The Zachman Framework is the oldest formally documented enterprise architecture framework. John Zachman published the original classification schema in 1987 in an IBM Systems Journal article titled “A Framework for Information Systems Architecture.” It predates TOGAF, predates ArchiMate, and predates the term “enterprise architecture” as it is used today.
Zachman’s contribution was conceptual rather than procedural. He observed that any complex artefact: a building, an aeroplane, an information system: could be described by answering six interrogative questions from six different audience perspectives. The result is a 6×6 matrix: 36 cells, each representing a distinct class of architectural description.
Understanding what Zachman is: and what it is not: is essential before deciding how to use it in a Sparx EA implementation.
The interrogatives form the columns of the Zachman matrix. They are the fundamental questions that describe any complex artefact:
Each column has a primitive model type that represents it at its most abstract: the What column is represented by an entity-relationship model at the conceptual level; the How column by a function model; the Where column by a network model; the Who column by a role model; the When column by a lifecycle model; the Why column by a motivation model.
The perspectives form the rows and represent the distinct audiences who need architecture descriptions:
Each cell in the matrix represents a unique class of architecture description. Cell (Row 1, What) is the executive-level list of things the enterprise cares about: a data inventory at its most abstract. Cell (Row 4, How) is the physical process specification: the specific technology that implements business functions in engineering-level detail.
Every architectural artefact an organization produces belongs in one or more cells of this matrix. The value of Zachman is not in filling every cell: most organizations will never need all 36: but in having a consistent vocabulary for describing where an artefact sits and what it describes.
This distinction is critical and frequently misunderstood.
Zachman is not a process. It does not tell you what to do first, what to do next, or how to sequence your architecture work. There is no Zachman method.
Zachman is not a methodology. It does not define deliverables, phases, governance mechanisms, or stakeholder engagement approaches. It does not specify how to run an Architecture Review Board or how to produce a Transition Architecture.
Zachman is not a modeling language. It does not define notation, diagram types, or the syntax for representing architecture content.
Zachman is a classification schema. It is a taxonomy: a framework for organizing and labeling what already exists, or what needs to exist, in an EA artefact library. It answers the question “what kind of description is this?” not “how do we produce it?”
TOGAF won the battle for EA methodology adoption. ArchiMate won the battle for EA modeling language adoption. Zachman was never competing in either of those battles, despite frequent comparisons. It operates at a different level of abstraction, and its value in 2026 is precisely in that distinction.
Zachman’s relevance today is not as an operational framework: it is as a conceptual map.
As EA repositories mature, they accumulate hundreds or thousands of artefacts across multiple frameworks, methodologies, and modeling languages. A large Sparx EA repository might contain ArchiMate diagrams from TOGAF ADM work, SysML models from systems engineering projects, BPMN processes from operational analysis, UML class diagrams from data architecture work, and capability maps from strategic planning exercises.
Zachman provides a classification vocabulary that cuts across all of these. An ArchiMate Application Layer diagram, a TOGAF Application Architecture document, and a component inventory spreadsheet all describe the same cell in the Zachman matrix (Row 3 or Row 4, What column). Knowing this means knowing they are substitutable descriptions of the same thing: useful for governance, deduplication, and coverage analysis.
When Zachman is used as an overlay classification, gaps become visible. If the executive perspective has strong content in the What, Why, and Who columns but almost nothing in the When and Where columns, that is an architectural coverage gap that leadership cannot see because there are no artefacts to reference.
Organizations running multiple frameworks simultaneously: a common reality in large enterprises: benefit from a cross-framework classification schema. Zachman is framework-agnostic, which means it can classify TOGAF outputs, DoDAF outputs, NAFv4 outputs, and custom framework outputs using the same vocabulary. This makes Zachman a useful lingua franca for multi-framework governance.
As EA repositories are connected to AI systems via EA GraphLink and MCP servers, the quality of repository metadata determines the quality of AI outputs. Zachman classification as a tagged value on artefacts gives AI systems a structured dimension for filtering and contextualising queries: “Show me all Row 2 (Business Management) artefacts in the How column that relate to customer onboarding” becomes a precise, navigable query rather than a keyword search.
