If you are an Architect who worked in large Enterprises, you will likely come across Enterprise Architecture Frameworks. Two popular ones are TOGAF , and Zachman . TOGAF is the more modern and open one I have experience with, although I contest that EA frameworks have generally been less useless than everyone thought they would be.
Most of the EA frameworks' ideas make sense and are undoubtedly helpful for someone getting started in managing a large enterprise. There is little formal discipline around IT Architecture, especially from a process perspective, and these frameworks help introduce some process rigor instead of accidental decision making.
Shift from Rigor to speed
Technology organizations have moved away from multi-year planning to incremental experimentation. It has become increasingly clear that there is no right decision in IT, there's the only right decision at a given time, and the smaller the steps are, the easier it is to correct.
It is much more rewarding for business and IT organizations to move fast , experiment, and correct as needed. Most of the multi-year plans are stale quickly because of the speed of innovation. Planning do not guarantee success howsoever rigorous you claim it to be.
A high production rate solves many ills— Elon Musk (@elonmusk) March 30, 2021
Poor communication tool - Those who matters most can't understand it
EA framework artifacts like Archimate diagrams (a diagramming language from TOGAF ) and other framework-specific architecture diagrams are not well understood by the stakeholders outside architecture.
If you are pitching an executive on a technology strategy, you will have to create a custom diagram/chart that speaks the language of the executive and explain it from their vantage point.
Decision making pushed to the edges
Many technology decisions that were centralized earlier are pushed to the edges; three forces are driving this
- Cultural, process shift like DevOps providing complete autonomy to the teams
- Everything as code - brings domains like infrastructure and middleware into the developer's toolset
- Cloud - Services are easily accessible; expanding to additional capabilities can be done without going through the complex procurement process.
Innovation from the Commons
Most of the innovative technologies are being built by developers for developers. These are not traditional IT Companies like IBM or Oracle trying to expand revenue but Organizations like Netflix , Square, and Google trying to solve their business needs.
In the old world, new technologies and solutions were "sold" to the architects and executives first, which was then "rolled out" across the Enterprise. The best and brightest solutions and ideas are sprinkled across many open source projects in Github, and solution advice is delivered through Stackoverflow, Hacker News, Reddit, or Twitter. The notion of specialties in IT has diluted with full-stack teams, and Lead Architects are often struggling to catch up with the influx of technologies coming at them.
My point is not to claim that Enterprise Architecture as a discipline is useless, but heavyweight enterprise Architecture frameworks are less effective than they used to be. No one cares if you have a great EA team and process but if you cannot respond quickly to the business's needs and attract the best talent.