by Ethan Pack

The enterprise architecture paradox

Feb 19, 2018
Enterprise ArchitectureTechnology Industry

When done properly, EA can be instrumental to unlocking the value in all initiatives—digital or otherwise.

abstract system architecture
Credit: Thinkstock

The much-anticipated The Cloverfield Paradox was recently released on Netflix.  While the surprise trailer and announcement received much praise, the movie itself was widely (and sometimes, thoroughly) panned.  I enjoyed it.  (I have an admittedly low bar for what constitutes an enjoyable movie.)  Regardless, I do like the idea of an anthology of loosely coupled concepts around a central theme.  It provides the freedom of countless creative options, but it also introduces the challenges of continuity and connectivity within the series over various timelines and contexts.

I also like paradoxes.

On the professional front, I’ve noticed that enterprise architecture (EA) is making quite the comeback.  According to a 2017 Harvey Nash/KPMG CIO survey, “architecture staged a comeback after several years of decline.”  A decline?  Then, why the sudden 26-percent uptick from the 2016 survey?  And, what were the causes of any decline?  A potential paradox, indeed.

While the enterprise architecture practice generally began in the IT unit at most organizations, enterprise architecture-like activities were also happening elsewhere.  Regardless of where EA originates within a given company, enterprise architects have historically struggled, not truly knowing and architecting their enterprise.

In the world of EA before the 21st century, companies were able to successfully operate in functional and departmental silos, so there was no real penalty to EA efforts that didn’t cover the enterprise.  Architects who only saw the IT areas—applications, technology/infrastructure, and data (sometimes, but usually not elevating into information)—were unable to connect the architecture to operational or financial results.  Many of us are familiar with the “you aren’t/weren’t invited because this isn’t/wasn’t an IT-related discussion” challenges that have hindered getting visibility into the human and machine interactions.  Conversely, EAs coming from the perspectives of business strategy or business process were not able to see, relate to, or did not have much interest in the IT-specific complexities and opportunities.

As the business and operational impact of technology continued to evolve, business leaders started seeing the state of their fractured enterprises and holding enterprise architecture departments to account.  Enterprise architecture efforts were being disbanded—increasingly so as Agile and Scrum trends were on the rise.  (I contend that Agile methodologies like Scrum only really work at the intersection of a supporting organizational culture and underlying architecture.  It is not sustainable in the long term if either of these two supports begin to diverge.  Agile works really well when there is little or no legacy architecture involved.)

It’s hard to justify your existence when your role or title reflects “enterprise architecture,” yet you were unable (for any reason) to actually do so and demonstrate business value.  That’s why the decline occurred.  But, we have to also consider the fact that there wouldn’t be any decline in EA at organizations that already figured out how to get value out of it.

Surviving EA practices did so by realizing what Robert Frost once wrote: “[People] work together whether they work together or apart.”  They were able to successfully identify the person-to-person, person-to-machine, and machine-to-machine linkages present across their enterprise.  They were able to successfully communicate these connections and how changes thereto can enable better business results.  They were able to continuously evolve their organization’s operating model to capitalize on technology’s evolution and societal changes.  (And, they did so with more agility and much less internal friction.)

The apparent flash-point of technological ubiquity simultaneously shines a spotlight on “the enterprise” while challenging the very definition of it.  Digital expands the reach of the enterprise and its surrounding ecosystem.  This brings some clarity to the interconnectedness of people, process, and technology inside the enterprise, but exposes new and yet-more-opaque areas of external business integration, omnichannel, chatbots, RPA-enabled business processes, a world of “things,” and AI.

Yes, this amorphous thing we call “digital transformation” is too an anthology of loosely coupled concepts.  Gartner’s research on what they refer to as the “digital business technology platform” reinforces this new landscape by laying out components like customer portals and apps, API management, core and back-office systems, analytics, the various “things” in the growing IoT, and algorithm and AI engines.

This brings us to the uptick in enterprise architecture.  I think organizations now have a better idea of what they’re looking for from an EA.  When done properly, it can be instrumental to unlocking the value in all initiatives—digital or otherwise.   Enterprise architecture becomes extremely irrelevant when it is only producing diagrams and documentation disconnected from successful business results.

So, was the digital transformation trend borne out of past success in EA efforts to meet technology-led change or is the future success in EA efforts a result of addressing digital disruption and transformation?  I believe the answer is a paradoxical knot of two yeses.  EA was able to sustain missing the enterprise in its past, but the digital imperative brings a Cloverfield monster-sized challenge.