Sparx EA supports Zachman implementation through two primary mechanisms: the built-in Zachman MDG extension and the use of tagged values as a classification overlay.
Sparx EA ships with a Zachman MDG (Model Driven Generation) extension that implements the 6×6 matrix as a native framework view. The extension includes:
To activate the Zachman MDG in Sparx EA, navigate to Specialize → Technologies → Manage Technologies and enable the Zachman Framework technology. Once active, Zachman diagram types appear in the New Diagram dialog.
The more practically useful Zachman implementation is as a classification overlay applied to existing artefacts. Rather than rebuilding an existing repository in Zachman terms, architects add Zachman classification tagged values to existing elements and diagrams.
A practical implementation adds two tagged values to every significant artefact in the repository:
These two tagged values, applied consistently, enable Zachman-based filtering without disrupting the existing modeling approach. An ArchiMate Motivation diagram receives Zachman_Perspective = Executive and Zachman_Interrogative = Why. An Application Layer diagram receives Zachman_Perspective = Architect and Zachman_Interrogative = What. The existing modeling work is unchanged; the Zachman classification is an additional metadata layer.
This overlay approach is the recommended Sparx Services implementation for organizations that have existing ArchiMate or TOGAF-based repositories. It adds governance value without requiring rework.
For mature implementations, the Zachman ontology can be used as a formal cross-reference schema: a metadata structure that links artefacts across different package hierarchies by their Zachman classification. This supports queries like: “Find all artefacts that describe the Engineer perspective of the How column, regardless of which project or package they live in.”
In Sparx EA, this is implemented using the Advanced Search (SQL-based) or EA’s built-in model search filtered on tagged values. With EA GraphLink, it becomes a structured API query that can populate a Power BI matrix view showing Zachman cell coverage across the entire repository.
These three are frequently compared and the comparison generates confusion because they operate at different levels.
| Dimension | Zachman | TOGAF | ArchiMate |
|---|---|---|---|
| Type | Classification schema | Process methodology | Modeling language |
| What it defines | What kind of description an artefact is | How to produce architecture | How to draw architecture |
| Sequence | None | ADM phases | Not applicable |
| Deliverables | None | Architecture documents | Diagrams/models |
| Governing body | Zachman International | The Open Group | The Open Group |
| Certification | Zachman Certified EA | TOGAF certified | ArchiMate certified |
| Relationship to others | Classifies TOGAF and ArchiMate outputs | References ArchiMate as notation | Referenced by TOGAF |
Zachman classifies. TOGAF processes. ArchiMate models. Used together, they form a comprehensive EA capability: TOGAF tells you what to do and in what order; ArchiMate gives you a language to describe what you find; Zachman gives you a taxonomy to organize what you produce.
The organizations that get the most value from Zachman in 2026 are those that have significant existing TOGAF and ArchiMate adoption and need a governance layer that spans their artefact library without disrupting it.
The most common mistake is using Zachman as a work breakdown structure: attempting to “fill in” cells sequentially as an architecture program. This leads to enormous amounts of modeling effort producing artefacts that serve the matrix rather than the business. The matrix is a classification tool, not a production checklist.
No organization needs all 36 cells populated. Most organizations work intensively in 8-12 cells and should have reference content in another 6-10. Attempting comprehensive coverage of the full matrix is a distraction.
Some architects implement Zachman as their primary modeling approach, creating Zachman-typed elements instead of ArchiMate elements. This loses the semantic richness and standardized notation of ArchiMate. Zachman should overlay ArchiMate, not replace it.
Applying Zachman classification to a repository that has no consistent MDG governance produces classification of noise. The underlying content must be reasonably well-structured before the overlay adds value.
Zachman classification and artefact governance are most commonly addressed in a Sparx Services Amplify engagement: which focuses on maximizing the value already present in an existing EA repository.
