Start With Purpose

The four purposes — start with one

The guide says an EA Capability serves “one or more” purposes; a mature practice often spans several. But it earns them in sequence, starting from one clear, sponsored purpose. None is more “real EA” than the others — the failure isn't serving one purpose or four, it's never naming any, or claiming all four before you can sustain one.

The four death traps

Naming a purpose is necessary but not sufficient. The guide catalogues how an EA Capability talks itself into a reboot — four recur often enough to name outright, and each has a cheaper fix than rebuilding the practice from scratch.

1. Trap 1

   Advising mistaken for deciding

   the most common trap

   EA advises; executives decide. Confusing “supporting a decision” with making it is, in the guide's words, “simply wishful thinking.” The fix: measure the value of good advice — recommendations accepted, risks surfaced early — and keep decision rights and outcome accountability where they belong.

2. Trap 2

   Standards over transformation

   an output, not a reason to exist

   The guide's most quotable finding: teams that deliver organizational transformation are statistically more successful than teams that focus on standards, reference architectures, and governance structures. A team known only for saying no is one a new CIO can cut without noticing a gap.

3. Trap 3

   Value before buy-in

   the revivalist trap

   Delivering value before securing a sponsor “has wrought many challenges” — the leader acts on assumption, and when the sponsor doesn't value the result, the team has burned scarce resource failing again. Secure financial allocation and resource commitment first.

4. Trap 4

   No named purpose

   the root cause

   The message that an EA Capability is a function of context and purpose “is often lost.” Without a named purpose the team has no test for what to work on, drifts toward documents nobody asked for, and can't answer the question that decides its funding: what would the enterprise do if this capability didn't exist?

The Purpose Test

Ten questions to answer before you establish — or reboot — an EA practice. They come straight from the guide's own guidance on context, purpose, and success measurement. Score one point per honest “yes.” Below seven, you are building on the same ground the last attempt failed on.

1. You can name which purpose(s) this capability serves, and the sponsor agrees — and if you're establishing or rebooting, you're starting with one.
2. There is a named executive sponsor who will defend the budget through a leadership change — not just the one who chartered it.
3. You can answer “what would the enterprise do if this capability didn't exist?” with something concrete, not a platitude.
4. The team knows it advises; decision rights and outcome accountability sit with the executives it serves.
5. Success is measured by decisions influenced and transformations delivered, not documents produced.
6. A charter states the extent of overlap with other functions, the shared responsibilities, and what EA will explicitly not do.
7. Buy-in was secured before value delivery — financial allocation and resource commitment are real, not assumed.
8. The roadmap is run like a real project: milestones, deliverables, and someone accountable for keeping it true to plan.
9. The capability is scoped to a purpose, not the org chart — the next reorganization shouldn't break it.
10. Someone tracks which point-of-view documents get read, and by whom — the earliest leading indicator of real influence.

Going deeper

Sources worth your time, roughly in the order a team should read them.

• The TOGAF® Leader's Guide to an EA Capability (G168)The Open Group guide this note is drawn from — establishing and evolving an EA Capability. Start here.
• The TOGAF® StandardThe architecture method itself — the framework the Leader's Guide interprets for practitioners.
• TOGAF Standard, 10th Edition — IntroductionA current, navigable overview of the ADM and the architecture domains referenced throughout.
• The Open Group Architecture ForumWhere the TOGAF standard and its companion guides are maintained and evolved.
• Association of Enterprise Architects (AEA)The professional body for EA practitioners — behind much of the research on EA success and failure.
• MIT CISR — Ross & Weill on EAHome of the research behind Enterprise Architecture as Strategy — tying EA to an operating model and business execution.
• A Digital Product-Centric Reference Operating Model (D-PROM)The Open Group's 2022 model for the shift from project/silo IT to product-centric operating — where a named purpose matters most.