Cracking the last silo

Enterprises know the ability to adapt quickly is essential, but can that agility extend to the core, proprietary systems that define a company?

Cracking the last silo
Thinkstock

Ever since the dot-com boom, enterprises have been trying to figure out how to reorganize themselves to meet the demands of a fast-changing, perpetually connected world. One model that many companies are finding persuasive is based on the workflow for the design, development, and deployment of software.

According to Dominic Price, a self-styled “work futurist” and head of R&D for the dev tool firm Atlassian, the basic unit of this new organizational model is the “two-pizza team,” a phrased coined by Jeff Bezos more than a dozen years ago. In a visit to InfoWorld’s offices last week, Price described the success he’s seen from applying this model to his own company.

Why a team no larger than two pizzas can feed? In Bezos’ original conception, one big reason was to cut down on the number of people team members needed to communicate with on a regular basis in order to get their jobs done. Reducing dependencies on communication to a manageable level boosts productivity.

Then there’s team loyalty, focus, and job satisfaction. In a recent survey of 2,000-plus U.S. employees that Price helped develop, respondents said they were more motivated by their team’s success than their company’s — or even their own. No less than 78 percent of those surveyed felt their primary team was very or extremely effective, and 69 percent trusted their teammates “very much” or “completely.”

As it happens, small teams align with modern web and mobile software architecture, where applications are broken into discrete services which can then be reassembled into other applications. Each service, usually accessible via RESTful API, has its own two-pizza team to maintain and revise it over time. Each team has the resources to dev, test, and deliver — and often deploy on its own. Continuous monitoring of the service provides the metrics and analysis on which a team’s continuing improvements are based.

But as Price notes, a model this decentralized can’t work without “strong North Star goals.” Leadership needs to make clear what the overall objectives and standards are for individual projects and the company. Plus, “guardrails” must be in place for security, avoiding duplication of effort, and so on.

What impact does this model have on the enterprise as a whole? Last week I also spoke with Bill Briggs, CTO of Deloitte Consulting and co-author of the firm’s annual Tech Trends report. This year’s edition talks about the “kinetic enterprise, an idea that describes companies that are developing the dexterity and vision required not only to overcome operational inertia, but also to thrive in a business environment that is, and will remain, in flux.”

The notion that enterprise IT should provide a platform for continuous change is a point Briggs and Price agree on. But from my own observation, in most enterprises the reorganization of work that Price describes has mainly been around the development and delivery of customer-facing web and mobile applications. As Briggs puts it: “It’s really the idea that I’m going to understand at a personal level my customers, their interests, behavior, their needs, their history. I’m going to have a human side of design that’s compelling and engaging and I’m going to be delivering things increasingly dynamically to them as individuals.”

Such systems of engagement lend themselves to Price’s future-of-work model. That’s in part because those applications need to change most often and they’re not mission-critical. If a recommendation service on an e-commerce site goes down for a few minutes, for example, the world isn’t going to end. But when you get into vertical software that controls the mixture of ingredients for pharmaceuticals or runs testing routines for jet engines, eventual consistency is not going to work. Ironclad processes need to be in place, not only guardrails.

These proprietary systems are what Briggs calls “the digital core.” They tend to resist change because they require precision and specialized knowledge that creates its own, often insular, culture. But Briggs thinks the digital core can be opened up, beginning with accessing the data created by those systems and applying new analytics technologies. “There’s so much richness in this information,” he says. We have to “challenge these very arbitrary and artificial boundaries.”

What Briggs calls the digital core needs the ability to adapt more quickly than it has in the past for enterprises to keep up with the pace of change. That saga will take a long time to play out — and won’t result in the entire world being reorganized into two-pizza teams.

But over time, integrating the digital core with the fast-paced technology development and innovative work models described by Price is the key to moving the entire enterprise forward at a rapid pace, not only the millennials building their web and mobile stuff. In most companies, those turf battles are simply the beginning, but how they play out will be crucial to whether enterprises can respond effectively to accelerating change.