Main Takeaway from Enterprise Architecture as a Strategy: Part One

June 20, 2026

I’m currently reading Enterprise Architecture as a Strategy by Jeanne W Ross, Peter Weill, and David C Robertson. I’ll likely be splitting my main takeaways into thirds, because the book is quite information dense.

If you’re interested in finding the other main takeaways, you can click on the enterprise architecture as a strategy tag listed above.

With that said, let’s get to it:

1. Enterprise Architecture (a strategy for business execution) is a business issue driven by Business Process Integration and Business Process Standardization, both of which happen to be enabled via technology. It takes the form of a high-level architecure aligning the company around a shared operational vision, and requires a shared understanding of how people, processes, and technology converge around data.

2. Executing via an Enterprise Architecture involves three steps which occur in this order:
   1. Defining the business Operating Model: Where Business Integration and Business Process Standardization takes place
   2. Designing the Enterprise Archicture: Translates Operating Model into clear design
   3. Adopting the business IT Engagement Model: Governance system that aligns local project decisions with company-wide objectives
3. The book defines four standard Operating Models:
   1. Diversification Model: Low Integration / Low Standardization. Business units run more or less independently. Minimal standardization and little to no data sharing.
      • To Build: Identify shared technologies for standardization and compliance.
   2. Coordination Model: High Integration / Low Standardization. Products / services offered without forcing standardization. Data shared across business units to present unified “common face” to customer.
      • To Build: Identify key segments / channels to be shared across business units, plus the necessary data to be shared to serve them and the technology required to do so, then consider whether any business process elements mst be included.
   3. Replication Model: Low Integration / High Standardization Standardized business processes replicated across the organization with business decisions made locally by each business unit.
      • To Build: Identify the key processes to be standardized and the technologies automating these processes, then identify linking technolgoies across units.
   4. Unification Model: High Integration / High Standardization Standard processes access shared data to make products / services available to customers.
      • To Build: Identify the key segments / channels the company serves, list key processes to be standardized and integrated, identify shared data required to link processes, show the technologies that either automate or link these processes.

   Trade-offs:

   • Process Standardization: Reduces variability, and increases throughput but limits local innovation
   • Data Integration: Links organization units together but increases overhead (increased complexity and risk of data governance).

4. Once the operating model of the business is determined, and the enterprise architecture is developed, IT will then engage with this high level architecture to develop separate infrastructure diagrams that map out granular applications, data, information, and technology based on company goals, on a project basis through a governed process.

5. This three step process (define the Operating Model, create the Enterprise Architecture, adopt an IT Engagement Model) is all done in the name of creating a strong foundation for the enterprise to deliver on its stated value or pivot gracefully. It does not need to happen all at once. It can be implemented over the course of multiple projects, one project at a time.

   Warning signs of a weak foundation often include:

   • Inconsistent answers given to the same customer by different parts of the company
   • Massive infrastructure investments to meet new regulatory or reporting needs
   • Lack of business agility where every new initiative must start completely from scratch
   • Missing data necessary for making critical product or customer decisions

Until next time.

Keith