Domain catalogue Products

Product classification

The trees that group products for sales, planning, sourcing and reporting, kept as versioned hierarchies with one leaf per golden item.

SOURCES, READ IN PLACESAP S/4HANAERPOracle Fusion CloudERPSalesforceCRMSnowflakelakehouseASSEMBLED1 · standardise2 · match on the domain keys3 · survive field by field4 · place in the hierarchies5 · run the column and row rules6 · publish the golden viewGOLDEN PRODUCT CLASSIFICATIONProduct classificationone id · every source key keptREAD BYGrids, pivots and drill-downswith the transactionsAPIs, SQL and agentsone policyWrite-back to the systemsacknowledged
Where the product classification record comes from, how it is assembled, and who reads it. Other systems have their own page.
3 days
of configuration to a production-grade version
estimate
4
attribute groups in the golden record
catalogue
2
match keys
catalogue
4
hierarchies the record sits in
catalogue
3
kinds of transactions that refer to it
catalogue
What it is

Today, and with the golden record

Today the sales hierarchy in the CRM, the product hierarchy in the ERP and the category tree in the web shop share nothing but the intent; a reorganisation takes a quarter.

With Lima the classification hierarchies are read where they live, matched on node code per hierarchy and name within a parent (fuzzy), assembled field by field and published as one golden record per real product classification, with every source key underneath. From any node to its leaves, to the items, to the transactions of those items, under any version are one click away, because the transactions sit in the same bed.

The golden record

4 attribute groups, and which source usually wins each

Survivorship is per field; the winning source below is the usual one, and a steward's verified edit outranks it.

Hierarchy nodescode, name, level, parent, valid dates, owner
wins: stewards
Leaf assignmentsgolden item → node per hierarchy
wins: ERP and stewards
Attributes by nodeplanning parameters, tax groups, margin targets inherited downwards
wins: planning and finance
Crosswalksnode codes per system
wins: generated and maintained
How records match

The keys, in order

node code per hierarchyname within a parent (fuzzy)

A leaf can sit in several hierarchies at once; each hierarchy has its own version history.

Pairs above the merge line merge; pairs between the lines go to the steward with the two records side by side and the transactions of each; the rest stay apart. Every merge can be undone. The golden records page explains the steps.

Where it sits

4 hierarchies

Several structures for the same record, each kept as versions with dates.

HierarchyLevels and use
Salesdivision → family → line → product
Planningplanning family → planning group
Sourcingcategory → sub-category
Statutorycommodity codes and tariff chapters

A pivot can roll the record's measures up any of them; a draft version is validated against the facts before it goes live. See hierarchies.

Rules at three levels

Sample rules from the template

Written in one language, tested on live rows before they are saved.

LevelRules
Columna level is between 1 and the hierarchy's depth
a node code is unique within a hierarchy version
Rowevery node except the root has exactly one parent per version
a node with children has no direct leaves in a strict hierarchy
Businessevery active item sits in exactly one leaf of each mandatory hierarchy
revenue by node equals the sum of its children's revenue
a moved node keeps its history under the old version
What refers to it

Transactions, measures and drill-downs

Transactions in the bedRead from
Sales by iteminvoiced lines
Forecast by planning groupplanning system
Spend by categorypurchase order lines
revenue by node and monthforecast accuracy by planning groupspend by categoryitems per leaf

Measures above are available in the pivot on any hierarchy level, by month or week, with calculated measures beside them. Drill-down opens from any node to its leaves, to the items, to the transactions of those items, under any version.

Security pattern

What is masked, who edits, who reads

Default policy for the domain

Hierarchy edits by the owning function; everyone reads; planning parameters by node editable by planners. The twelve layers apply on every door; see security.

The lookup API

Pass the keys you know, get the golden record

request
POST /api/v1/golden/hierarchy/product-sales/nodes
{
  "as_of": "2026-04-15",
  "node": "PWR-SUP"
}
response
200 OK
{
  "version": "v13",
  "children": 6,
  "leaves": 412,
  "revenue_ytd": 18240000
}

The same call creates the record when nothing matches; related facts are expanded through the cross-reference; page sizes and limits are published per endpoint. See the headless API.

On your systems

One page per system

The objects with their real names, the keys in that system's fields, the write interface and what to set up first.

A day with it

Illustrative

Names, places and figures are illustrative or anonymised; the pattern is the catalogue's.

A consumer-goods company moved 1,200 products between families for a new go-to-market. The new version went live on a Monday; reports for the previous quarter still rolled up the old way, and the board pack showed both.

illustrative scenario · Product classification
Terms on this page

VersionA hierarchy as it stood between two dates; history rolls up under the version valid at the time.
Ragged hierarchyA tree whose branches have different depths.
Inherited attributeA value set on a node and read by every descendant.
Golden recordThe one record for a real product classification, assembled from the source records, with every value carrying its source, rule and time.
Cross-referenceThe list of source keys that belong to one golden record.
Questions

Does this domain copy data out of the systems?

The golden record is a governed view over rows read in place; a governed replica in your lake is used only where a source is too slow or too limited to query live, and the column says so.

How long to the first golden view?

One working day for the first two sources; 3 working days to a production-grade version with rules, hierarchies, security and write-back tested.

Which industries use it most?

Consumer goods, High-tech and semiconductor, Industrial manufacturing, Telecom. The industry pages show the domains each starts with.

See your first domain built on your data

Bring two systems that hold the same customers, suppliers or materials. In one working day we connect them, run the match on live rows, review the queue with your steward and publish the first golden view with its API.