Why Enterprise Architecture Still Matters: ROI Arguments, Misconceptions, and a Path Forward

When a large enterprise asked for proof of the ROI of re‑implementing Enterprise Architecture (EA) after discarding its ERP‑centric information system, the author faced the challenge of convincing senior management of EA's strategic value.

The company had previously hired architects but chose a massive ERP solution as the backbone of its IS, leading to the belief that a default, ERP‑driven EA would suffice, with competitive advantage based on cost reduction and operational productivity.

After a recent business and IT review identified ERP as a bottleneck for new business lines, the organization decided to overturn ERP in many strategic IS areas, yet it lacked dedicated EA architects; project leads were making ad‑hoc architectural decisions, and the EA ROI request was intended to secure sponsorship to overcome operational resistance.

The author’s response was not a dismissive "are you kidding?" but a serious attempt that ultimately failed to win the deal.

He reflects on three failed arguments: (1) the belief that without an architect, IS architecture can still succeed; (2) the assumption that benefits from ERP‑based EA will automatically transfer when ERP usage declines; and (3) the expectation that C‑level executives will inherently support EA without concrete evidence.

The difficulty of quantifying EA’s ROI stems from its abstract nature; while overall ROI for new business strategy can be estimated, isolating the EA contribution is challenging, especially when using default or poorly managed EA methods.

EA costs are primarily labor (architecture work, composition, governance, support, communication) and tool expenses, plus indirect costs such as delayed time‑to‑market.

Even a default EA incurs these costs, and without managed decision‑making, organizations cannot effectively build their information systems, leading to hidden expenses and distrust in EA.

Common misunderstandings include the notion that architects must perform all architectural work and the "ivory‑tower syndrome" where architects produce artifacts only for themselves, resulting in costly, misaligned outputs.

To revitalize EA, the author suggests achieving 80% of EA benefits with only 20% of the cost by re‑establishing strategic control over architecture decisions and avoiding decentralized local decisions.

He proposes using the 5W2H framework (What, Why, Where, When, Who, How, How Much) to clarify the purpose, scope, and cost of architectural changes before major transformations, aligning with TOGAF’s Architecture Vision phase but focusing on the overall enterprise rather than isolated architecture artifacts.

Finally, he emphasizes that EA should be treated as the glue linking business needs, SI products, people, and resources, enabling the company to achieve its next strategic objectives.