A typical Zachman-related Amplify workstream includes:
Amplify engagements start at $45K and are scoped after a Discover engagement. Organizations with large repositories or multiple conflicting framework approaches benefit most.
What is the Zachman Framework in simple terms?
The Zachman Framework is a classification schema: a 6×6 matrix that provides a consistent vocabulary for describing what any architectural artefact is about. The six columns (What, How, Where, Who, When, Why) describe the type of question an artefact answers; the six rows (Executive through Technician) describe the audience perspective it serves. It is not a method or a process: it is a taxonomy for organizing EA artefacts.
Is the Zachman Framework still relevant in 2026?
Yes, for specific purposes. Zachman is not relevant as an operational framework: TOGAF is the dominant EA methodology and ArchiMate is the dominant modeling language. Zachman is relevant as a governance and classification tool for organizations with large, multi-framework EA repositories that need a cross-framework taxonomy. Its value has shifted from foundational framework to governance overlay.
How does Zachman differ from TOGAF?
Zachman classifies; TOGAF processes. TOGAF (The Open Group Architecture Framework) provides a methodology: the Architecture Development Method (ADM): that tells architects how to produce architecture: what phases to run, what deliverables to produce, what stakeholders to engage. Zachman provides no methodology; it only describes where any produced artefact sits in a classification space. They are complementary, not competing.
Does Sparx EA have built-in Zachman support?
Yes. Sparx EA includes a Zachman MDG extension that can be activated from the Technology Manager. It provides Zachman diagram frames showing the 6×6 matrix, Zachman stereotypes, and a Zachman-specific toolbox. For most mature implementations, Sparx Services recommends supplementing the built-in support with a tagged value overlay on existing ArchiMate and TOGAF content rather than rebuilding content in Zachman-native form.
Can I use Zachman alongside ArchiMate in Sparx EA?
Yes: and this is the recommended approach. ArchiMate provides the modeling language and semantic richness; Zachman provides the classification overlay. In practice this means adding Zachman_Perspective and Zachman_Interrogative tagged values to existing ArchiMate elements and diagrams. The two frameworks operate at different levels and do not conflict.
How does Zachman classification help with AI tools like EA GraphLink?
When Zachman classification is consistently applied as tagged values in the Sparx EA repository, EA GraphLink can expose Zachman perspective and interrogative as filterable dimensions in its API. This enables structured AI queries: “show me all Engineer-perspective How-column artefacts related to order processing”: that return precise results rather than keyword-matched noise. Zachman classification is a repository governance investment that directly improves AI query quality.
What is the difference between Zachman rows and TOGAF architecture domains?
They are different dimensions. TOGAF architecture domains (Business, Data, Application, Technology) correspond roughly to interrogative columns in Zachman (What, How) at different levels of abstraction. TOGAF domains describe the subject matter of architecture; Zachman rows describe the audience perspective. An Application Architecture document in TOGAF could appear in Row 2 (Business Management), Row 3 (Architect), or Row 4 (Engineer) depending on its level of abstraction and intended audience.
Which Sparx Services engagement covers Zachman implementation?
Zachman classification and artefact governance are typically addressed in an Amplify engagement ($45K–$160K), which focuses on maximizing the value of an existing EA repository. Amplify includes MDG profile extension, artefact classification, coverage analysis, and EA GraphLink integration. If the repository is not yet well-governed, a Deploy engagement to establish MDG foundations is a prerequisite. Contact Sparx Services to discuss the right starting point.
If your Sparx EA repository contains significant content but lacks the governance structure to make it navigable, reportable, or AI-ready, an Amplify engagement is the right starting point.
Sparx Services will audit your repository, extend your MDG profiles with classification metadata, apply Zachman coverage analysis, and configure EA GraphLink to surface the results in Power BI dashboards and AI queries.
Talk to Sparx Services about an Amplify engagement
Amplify engagements start at $45K. Scoped after an initial Discover workshop.
Talk to a Sparx Services architect about where your organization is on the journey and what the next stage looks like.