As enterprises splinter, so does enterprise architecture (but in a good way)

Enterprise architecture (EA) is entering a new phase, driven by the ongoing splintering of enterprises. The discipline has been carefully crafted over the past couple of decades as a process-driven, well-documented exercise intended to develop a roadmap to prevent organizations from spending big money on technologies that don't produce value for the business.

This model of EA worked well within established enterprises that viewed IT as a budget item that needed to be justified and monitored -- and constantly explained to executives. Nowadays, however, IT is integral to everything the business does, and is viewed as a strategic weapon. This moves EAs out of their roles as "explainers" and into more activist, "driver" roles.

Tom Graves, a self-identified "enterprise architect, business-anarchist, confusionist,"whose work I greatly respect, says a major disruption is reshaping the role of EA and the mission of enterprise architects. In a recent post, he explains what's happening to the discipline as of late:

A shift away from big EA: EA has developed, especially in the US, almost as an engineering specialty that is a component of a large corporate culture. "Probably the only viable way to link everyone and everything together is through formal documents and definitions," Tom muses, citing a statement from leading EA advocate John Zachman that "engineering an enterprise is the same as engineering an aircraft." A side effect of big EA, of course, that it "also drives everything towards hierarchical, top-down, Waterfall, and organization-centric," Tom warns.

Many EA initiatives going forward, however, may require a rethinking of what Tom calls "small-country" cultures - which I assume is both in the literal and figurative sense. That is, adopting the techniques used in companies in smaller nations, but also smaller companies with more distributed or entrepreneurial cultures. These are cultures that emphasize collaboration or consensus among teams versus top-down dictates, and are likely not to have large EA staffs or support functions. "There simply isn't time to produce, digest and use documents," Tom explains. This needs to be reflected in EA toolsets as well, since most are configured for the larger EA frameworks.

Cloud is driving the new EA: Increasing adoption of cloud services -- versus building on-premises systems -- requires a new EA emphasis. "Increasing use of public cloud lessens the importance of in-house big-IT," Tom says. In turn, this demands a greater understanding by IT and their EA counterparts of the business implications and customer impacts to technology investments. The extended, connected organization is creating "increasing complexities in supply-chains and value-webs mandate integration across the whole end-to-end flows, requiring interweaving, rather than separation, of all IT-based, machine-based and human-based elements of flows," he states.

Enterprise architects need to start thinking outside the framework box: "The mainstream EA frameworks that we now have are almost perfectly unsuited for the real EA contexts that we now face," Tom explains, noting that many EAs from this point forward will be forced to "unlearn just about everything they've been taught about EA and start again from scratch."

Business and IT are no longer separate worlds that EAs need to bridge: Something interesting has happened over the last five years: IT has become the business, and the business has become information technology. The anointed role of EAs has been to bridge these two worlds, but as they become one, this requires EA take a more proactive role higher up the organizational hierarchy -- perhaps interacting directly with the CEO and CFO. Tom cites developments such as cloud-based shadow-IT, virtualization and serverless computing, the Internet of Things, mobile computing, artificial intelligence, and robotics as elevating technology into the boardroom.

Another piece of this evolution is the need for the EA role to encompass the evolution of work and skills requirements in organization. "Without appropriate balance on these themes, the vaunted AI/robot future risks fizzling out for lack of skills to build or maintain it; burning out in destructive fashion from failure to prepare for unintended consequences; or triggering an all-too-understandable Luddite-style reaction," Tom warns.