Postmodern Software

One of our portfolio CEO’s made me laugh — and think — this week by describing a new investment opportunity I’m pursing as “very postmodern”.

As an undergrad American Studies / Lit geek I spent many happy hours puzzling over the ambiguous and perspective-shifting fiction of writers likeĀ Don DeLillo and Donald Barthelme, but before this week’s conversation I hadn’t heard the “postmodernist” label applied to the current state of the software business.

Of course, as soon as I typed the phrase “postmodern software” into Google, I turned up a fascinating set of citations that traced the term as far back as 2009 (with a focus on software development methods and processess), but progressing to more recent citations that describe the entire polyglot IaaS / SaaS / PaaS landscape as “postmodernist”.

The idea hit the mainstream in 2011, when Gartner analyst Anne Thomas Manes asserted in an industry presentation (at Gartner’s 2011 AADI Summit) that contemporary IT architecture design solutions require “postmodern” principles. Here’s the description from the conference agenda for her track:

“Externalization, consumerization, and democratization are disrupting business-as-usual in IT organizations. Postmodern IT requires a conscious break with the past and the rejection of old narratives. This Summit track explores the latest trends in application architecture.”

There’s more than tongue-in-cheek lit-geekery at work here: the enterprise software business is currently undergoing a rapid phase of disaggregation and abstraction, both “vertically” (in the form of multi-platform / polyglot Infrastructure- and Platform-as-a-Service environments), and “horizontally” (from integrated, single-vendor enterprise systems to highly fragmented, multi-vendor “best of breed” SaaS environments).

As the pace of software innovation and disruption continues to accelerate, even technically competent enterprises are losing the battle for “single perspective” centralized IT control and provisioning regimes — organizations just can’t ship software fast enough to meet customer demand without loosening their grip on platform selection, tooling and supporting infrastructure services among their many divisions and business units.

What does “enterprise IT” even mean when every user — both technical and non-technical — is empowered to chose her own software tools, vendors and and services from the open source / cloud supermarket?

As with literary postmodernism, only the reactionaries believe the user-centric genie can be put back in the bottle. The functional needs of centralized Enterprise IT — security/access control, provisioning, business intelligence and cost management — aren’t going away, but successful IT managers will have to embrace radically new methods of delivering these services that actually enable agile, user-driven self-provisioning rather than impeding it.

As an investor, I’m excited to see so many entrepreneurs tackling the knotty technical problems embodied in the user-driven enterprise. Companies like AppFog (a Founders Co-op portfolio company), and Cloudability (a TechStars Cloud and PIE PDX grad) are lighting the way for me on this topic, but I’m actively on the lookout for more.

  • http://www.verespej.com/ Hakon Verespej

    What a great thought-provoking commentary. What do you see as some of the high-level thinking you encourage entrepreneurs to consider in this regard? Some things that come to mind might be that there will be more small highly customized businesses, think more about the end user even when building for enterprise, completely new opportunities now exist in terms of new needs for security, management, auditing, etc.

    • http://www.crashdev.com Chris DeVore

      Thanks for the note Hakon + hope you’ve been well. The companies I see doing interesting work on this theme are thinking about how to enable distributed + user-driven software environments while still allowing the enterprise to be managed as an enterprise (i.e., through some level of unified auth / data / analytics). Lots of hard technical problems here, but also a ton of value to be created.