Enterprise Hierarchies and Structures

Top  Previous  Next

Theme Concepts

This theme includes subject areas that identify, name and define different hierarchies and classification structures used to describe retailer business unit groupings, merchandise hierarchy (categories), geolocation hierarchies along with relationships to the detail entities to be classified.

Retail enterprises are complex business and social systems.  They have to identify, name and describe a wide range of different people, places, things, processes and concepts and figure out how to organize them into a living, coherent economic organism.  One key mechanism for managing this complexity is to break these different entities into multi-leveled categories.  The classification of items using a hierarchy of named categories establishes a taxonomy.  Retail enterprises encompass a number of different kinds of taxonomies used to organize their thinking and actions around products, assignment of responsibility and authority, geography, positions and jobs, calendars, etc.   The ARTS ODM V7.0/7.1 contains a number of subject areas designed to help retailers set up taxonomies to help organize their decisions and activities in ways that they can be managed at an aggregate AND detailed level based on the business problem they need to solve.

Major Retail Taxonomies

Item Taxonomy

Item taxonomy addresses the classification of all of the services and tangible merchandise items sold by a retailer.  In ARTS, the Item is the single lowest common denominator that enables retailers to set up multiple product taxonomies.  Also, each item must be uniquely identified across the entire enterprise -- any duplicate or redundant items will break the data model.

The Product (Items and Services Sold) topic presents a detailed explanation of how ARTS represents product taxonomies using a set of item hierarchy entities.  The Logical 10200 - Enterprise - Merchandise View presents an entity model that shows the product and enterprise responsibility delegation taxonomies and how they fit together.

Enterprise Responsibility Delegation Taxonomy

As discussed earlier, retail enterprises are complex business and social structures.  The ability to delegate authority and responsibility is an important mechanism to enable retail businesses to scale up.  In ARTS, the concept of authority and responsibility delegation is represented in the Logical 10115 - Enterprise Business Unit Groups Hierarchy View subject area.  The idea is to decompose a retail enterprise into BusinessUnitGroup entities which each embody a defined scope of authority and responsibility to plan, execute and measure retail business activity.  Most retailers choose to align their delegation structures along geographic dimensions which can blur the notion of geographic taxonomy (discussed next) with delegation taxonomy.  In ARTS these are two distinctly different views of the enterprise.

The concept of separation is critical when developing measurement and reporting systems.  Profit and lost accountability should be aligned with the BusinessUnitGroup taxonomy structure because its express purpose is to model the delegation of responsibility and authority to named entities (BusinessUnitGroup) that make up the retail enterprise.

Geographic Taxonomy

The world is a big place and it helps to decompose it into a hierarchy of named regions and areas.  As an example a world wide retailer may break the world down by continent, country, primary country subdivision, etc.  ARTS does this using international ISO-3166-1 (a standard for identifying and naming countries) and ISO-3166-2 (a standard for identifying the primary political subdivisions of countries into states, provinces, territories, prefectures, etc.).  These two standard geographic area designations are related hierarchically as country -> primary political subdivision.  ARTS allows retailers to embed these standards into their own geographic taxonomies.

For example in the United States, the Census Bureau provides a very useful decomposition of the US down into regions, divisions, states, counties and several lower levels to the street level as illustrated here.  This provides a ready-made way for retailers to divide their business geographical areas down in a way that aligns their data aggregations with US Census statistics.

Figure 71 - US Census Geographic Taxonomy Example


ARTS provides multiple geographic hierarchies in part because each country will use a different taxonomy schema.  The lowest common denominator that ties these different taxonomies together is the set of entities Site, GeoLocation, BusinessUnitSite and BusinessUnit.  These entities create the atomic-level integration between the Enterprise Responsibility Delegation Taxonomy discussed earlier and the Geographic Taxonomy discussed here.

Geographic taxonomies provide a different reporting view from the delegation taxonomy.  Geographic taxonomies enable retailers to develop insights into sales and factors behind sales (as well as operating expenses) for decision support purposes. Decision support as used here means evaluation and, if necessary modification of business operations and strategy to improve performance.  This is different from the accountability purpose discussed under Enterprise Responsibility Delegation Taxonomy.

The Logical 10108 - Enterprise Geolocation presents the ARTS entities used to implement a geographic taxonomy for retail enterprises.

Position Taxonomy (the Organization Charts)

The classic representation of an enterprise organization consists of a hierarchical arrangements of positions based on reporting relationships.  That view of the retail enterprise is supported in the Logical 11100 - Worker - Macro View.  In this view, the emphasis is on supporting the human resource functions of a retail enterprise.  It is different from the Enterprise Responsibility Delegation Taxonomy discussed earlier in that it emphasizes position reporting links rather than profit and loss responsibility.  Also the Position Taxonomy is more granular because it addresses levels down to individual positions not business units (i.e. stores, warehouses, administrative offices which are higher levels).

Calendar Taxonomy

A calendar is not typically thought of as a taxonomy.  But it is in fact the decomposition of time periods into multiple, lower level time periods.  The Gregorian calendar breaks Centuries into decades, decades in to years, years into months, and months into days.  Retail calendars like the NRF 4-5-4 calendar break years into seasons, seasons into quarters, quarters into periods (similar to calendar months), periods into weeks and weeks into days.  ARTS represents calendars as a hierarchy of named time periods.  ARTS supports multiple calendars with the understanding that the common denominator is the business day. Logical 06100 - Calendar View presents the calendar hierarchy model used by ARTS.

Future Taxonomy Extensions

As ARTS extends its coverage further into customer behavior, promotion design, planning and execution, social networking, and beyond new taxonomies will be added to the Data Model.  Also, these taxonomies will serve as the foundation for creating multi-tier Data Warehouse dimensions used to support roll-up and drill-down reporting.  In addition to adding taxonomies, ARTS will address the need to enhance the semantic aspects of retail information modeling - particularly as it relates to integrating and aligning XML schemas using resource definition/ontological models to supplement the relational models in ARTS ODM V7.0/7.1.

POS Department

The Logical 10300 - Enterprise - POS Department View subject area presents a less formal hierarchy used to decompose a business unit (store) into areas based on common point of sale attributes.  The name "department" in this context is a semantically overloaded term that designates both an area and a category of merchandise.   The merchandise category may or may not correspond to the formal product taxonomy of the retailer.  POS Department is DIFFERENT from Location and MerchandiseHierarchyGroup even though it may cover both areas.  A typical example is a retail grocery store.  The retail grocery store (BusinessUnit) may have a meat POS department, a diary POS department, a Produce POS department, a general merchandise POS department, a frozen food POS department, etc.  The names of these POS departments reflect categories of merchandise but actually reflect selling floor areas (Locations) designated around different point of sale workstations.  Merchants typically will use the product taxonomy as the basis for making buying decisions.  In-store assignment of responsibility for front-line management will typically revolve around the POS Department.

The POS department hierarchy is implemented as a self referencing foreign key.  Its structure is simpler than the join table strategy used to model product and location taxonomies.  The treatment of POS department in the ARTS Data Model is likely to change in the future as planograms are addressed.  Planograms represent a highly structured merging of intra-site locations with merchandise hierarchies down to the item/SKU level of granularity.  They represent a much more formal treatment of this blending of location-merchandise than is reflected in the current POS Department View.