Service · Data and analytics

Data and reference data management

Reports fail to match not because the system is bad but because one counterparty has three records and a product unit is sometimes a piece and sometimes a pack. The cure is not software but rules and named owners.

Submit a request What is included

Duplicates

found and merged

Rules

who creates records and how

An owner

for every reference list

Checks

at entry, not afterwards

What is included

Technical tools help, but the foundation is organisational: every reference list needs a person responsible for it.

Discuss the scope

Reference data audit

Duplicates, empty mandatory fields, and inconsistent names and units.

Maintenance rules

How an entry is named, which fields are mandatory and who may create new ones.

Merging duplicates

With the links preserved: documents must not lose their references.

Owners

The owner of a reference list is a specific person, not a department.

Checks at entry

Validations that prevent creating a duplicate or skipping a mandatory field.

Regular review

A quality report: how many new duplicates and where fields are empty.

How the work runs

The audit takes one to two weeks. Cleaning up depends on volume and usually runs over several months in stages.

Audit

We measure the scale: how many duplicates and how many records contain errors.

Rules

Agreed with the people who actually maintain the data rather than handed down from above.

Cleanup

Merging duplicates by priority: first what gets in the way of reporting.

Control

Validations at entry and a regular data quality report.

Cleaning up reference data without rules for maintaining it is pointless. Six months later the duplicates are back, because a salesperson creates a new record after failing to find the old one. Rules and entry checks first, then the cleanup, otherwise the work is done twice.

Frequently asked questions

Why do duplicates appear at all?

Because finding an existing record is harder than creating a new one. A salesperson searches for a company name, fails to find it because of quotation marks and spaces, and creates another. The cure is decent search and a ban on creating records without a check.

Can duplicates be merged automatically?

Candidates can be found automatically, yes. Merging is better done with a human check: similar names do not always mean the same thing, and an incorrect merge ruins the document history.

How long does this take?

The audit is quick; the cleanup depends on volume. For a counterparty list of several thousand records it is weeks of work, but it can run alongside normal operations, starting with the active entries.

Other services in this area

The whole area

We will put your data in order

Tell us where the reports disagree and which reference lists raise questions. We will start by auditing the scale of the problem.

Phone
+7 (495) 978-63-36

Email
office@apply.ru

Office
Moscow, 108 Profsoyuznaya St.

I consent to the processing of personal data