner and a history, and the scorecard shows the excused count beside the score. A domain whose excusals rise is visible to the governance lead.

How often are scores computed?

On every run of the rule agents, which is nightly by default and on change for the fields that carry the most weight. Trends are kept per day.

See quality run on your own domain

Bring one domain and two systems. In a week the sources are profiled, the ten costliest fields are under rules, the first corrections are written back with their acknowledgements and the publish gate is on.