/

The basics

A simple introduction to master data and why things get messy day to day

Most people working in a municipality, region or larger organisation run into problems every day that come from poorly ordered master data — often without knowing that is what it is.

Here we explain what master data is, why the problems arise, and four rules of thumb for solving them.

From chaos to automation

  1. Chaos

    Fragmented systems and conflicting concepts

    Information sits in separate systems and silos. Departments keep their own registers and use different words for the same thing, with no shared sources.

  2. Insight

    Information building blocks

    People agree on what the basic concepts mean – organisation, unit, person, employee, service and configuration item – whatever system is used.

  3. Patterns

    Information patterns & the organisation tree

    The building blocks are connected with clear relationships. The organisation tree is the hub that employees, roles, processes and services link to.

  4. Structure

    Business rules, code lists & source governance

    Each information set gets a designated source, and rules and code lists decide what applies and who may change it.

  5. Automation

    Understanding that enables change

    With the master data in order, access, approvals and integrations can work without manual workarounds.

Recommended starting point: the organisation tree

The organisation tree is the backbone that holds the entire organisation together. It is through this hierarchy that business goals cascade down and reporting aggregates upwards to responsible managers, based on information about employees, process roles, approval chains, cost centres, IT components and services that all link back to the tree.

Does this sound familiar?

  1. 1

    The same information is registered in different systems.

    Employees, units or suppliers are typed in again in every business system instead of being taken from a shared source.

  2. 2

    Hard or impossible to integrate, even though the systems have integration features.

    The systems have technical APIs, but their data models do not speak the same language. Getting two systems to agree on what a “unit” or an “assignment” is takes custom work that often breaks at the next update.

  3. 3

    Business goals delivered to and from managers are handled manually and with great uncertainty.

    When the organisation tree and line of managers differ between systems, goal follow-up and reporting end up in spreadsheets, with a high risk of wrong figures.

  4. 4

    Time-consuming manual work to give new staff access to the right resources quickly.

    Access, email, key cards and work tools are arranged through email chains and manual approvals, taking time from IT, HR and the new employee.

  5. 5

    System content keeps differing, with managers either registered manually or different in each system.

    When a manager leaves, changes role or starts, it does not flow through automatically. Different systems show different managers, and errors are fixed by hand afterwards.

What these situations share is that the same information is registered in several places, in different ways, with no clear rule for who decides what applies.

What is master data?

Master data is the information the whole organisation shares and that rarely changes minute to minute:

  • Who we are: units, departments, administrations and workplaces.
  • Who works here: people and their employments, assignments or other engagements.
  • Who we deal with: customers, residents, suppliers and partners.
  • Where things are: properties, visiting addresses and geographical areas.

Transactions — a case, an invoice, a decision — come and go. Master data is the building blocks those transactions rely on.

The organisation sees the world from the inside out. The organisation tree consists only of units. A person is never inside the organisation, just as an external company is not. People and external parties are connected through an agreement – employment, assignment, procurement contract – and only that connection creates positions, process roles and digital identities.

Four rules of thumb for order and automation

Getting master data in order is not about buying a new all-encompassing IT system, but about a shift in four steps:

1. Define how the information fits together

Decide how information should fit together to enable and optimise automation. This is done with information patterns, building blocks and business rules. Without a shared structure, systems do not know what to automate against.

2. Establish the processes and sources

Decide which process creates each information set and, where possible, which system creates, changes and removes it — the master data source or sources. Other systems read.

3. Set up an organisation that checks the rules over time

A one-off clean-up project is not enough. You need an organisation with responsibility and mandate to check continuously that information follows the agreed rules.

4. Adapt the receiving systems (the ordeal)

Systems that subscribe to the information sets need to adapt to the shared structure. This is the ordeal every organisation has to go through. Above all it concerns HR and finance systems, which often consider themselves to hold the digital description of the organisation.

A small plain-language glossary

Master data
Shared information used in several processes and systems, such as person, organisation and place.
Master data source
The system or process that has the right to create, change and remove a particular piece of information.
Subscriber
A system that receives and uses master data but may not change it.
Information pattern
An agreed blueprint showing how concepts relate and which rules apply between them.
Code list
A fixed list of approved values, such as employment types or county codes, so all systems mean the same thing.